ND color correction in DV?
Avid Fan
image was cooled out, light was very blue and other colors were washed
out. How do I correct the color and still benefit from the ND? I
tried putting a hoya 85 on as well, but the result was unnatural. Any
suggestions? I'm fairly new to this, so any advice would be
appreciated.
Thanks,
Avid Fan
C2001
I use Tiffen's ND3 with the Sony VX1000, in addition to the camera's
built-in ND filter (I think the XL1 has a built-in ND also). The results
are *spectacular* and I have never experienced difficulty with accurate
color balance using the ND3.
That extra dose of ND _really_ stretches the camera's contrast range,
reduces noise and (IMO) enhances color acuracy and related subtlety.
If you don't have difficulty balancing your XL1 in other situations, you
might try balancing it before you install the ND. I'm more inclined to
think, however, that the ND6 is a bit too strong on top of the camera's
built-in filter.
Charles
Alexander Ibrahim
you have applied.
If you change filters...do another white balance. Heck, even if you swap to
another supposedly identical filter, do a manual white balance.
FWIW ND filters are not supposed to alter color balance, but they often do
anyway, especially on cheaper filters. In practice small deviations in
manufacturing often create a color cast, hence the above advice.
When doing your white balance, use a quality photo-white card. Kodak makes a
set that includes 2 8x10 cards and a smaller card. They are 18% grey on one
side and 90% white on the other site. In a pinch these can also be used as a
reflector.
(Kodak Catalog Number E152 7795)
--
Alexander Ibrahim
http://www.zenera.com
I decline unsolicited commercial email (SPAM)
If you send it to me I will invoice you $500
"Avid Fan" <krey...@hotmail.com> wrote in message
news:46DFCA74D117AD08.80F6E898...@lp.airnews.net...
Sswitaj
Avid Fan <krey...@hotmail.com> writes:
>When using an XL-1, I applied a tiffen nd6 filter and the resulting
>image was cooled out, light was very blue and other colors were washed
>out.
What's your iris at? Was there room to open up two stops before you started? Is
it possible that you're underexposed now, pusing the camera in to a high gain
situation?
- Steve
R.J.F. Stewart
> I use Tiffen's ND3 with the Sony VX1000, in addition to the camera's
> built-in ND filter (I think the XL1 has a built-in ND also). The results
> are *spectacular* and I have never experienced difficulty with accurate
> color balance using the ND3.
>
> That extra dose of ND _really_ stretches the camera's contrast range,
> reduces noise and (IMO) enhances color acuracy and related subtlety.
Can you explain how the use of an ND filter can reduce noise in a
television camera, and have you patented the method yet? As a television
engineer for 33 years, I'd be very interested to know this.
Rod.
Alexander Ibrahim
writers observation.
OTOH, if you drop the gain on an XL-1 to -3dB then you do increase SNR to
around ~56dB, so I always do that before I drop an ND filter on. It even
lets me lower my f-stop so I can get shorter depth of field...how
convenient.
If you plan on using depth of field to direct audience attention, you had
better be at f3 or less on MiniDV cams like the XL-1. You probably need to
back off from the subject a bit too. Even at f1.6 my XL-1 can give deep
focus. Since I have no budget, my sets are a disaster...so I am counting on
the blurred background to save me...
MiniDV does not always save you money. If I was not ready or able to move
the camera for depth of field, I'd have to use color to direct attention AND
have great sets.
Shooting a quality movie on MiniDV is just as hard, and sometimes harder
than shooting 35. That is why it sucks so often.
--
Alexander Ibrahim
http://www.zenera.com
I decline unsolicited commercial email (SPAM)
If you send it to me I will invoice you $500
"R.J.F. Stewart" <escap...@08002go.com> wrote in message
news:VA.0000004...@08002go.com...
David Mullen
below maximum lens aperture and have to start boosting the signal.
The only affect it has is on depth-of-field, which can also affect lens
resolution (some lenses are softer at maximum or minimum aperture) and
APPARENT contrast -- not true contrast. Meaning that the more the background
goes out-of-focus, the lower in contrast it appears. Also, colors become
softer & less harsh (and less attention-grabbing) as they go out-of-focus.
ND filters are extremely useful in film & video photography in order to
control what f/stop you shoot at and how much depth-of-field you are
getting. But they don't actually change the dynamic range of the image nor
do they improve noise; however, by throwing more of the frame out-of-focus,
they can make certain telltale artifacts less visible, thus giving the
illusion of things like less contrast.
David Mullen
C2001
> I'm not an electronics engineer, but I'd speculate that the reduction of noise
> (and it's not my imagination; I did back-to-back testing) is simply because
> smoothing the contrast ratio reduces stress on the image sensors. Most
> videographers are well aware that too little *and* too much light result in
> video noise.
>
> On dazzle-bright days, I found that using the Tiffen HD3 wiht the VX1000
> created a markedly more film-like look, simply because the camera could handle
> contrast so much more effectively -- the really bright spots never burned yet
> the shadow detail was also tremendous.
Neglected to mention that I started experimenting with additional ND filtration
after reading about John Seale's use of ND filters for "The English Patient" in
American Cinematographer.
C.
C2001
I'm not an electronics engineer, but I'd speculate that the reduction of noise
(and it's not my imagination; I did back-to-back testing) is simply because
smoothing the contrast ratio reduces stress on the image sensors. Most
videographers are well aware that too little *and* too much light result in
video noise.
On dazzle-bright days, I found that using the Tiffen HD3 wiht the VX1000
created a markedly more film-like look, simply because the camera could handle
contrast so much more effectively -- the really bright spots never burned yet
the shadow detail was also tremendous.
Charles
David Mullen
noise
>(and it's not my imagination; I did back-to-back testing) is simply because
>smoothing the contrast ratio reduces stress on the image sensors. Most
>videographers are well aware that too little *and* too much light result in
>video noise.
>
>On dazzle-bright days, I found that using the Tiffen HD3 wiht the VX1000
>created a markedly more film-like look, simply because the camera could
handle
>contrast so much more effectively -- the really bright spots never burned
yet
>the shadow detail was also tremendous.
I'm sorry, but this sounds a little like psuedo-science here. Think about
what happens in the camera when using an ND filter.
The amount of light coming OUT of the rear of the lens, past the lens iris,
and finally hitting the CCD's -- whether you used an ND.90 and shot at an
f/2.8 or no ND filter and shot at f/8 -- IS THE SAME. The intensity is the
same and the brightness of highlights & shadows is the same. The contrast
is the same, except for the fact that out-of-focus areas will be lower in
contrast due to the blending of bright & dark -- and the wider aperture
version allows more of the frame to be out-of-focus. There also might be
more lens flare when shooting wide open that will reduce some of the black
levels.
Lower depth-of-field in video also mimics the depth-of-field of 35mm more,
which is generally lower due to the longer focal lengths necessary -- hence
the greater "film look". But this is not because contrast is lowered.
The dynamic range of the video image is NOT increased by use of ND filters,
unless without ND filters, you cannot expose the image properly. However,
lower depth-of-field might create a more out-of-focus background, which will
LOOK lower in contrast. But the in-focus subject has the same contrast and
the dynamic range permitted by the CD's & recording format is not changed by
the use of ND filters, other than some possible loss of black levels from
lens flare from shooting wide open on some lenses.
ND filters are not magic -- they just cut the overall amount of light. They
don't compress or extend contrast. The only issue is how much the image is
changed by being able to shoot at wider lens apertures, and how that affects
PERCEIVED contrast. But on the in-focus subject, you won't find that bright
highlights are suddenly less clipped or that shadow detail is suddenly
extended through the use of ND filters, assuming that you are still exposing
for the same IRE level on the subject (in other words, you aren't actually
underexposing highlights in one version versus another).
Pola filters used as ND filters are another issue, since they can affect
glare off of highlights...
David Mullen
C2001
David, at one point in your post you say the light coming out of the rear of the
lens is the same _with or without an ND filter_, *then* a few paragraphs later
you state that ND filters "cut the overall amount of light".
They may not directly compress or extend the contrast of the light going through
the lens, but the way they work definitely makes it easier for the CCD to handle
high-contrast and/or dazzle-bright outdoor light -- and I have the A/B
comparison tapes to prove this.
Polarizers (depending on how they are implemented) may reduce reflection, but
this is not necessarily the same as glare -- ie, a reflection is not
_automatically_ glare; OTOH glare is usually the result of a surface that is
reflecting a high level of ambient (usually sun) light. Also, polarizers may
screw up color accuracy and/or consistency.
When polarizers are used to reduce or eliminate reflections the reason usually
is to eliminate an unwanted reflection rather than to eliminate glare.
As I mentioned, I decided to experiement with additional ND filtration after
reading about the use of ND filters for the desert sequences in "The English
Patient".
Charles
R.J.F. Stewart
> David, at one point in your post you say the light coming out of the rear of the
> lens is the same _with or without an ND filter_, *then* a few paragraphs later
> you state that ND filters "cut the overall amount of light".
I think you do really know what he means. Of course ND filters cut the overall
amount of light. So does the iris, but it has the side-effect of altering depth of
field as well. Provided there is sufficient light available, the two methods can
used in combination to adjust depth of field for artistic reasons. What *does*
remain constant - or at least within a limited range - is the amount of light
falling on the light-sensitive CCD in order to maintain signal levels within the
required limits. No adjustment is made to the gain of the electronics, which is
why I would like to know what mechanism is alleged to affect the noise level,
because I can't think of one.
Rod.
Alexander Ibrahim
ND filters do lower the overall amount of light, if you do not alter
exposure elsewhere.
However the reason to use ND filters is, as David said, so that you can
control your aperture better. You want a lower f2.8 for certain film effects
and to avoid that deep focus look, as when you have no money for sets.
In fact I rarely wish to use more than ~f3. Like David said, using the same
lighting conditions a ND.90 filter at f2.8 looks like no filter at f8.
Notice the difference in exposure settings ?
As to noise...it changes relative to camera gain levels. On may cameras you
can not set this manually. Camera's like the XL-1 have a setting knob that
allows me to switch gain to -3dB, 0dB, +6dB, +12dB and Automatic Gain
Control. (AGC) Holding this setting at any given number will give you the
same noise levels. On less controllable cameras, it is set automatically,
and using an ND filter may well create MORE noise...
--
--
Alexander Ibrahim
http://www.zenera.com
I decline unsolicited commercial email (SPAM)
If you send it to me I will invoice you $500
"C2001" <charle...@sympatico.ca> wrote in message
news:3B154774...@sympatico.ca...
David Mullen
through
>the lens, but the way they work definitely makes it easier for the CCD to
handle
>high-contrast and/or dazzle-bright outdoor light -- and I have the A/B
>comparison tapes to prove this.
The CCD doesn't know that an ND filter is being used -- for example, whether
or not the camera has an ND.9 filter and shoots at f/2.8 or no ND filter and
shoots at f/8, the amount of light reaching the CCD is the same. So why
would the same amount of light reaching the CCD now suddenly be easier for
the CCD to handle???
The only difference between the two situations is the f/stop being used. So
then you have to believe that the f/stop has an affect on contrast. So then
you have to believe that the size of the hole at the back of the lens
affects the dynamic range of the image -- and that somehow now that a
highlight over 100 IRE and therefore clippped when shooting at the deeper
stop is somehow under 100 IRE and not clipped in the wider aperture.
If you looked at a waveform monitor, you would see that the brightest and
darkest areas of the frame were at the same video levels in both
scenarios -- while if the ND filtration were truly causing a lowering of
contrast, the range would be compressed. Instead, what you see when
shooting with ND filters and working at a wider aperture is that the
out-of-focus areas are causing the signal to put more information in the
middle range -- because blurred objects sort of blend into a medium
intensity -- but the extreme ENDS of the signal are at the same point. And
the contrast of the in-focus area, if one could eliminate the out-of-focus
background info from the signal, would look the same.
Unless you can come up with a plausible explanation as to why there should
be a decrease in the dynamic range of the image when using an ND filter and
opening up the iris. Like I said, there can be some loss of blacks at wide
apertures due to flare, and more of the image is out-of-focus. This can
lead to a more pleasing image with less perceived harshness, but you haven't
really compressed the dynamic range of the in-focus areas of the image. You
made areas of the frame look less contrasty, you maybe have lifted some of
the blacks slightly, depending on the amount of flare, and therefore there
will be the perception that now the image is lower in contrast. But for the
in-focus areas of the frame, the video levels are the same unless you change
the exposure beyond the adjustment for the filter.
David Mullen
C2001
I've read the three replies to my earlier post with interest and respect and,
frankly, since (among other considerations) this NG is primarily concerned with
film photography rather than video, I don't wish to protract an apples/oranges
argument in which I sense different people are saying essentially the same
things, but using different wording, or else imputing different contexts and/or
perspectives and therefore imagining they are being somehow contradicted.
First (and for the record), using the VX1000, with the camera's built-in ND
filter turned on, _and_ an additional Tiffen ND3, exposure (bright outdoors,
that is) is pretty well limited to f8 and f9.6 (f11 will be underexposed, unless
for some reason you want crushed blacks and zero shadow detail).
I'm fully aware that ND filtration affects depth of field (and, accordingly, in
some situations overall image sharpness [depending on the image itself and how
much detail is included]). However, this is *not* the difference to which I was
referring (never mind John Seale, in discussing his work for "The English
Patient" [for the bright, _very_ glary desert shoots, he used moderate ND
filtration for earlier morning and later afternoon and high levels of ND
filtration for mid-day and mid-afternoon; I don't recall offhand whether or not
he used polarizers)]).
Nevertheless, your assumption that the "same" amount of light reaches the CCD
(or film stock) with and without ND filtration because of adjustments in
exposure is, IMO, simplistic.
Careful use of ND filtration (for both film and, at least, high-end prosumer and
industrial video cameras [and the VX1000 arguably eclipses most analog
industrial cameras]) enables the film or CCD to handle extremes in contrast, or
just extreme brightness (such as strong sun reflection from sand, snow, or even
light-colored stone or concrete buildings) more effectively. Judicious use of
ND filtration will results in "bright" areas of the image that will _not_ "clip"
(to use video jargon) or burn, but will still look genuine and not
underexposed. Simultaneously, there will be more shadow detail in darker areas
of the same frame (if the shot encompasses darker areas).
It was back in October, 1997, that I took my camera out one Saturday afternoon,
because it was an exceptionally beautiful autumn day and I suspected (correctly)
it would be the last good weekend of the year (we had snow four days later) for
outdoor shooting at certain favourite sites. I was fairly pleased with portions
of the tape but _did_ notice some glare and burning (probably "clipping") white
levels when I watched it. I had recently read the article in "AC" about Mr.
Seale's work for "The English Patient" and since I'd bought (mainly on a hunch)
a Tiffen ND3, decided the next day (Sunday) to repeat the shoot, but with the
extra filter. First, it didn't take me long to realize exposure had to be
radically adjusted. But I could tell even from the viewfinder that I was
getting a different look from the previous afternoon. (And in case anyone
wonders, weather and light were literally identical and I was shooting both
afternoons from about 2:00 pm to 5:00 pm.)
When I viewed the second tape I noticed immediately that there was no glare from
even the brightest surfaces (mainly, road pavement [which from certain angles
may be highly reflective] and light-toned buildings, especially hi-rises), and
-- always -- greater shadow detail. I don't own (or require) gear to make
precise video noise measurements (my eyes are pretty good at this), and I also
noticed most of the second tape tended to have a smoother, more silky texture
which I was aware indicated an improvement in luma and chroma noise levels.
Perhaps the most interesting single shot was one, facing north, of a not overly
high sandstone church's bell tower, side/back lit by the sun. On Day One, to
make the texture and color of the sandstone (which had no direct illumination)
discernable, I had to open the exposure to a point where the background blue sky
was overexposed and was bleached white.
With the Tiffen ND3 on Day Two, I got beautiful sandstone *and* the beautiful
blue sky in the background.
To be frank, I simply don't care about technical arguments as to whether or not
there is or was more or less or the same amount of light reaching my camera's
CCD when I did these two shoots (or at least certain very closely matched shots
from them): the bottom line is that using the ND3, exposure latitude *and*
chroma range were dramatically improved and I'm pretty sure most experienced
video or film photographers would agree the Day Two tape produced finer
results. (And no one would require a vectorscope to confirm the difference!)
Charles
C2001
You cannot get blood from a stone and I think that (especially for an electronics
engineer) you are being simplistic.
"Gain" is only one factor in the noise-contributing equation (though I'm well aware
many users of the VX1000 always use the reduced gain setting when shooting outdoors in
bright light [I'd certainly recommend this if you don't have ND filtration over and
above the camera's biult-in filter]).
A CCD will register video noise and generate undesirable artifacts when _too_ much or
_too_ little light is being supplied to it. This has nothing to do with any
adjustment in"gain", but with changes in light levels.
Charles
C2001
> David is right, you are mistaken somewhere in your testing.
>
> ND filters do lower the overall amount of light, if you do not alter
> exposure elsewhere.
>
> However the reason to use ND filters is, as David said, so that you can
> control your aperture better. You want a lower f2.8 for certain film effects
> and to avoid that deep focus look, as when you have no money for sets.
>
> In fact I rarely wish to use more than ~f3. Like David said, using the same
> lighting conditions a ND.90 filter at f2.8 looks like no filter at f8.
> Notice the difference in exposure settings ?
You are focussing (sorry -- bad pan [<g>]) on only *one* of numnerous scenarios
for using ND filtration.
As I've indicated before, my "testing", which you regard as dubious, was
originated from reading about the work of one of today's formost
cinematographers (I believe John Seale knows a lot more about photography than
any pro broadcast engineer [at least if typical broadcast NTSC for the past 30
years is a measure of quality!]).
> As to noise...it changes relative to camera gain levels. On may cameras you
> can not set this manually. Camera's like the XL-1 have a setting knob that
> allows me to switch gain to -3dB, 0dB, +6dB, +12dB and Automatic Gain
> Control. (AGC) Holding this setting at any given number will give you the
> same noise levels. On less controllable cameras, it is set automatically,
> and using an ND filter may well create MORE noise...
Obviously, you cannot use ND filters with a video camera that functions only in
auto mode and -- obviously -- this is not what I was discussing.
Charles
David Mullen
CCD
>(or film stock) with and without ND filtration because of adjustments in
>exposure is, IMO, simplistic.
It IS simple. An ND.9 filter cuts out three stops of light. If you adjust
by opening up the iris by three stops, the same amount of light passes to
the CCD or film, unless further altered by shutter speed. What else do you
think an ND filter does beside cut the overall level of light?
>Careful use of ND filtration (for both film and, at least, high-end
prosumer and
>industrial video cameras [and the VX1000 arguably eclipses most analog
>industrial cameras]) enables the film or CCD to handle extremes in
contrast, or
>just extreme brightness (such as strong sun reflection from sand, snow, or
even
>light-colored stone or concrete buildings) more effectively.
How? Assuming that you are EQUALLY adjusting the f/stop to account for the
light loss of the ND filter. If the brightest area is three stops
overexposed and you use an ND.9 (three stops loss) and open up the iris
three stops to compensate, in order to get the same exposure on the main
subject, are you saying that the area that was three stops overexposed is
now only two stops overexposed?
>Judicious use of
>ND filtration will results in "bright" areas of the image that will _not_
"clip"
>(to use video jargon) or burn, but will still look genuine and not
>underexposed. Simultaneously, there will be more shadow detail in darker
areas
>of the same frame (if the shot encompasses darker areas).
How is this possible? A simple grey filter -- "neutral density" meaning
that it cuts all color wavelengths equally, thus just cutting the overall
amount of light passing through -- exactly how does that cause the dynamic
range of the image to become compressed? How does it make the brightest
areas darker and the darkest areas lighter, but leave the middle the same
brightness relative to each other?
Let's say a face spot meters at f/16 with no filter for a "correct"
exposure, and you're placing it at 70 IRE. You have a white object that is
110 IRE and is being clipped. You put in an ND.9, which cuts three stops of
light, and open up the same amount, to f/5.6. So once again, the face is at
70 IRE. Now you're telling me that the white object is no longer at 110
IRE, but has somehow dropped to, let's say, 98 IRE???
>I had recently read the article in "AC" about Mr.
>Seale's work for "The English Patient" and since I'd bought (mainly on a
hunch)
>a Tiffen ND3, decided the next day (Sunday) to repeat the shoot, but with
the
>extra filter.
"The English Patient" HAD to use ND filtration -- he shot the whole movie on
500 ASA stock (5298) in bright sunlight, requiring heavy ND and polas to
boot just to work at a normal f/stop range. Plus John Seale is a big user
of ND grad filters to control AREAS within the frame (like a hot sky). I
read the article too -- plus a lot more about John Seale over the years (in
fact, I've read every issue of "American Cinematographer" back to 1927...)
>To be frank, I simply don't care about technical arguments as to whether or
not
>there is or was more or less or the same amount of light reaching my
camera's
>CCD when I did these two shoots (or at least certain very closely matched
shots
>from them): the bottom line is that using the ND3, exposure latitude *and*
>chroma range were dramatically improved and I'm pretty sure most
experienced
>video or film photographers would agree the Day Two tape produced finer
>results. (And no one would require a vectorscope to confirm the
difference!)
It sounds clearly to me that you used an ND.3 filter without compensating
for the one stop light loss, and therefore brought down your highlights to a
lower level, thus allowing you to hold more info in the brightest areas and
hold chroma better. Shadow detail would be darker as well, but since video
has better latitude for underexposure, compensation might have not been
necessary. This is a common trick -- a slight overall underexposure -- for
video photography. It doesn't necessarily mean that the image LOOKS
underexposed.
Or simply you had more sunlight than you could expose properly for, and
using an ND.3 thus brought it down to the range that your lens could handle.
Thus before you used the ND filter, you were not correctly exposing the
image to begin with (let's say that your lens has a minimum aperture of f/16
but the light was an f/22 -- thus without the ND.3 filter, your lens could
only stop down to f/16 and everything would be one stop overexposed. By
adding the ND.3, now the image is being correctly exposed at f/16.)
ND filters are used everyday in video and motion picture photography, either
to allow filming in light that is too bright for the lens to compensate for,
or to control the amount of depth-of-field. They are a valuable tool and
people shooting video should own them and learn to use them. But they
should also understand WHAT the filter does exactly. It does NOT change the
dynamic range of the image, other than how depth-of-field can change
perceived contrast. It is simply designed to cut the amount of light
passing through it. How you compensate for that or use it creatively is up
to you.
If you think that an mild ND.3 filter (only ONE STOP of light loss!) can
dramatically change the latitude and color reproduction of a video camera,
you really don't understand the basics of photography or exposure. You
could have gotten the same results simply by stopping down the lens. And if
there was too much light to allow you to stop down anymore, then the ND
filter was simply cutting the light level so you COULD expose correctly. It
wasn't changing the dynamic RANGE from dark to light -- it was cutting the
overall level. The range remained the same -- if it were ten stops from the
brightest to the darkest object, it was still ten stops. The ND filter
wouldn't have altered it to a nine stop range, let's say.
ND Grad filters, on the other hand, CAN bring down bright areas along one
side of the frame to a level closer to the other areas, and thus affect the
dynamic range of the image.
David Mullen
David Mullen
>originated from reading about the work of one of today's formost
>cinematographers (I believe John Seale knows a lot more about photography
than
>any pro broadcast engineer [at least if typical broadcast NTSC for the past
30
>years is a measure of quality!]).
Let me put it this way -- I don't know as much as John Seale does, but I
have photographed twenty feature films, 17 in 35mm, 1 in Super-16, and 2 in
high-definition video. So I suspect that there is a possibility that I
understand what a ND filter does to the image...
Are you aware that John Seale is not alone in using ND filters, but that
they are possibly the most COMMONLY USED filter on film & video productions?
Also, rereading the article in A.C. on "The English Patient" (January 1997),
nowhere does he claim that ND filters alter the dynamic RANGE of the
image -- only that they drop the light level down to a practical range for
his lenses. "...he used Kodak 5298, throwing in ND filters during the day
to bring down the ASA. 'I generally use an 85ND9. If it is still too
bright, I'll go to an 85ND12, which will get me down to between 40 and 60
ASA, a very usage range for daylight exteriors.' "
What he means is that 500 ASA is too fast for daylight levels, and thus
using ND filters brings the light levels down to a range that he can
correctly expose for. An 85 filter cuts 2/3 of a stop (thus down to 320
ASA) and ND9 cuts down three stops of light, so down to 40 ASA.
In direct sunlight in the desert, 40 ASA at 24 fps / 180 degree shutter
often gets you an f/11 to f/16. So sometimes he switched to an 85ND12,
which would get him 20 ASA / f/8 to f/11 -- but since he also exposing more
for the shade than the direct sun, he was probably opening up even more.
And he might have also been rating his stock slightly slower than 500 ASA as
a base sensitivity.
All of this has nothing to do with contrast and dynamic range, just overall
light levels. Although overexposing the negative and printing down does
affect contrast. But this is a film technique that doesn't work for video,
where overexposure is bad.
David Mullen
C2001
> >As I've indicated before, my "testing", which you regard as dubious, was
> >originated from reading about the work of one of today's formost
> >cinematographers (I believe John Seale knows a lot more about photography
> >than any pro broadcast engineer [at least if typical broadcast NTSC for the
> >past 30 years is a measure of quality!]).
>
> Let me put it this way -- I don't know as much as John Seale does, but I
> have photographed twenty feature films, 17 in 35mm, 1 in Super-16, and 2 in
> high-definition video. So I suspect that there is a possibility that I
> understand what a ND filter does to the image...
You are discussing film -- I am discussing video. (HiDef video is still an
emerging technology which, as you well know, is nowhere near being
standardized.)
> Are you aware that John Seale is not alone in using ND filters, but that
> they are possibly the most COMMONLY USED filter on film & video productions?
Of course (I wasn't born yesterday).
> Also, rereading the article in A.C. on "The English Patient" (January 1997),
> nowhere does he claim that ND filters alter the dynamic RANGE of the
> image
I never said that he did.
> -- only that they drop the light level down to a practical range for
> his lenses. "...he used Kodak 5298, throwing in ND filters during the day
> to bring down the ASA. 'I generally use an 85ND9. If it is still too
> bright, I'll go to an 85ND12, which will get me down to between 40 and 60
> ASA, a very usage range for daylight exteriors.' "
>
> What he means is that 500 ASA is too fast for daylight levels, and thus
> using ND filters brings the light levels down to a range that he can
> correctly expose for. An 85 filter cuts 2/3 of a stop (thus down to 320
> ASA) and ND9 cuts down three stops of light, so down to 40 ASA.
*EXACTLY* my point in video context! A film stock being too fast for daylight
levels is very parallel, if not identical, to the limited ability of most video
CCDs to handle really bright light levels. So what on earth are you arguing
about? Your quote from "AC" supports my own observations 100%.
It was my own intellectual curiosity that prompted me to see if something that
worked in a film context might also work for video -- and to my delight I
discovered that it _did_. I don't need unrelated technical mumbo jumbo, or
esotieric engineering explanations, or a @#$%^&* vectorscope (!!) to confirm
what my own eyes (and others') confirmed in an instant. (BTW, when I did my
experiment back in 97, my reference set at the time was an ISF-calibrated Sony
XBR100 [the results are even more obvious on my 20" Sony PVM [also
ISF-calibrated] using component inputs from my Sony DSR-80 VTR or SDI from my
Avid Xpress.)
> In direct sunlight in the desert, 40 ASA at 24 fps / 180 degree shutter
> often gets you an f/11 to f/16. So sometimes he switched to an 85ND12,
> which would get him 20 ASA / f/8 to f/11 -- but since he also exposing more
> for the shade than the direct sun, he was probably opening up even more.
> And he might have also been rating his stock slightly slower than 500 ASA as
> a base sensitivity.
>
> All of this has nothing to do with contrast and dynamic range, just overall
> light levels. Although overexposing the negative and printing down does
> affect contrast. But this is a film technique that doesn't work for video,
> where overexposure is bad.
Again, you continue to argue apples and oranges. Light levels *directly* impact
on requirements for contrast range in both film and video. You are now
contradicting yourself and (frankly) rather pathetically grasping at straws
because you realize (you're no fool) you've thoughtlessly boxed yourself into a
corner from which there is no easy, face-saving escape.
And why do you suddenly mention overexposing the negative and printing down as a
film technique that doesn't work for video? Of course I know about this, but it
has nothing to do with this particular discussion of ND filters -- it smacks of
another desperate red herring (on top of apples and oranges).
Is there are chef here who'd like to create a gourmet salad with apples, oranges
and red herring? We could serve it at an outdoor barbeque on a bright sunny day
and video it with ND filters!
<g>
Charles
C2001
> >Nevertheless, your assumption that the "same" amount of light reaches
> >the CCD (or film stock) with and without ND filtration because of
> >adjsutments in exposure is, IMO, simplistic.
>
> It IS simple. An ND.9 filter cuts out three stops of light. If you adjust
> by opening up the iris by three stops, the same amount of light passes to
> the CCD or film, unless further altered by shutter speed. What else do
> you think an ND filter does beside cut the overall level of light?
It doesn't perhaps *directly* do anything more -- BUT, because it cuts down
overall light level, it enables alternative flexibility in other aspects of the
image capture process. Since I've already explained this several times, I'm not
interested in repeating what I know I have described very clearly. Moreover,
since you have *not* seen the tapes in question you are being very presumptuous
(and rather rude) in your arguments.
> >Careful use of ND filtration (for both film and, at least, high-end
> prosumer and
> >industrial video cameras [and the VX1000 arguably eclipses most analog
> >industrial cameras]) enables the film or CCD to handle extremes in
> contrast, or
> >just extreme brightness (such as strong sun reflection from sand, snow, or
> even
> >light-colored stone or concrete buildings) more effectively.
>
> How? Assuming that you are EQUALLY adjusting the f/stop to account for the
> light loss of the ND filter. If the brightest area is three stops
> overexposed and you use an ND.9 (three stops loss) and open up the iris
> three stops to compensate, in order to get the same exposure on the main
> subject, are you saying that the area that was three stops overexposed is
> now only two stops overexposed?
Perhaps -- I'm not interesting in arguing about whether or not my combination of
ND filtration and exposure settings do or don't match (from a textbook/academic
view) what I'd use without ND filtration.
> >Judicious use of
> >ND filtration will results in "bright" areas of the image that will _not_
> "clip"
> >(to use video jargon) or burn, but will still look genuine and not
> >underexposed. Simultaneously, there will be more shadow detail in darker
> areas
> >of the same frame (if the shot encompasses darker areas).
>
> How is this possible? A simple grey filter -- "neutral density" meaning
> that it cuts all color wavelengths equally, thus just cutting the overall
> amount of light passing through -- exactly how does that cause the dynamic
> range of the image to become compressed? How does it make the brightest
> areas darker and the darkest areas lighter, but leave the middle the same
> brightness relative to each other?
I never claimed I knew *how* this was possible (again, I'm not an electronics
engineer). I've seen the results first hand (which you, David, have NOT --
which gives very limited credibility to your arguments).
> Let's say a face spot meters at f/16 with no filter for a "correct"
> exposure, and you're placing it at 70 IRE. You have a white object that is
> 110 IRE and is being clipped. You put in an ND.9, which cuts three stops of
> light, and open up the same amount, to f/5.6. So once again, the face is at
> 70 IRE. Now you're telling me that the white object is no longer at 110
> IRE, but has somehow dropped to, let's say, 98 IRE???
This is not the scenario I was discussing and (unlike you) I have no interest in
arguing about matters where I know I don't have direct experience.
> >I had recently read the article in "AC" about Mr.
> >Seale's work for "The English Patient" and since I'd bought (mainly on a
> >hunch) a Tiffen ND3, decided the next day (Sunday) to repeat the shoot,
> >but with the extra filter.
>
> "The English Patient" HAD to use ND filtration -- he shot the whole movie on
> 500 ASA stock (5298) in bright sunlight, requiring heavy ND and polas to
> boot just to work at a normal f/stop range. Plus John Seale is a big user
> of ND grad filters to control AREAS within the frame (like a hot sky). I
> read the article too -- plus a lot more about John Seale over the years (in
> fact, I've read every issue of "American Cinematographer" back to 1927...)
I don't question your expertise on film technology. But it's very obvious (even
from the credentials you have indicated) that your experience with video is
quite limited.
> >To be frank, I simply don't care about technical arguments as to whether or
> >not there is or was more or less or the same amount of light reaching my
> camera's
> >CCD when I did these two shoots (or at least certain very closely matched
> shots
> >from them): the bottom line is that using the ND3, exposure latitude *and*
> >chroma range were dramatically improved and I'm pretty sure most
> experienced
> >video or film photographers would agree the Day Two tape produced finer
> >results. (And no one would require a vectorscope to confirm the
> difference!)
>
> It sounds clearly to me that you used an ND.3 filter without compensating
> for the one stop light loss, and therefore brought down your highlights to a
> lower level, thus allowing you to hold more info in the brightest areas and
> hold chroma better. Shadow detail would be darker as well, but since video
> has better latitude for underexposure, compensation might have not been
> necessary. This is a common trick -- a slight overall underexposure -- for
> video photography. It doesn't necessarily mean that the image LOOKS
> underexposed.
This may or may not be an accurate explanation of why I was able to achieve
dramatically better results with additional ND filtration on bright days. Your
last sentence is interesting because the 'look' I accomplished (which I've used
heavily now for almost four years) is somewhat similar to (at least my)
perception of wearing sunglasses.
> Or simply you had more sunlight than you could expose properly for, and
> using an ND.3 thus brought it down to the range that your lens could handle.
Exactly -- without the ND3 the VX1000 can handle extreme bright light fairly
well, but there is still some burning which gives the image a more familiar
'video' look. It's one reason the VX1000, for all its merits, still isn't the
DSR300 or DigiBeta.
>Thus before you used the ND filter, you were not correctly exposing the
Correctly? Have you spent more than 60 hours shooting with a VX1000?
> image to begin with (let's say that your lens has a minimum aperture of f/16
> but the light was an f/22 -- thus without the ND.3 filter, your lens could
> only stop down to f/16 and everything would be one stop overexposed. By
> adding the ND.3, now the image is being correctly exposed at f/16.)
The VX1000 is limited to f11. Perhaps you regard this as "incorrect"? If so,
please take this up with Sony -- not me.
> ND filters are used everyday in video and motion picture photography, either
> to allow filming in light that is too bright for the lens to compensate for,
> or to control the amount of depth-of-field. They are a valuable tool and
> people shooting video should own them and learn to use them. But they
> should also understand WHAT the filter does exactly. It does NOT change the
> dynamic range of the image, other than how depth-of-field can change
> perceived contrast. It is simply designed to cut the amount of light
> passing through it. How you compensate for that or use it creatively is up
> to you.
I can't argue here since this is precisely what I do. I daresay that with a
more advanced camera ND filtration requirements would be reduced and/or
different.
> If you think that a mild ND.3 filter (only ONE STOP of light loss!) can
> dramatically change the latitude and color reproduction of a video camera,
> you really don't understand the basics of photography or exposure.
Now you are being condescending and offensive. I'm not John Seale or Vitorio
Storaro, but I know more than enough about the basics of photography to
appreciate and (with proper explanations [which is why I read "AC"]) understand
most if not absolutely all that they do.
> You could have gotten the same results simply by stopping down the lens.
You are 100% WRONG about this. I did stop down the lens for my Day One shoot
and the results, while OK, were often glary and not nearly as good as what I got
with the Tiffen ND3.
One of the people who marvelled at this result was Jennifer Dickson, one of
Canada's pre-eminent artist-photographers (also etcher and water colorist).
Jennifer is the only Canadian artist ever elected to Britain's Royal Academy and
she knows as much about the basics of photography as anyone. She has an
international reputation and her works are in major museum collections
world-wide, including New York's Met.
> And if
> there was too much light to allow you to stop down anymore, then the ND
> filter was simply cutting the light level so you COULD expose correctly. It
> wasn't changing the dynamic RANGE from dark to light -- it was cutting the
> overall level. The range remained the same -- if it were ten stops from the
> brightest to the darkest object, it was still ten stops. The ND filter
> wouldn't have altered it to a nine stop range, let's say.
All the ND filter did was enable settings that enabled the camera to produce
video that very clearly had more contrast range. Since you haven't seen the
video, you aren't in a position to argue with me. Throwing (in several
instances) off-topic technical mumbo-jumbo at me is pointless. You're starting
to sound like the psychiatrists who for decades insisted that MS was nothing
more than an exotic form of repression-induced hysteria. They couldn't accept
anything they didn't understand, or which fell outside their comfortable pews of
familiar frames of references.
> ND Grad filters, on the other hand, CAN bring down bright areas along one
> side of the frame to a level closer to the other areas, and thus affect the
> dynamic range of the image.
I haven't tried ND grad filters. I'd be interested but would prefer to avoid
scenarios that require them.
I think it's time to go to bed......
Charles
David Mullen
daylight
>levels is very parallel, if not identical, to the limited ability of most
video
>CCDs to handle really bright light levels. So what on earth are you
arguing
>about? Your quote from "AC" supports my own observations 100%.
That wasn't your point at all! We weren't talking about simply lowering the
overall light LEVEL -- you were suggesting that using an ND filter altered
the CONTRAST RANGE of the image so that both bright highlights AND dark
shadows were compressed into a range that the CCD could handle better. John
Seale merely used ND filters because his film stock was too fast to be able
to record an image without them. He wasn't altering his contrast through
use of ND filters.
If all you are saying is that sometimes it is too bright outside to be able
to correctly EXPOSE an image without ND filtration, then that's rather
obvious. But you have been suggesting that the ND filter somehow compresses
the dynamic RANGE so that both highlight AND shadow information are now
within a narrower range that that CCD can handle better -- not that ND
filters lower the overall level so that you can expose the picture correctly
because without it, you're beyond the ability of the lens to be able to stop
down enough.
>Again, you continue to argue apples and oranges. Light levels *directly*
impact
>on requirements for contrast range in both film and video. You are now
>contradicting yourself and (frankly) rather pathetically grasping at straws
>because you realize (you're no fool) you've thoughtlessly boxed yourself
into a
>corner from which there is no easy, face-saving escape.
Overall light level is a different issue than contrast range. Using ND
filters affects all light levels EQUALLY -- it does not change the RATIO
between the brightness of the highlights versus the shadows. Their relation
to each other remains the same -- if the key was four stops brighter than
the fill, it stays that way no matter what ND filter you are using. If a
bright object was five stops hotter than the main subject you're exposing
for, using an ND filter does not drop that difference down to only four
stops.
The only way that "light levels directly impact on requirements for contrast
range in both film & video" is if you change the light level of one portion
of the image but not another -- for example, if you added more or less fill
light but kept the key light the same. Changing the overall light level by
use of an ND filter does not alter the contrast range.
I keep saying the same things over and over - I haven't contradicted myself
and I haven't boxed myself in at all.
What exactly do you do for a living? And how come you know so much about
shooting hi-def video as to believe that the rules for how ND filters affect
the image are different than for standard-def video? I've also shot
professional NTSC video in beta-SP, Mini-DV, and Digital Betacam -- just not
entire features in standard-def, which is why I didn't mention it.
All I'm asking is for a logical explanation from you as to how an ND
filter -- essentially a grey piece of glass -- changes the contrast of
image -- the range from the hottest highlight to the darkest shadow -- and
not just the overall illumination level. How does it take a range that the
camera can't handle -- let's say that's 12 stops from bright to dark -- and
make something less extreme, like only 11 stops? It doesn't. If you use an
ND filter to bring down the brightness overall -- basically to reduce the
overall exposure, just like stopping down -- then you will hold more
information in the brightest areas but you will lose information in the
darkest of areas.
It's becoming clear to me that your using an ND.3 filter to simply knock
down the exposure, and thus bring down everything equally and hold detail in
bright objects and color information better -- but for some reason, you
think you are actually increasing the latitude of the camera.
It's also clear that the difference between the concept of overal
illumination level and that of contrast ratio is unclear to you, since you
keep confusing them. ND filters affect the first but not the second.
I've tried to explain the same thing in as many ways as possible -- I
suppose that IS pathetic and "grasping at straws", since clearly you're in
no mood to learn something. I'm just wasting my time.
I just hate misinformation passed on the Internet SO much that I feel
compelled to stamp it out when I see it, before someone else is corrupted by
it. I can only hope that others who have been reading this exchange have
learned something. What particular irritates me is the use of John Seale's
name to back up your theories, when he makes no such claims in the article
you keep pointing out. So not only are you passing along misinformation,
you compound it by misunderstanding and misrepresenting someone else's
comments as well.
I'm only sorry that my teaching skills are not good enough to explain this
subject more clearly. If anyone out there has been confused, I apologize.
If someone needs some clarification, they can contact me directly.
David Mullen
David Mullen
electronics
>engineer). I've seen the results first hand (which you, David, have NOT --
>which gives very limited credibility to your arguments).
>Have you spent more than 60 hours shooting with a VX1000?
Let me get this straight -- you're assuming that I've never used an ND
filter on a video camera and looked at the results???
60 hours -- that's only five days of full-time work.
As for hours put in, I've put in about 500 hours on a Sony HDW-F900 -- but I
suppose that doesn't count as video experience to you since it is hi-def
video. As for my hours with standard-def video gear, it's been on & off for
the past twelve years so I'm not in a mood to add it up. I have shot about
24 hours on a Sony VX-1000, and about 100 hours on a Canon XL-1, and
hundreds of hours on beta-SP gear. Yes, I mainly shoot film but I wouldn't
say I'm a novice at video. Unlike you, I assume, I don't primarily earn my
living by just shooting video, although my last two feature films have been
in hi-def video, so it's shifting that way.
>I don't question your expertise on film technology. But it's very obvious
(even
>from the credentials you have indicated) that your experience with video is
>quite limited.
I've practically given you all my credentials & credits, and my full name is
at the bottom of all my posts. Perhaps you'd care to do likewise? As for
my video work, a digital feature I shot called "Jackpot" will be released by
Sony Pictures Classics theatrically on July 27 so maybe you can check it
out. There is also an article on it in the newest RES Magazine.
David Mullen
R.J.F. Stewart
> A CCD will register video noise and generate undesirable artifacts when _too_ much or
> _too_ little light is being supplied to it. This has nothing to do with any
> adjustment in"gain", but with changes in light levels.
My understanding of "noise" in this context is unwanted random voltage variations,
originating not from the input but from the electronics itself. How does it vary within
the electronics of a camera with variations in light level? Do you also think it varies
if we use optical means to maintain the electronic signal levels within limits, which
would effectively maintain light levels on the CCD within limits too?
Rod.
R.J.F. Stewart
> Nevertheless, your assumption that the "same" amount of light reaches the CCD
> (or film stock) with and without ND filtration because of adjustments in
> exposure is, IMO, simplistic.
If the output signal level is the same, and no electronic adjustments have been
made, then the light falling on the CCD *must* be the same. It's more than
*simplistic* - it really is simple. A TV camera is an electronic device that is
sensitive to light, like a lightmeter. If you don't alter its sensitivity, then
ther same output "reading" indicates the same light level input.
The photographic character of the image will be affected by the use of irises
and filters, but as long as the signal level is the same, the *amount* of light
must be the same. Likewise the signal to noise ratio, which is what started
this discussion.
Rod.
R.J.F. Stewart
> As I've indicated before, my "testing", which you regard as dubious, was
> originated from reading about the work of one of today's formost
> cinematographers
Please explain how you test for noise levels, and what figures you get both with
and without ND filters.
> (I believe John Seale knows a lot more about photography than
> any pro broadcast engineer [at least if typical broadcast NTSC for the past 30
> years is a measure of quality!]).
He might do, but does he know how to measure signal to noise ratio in a
television camera?
Rod.
CT
> David Mullen wrote:
>
>
> *EXACTLY* my point in video context! A film stock being too fast for
daylight
> levels is very parallel, if not identical, to the limited ability of most
video
> CCDs to handle really bright light levels. So what on earth are you
arguing
> about? Your quote from "AC" supports my own observations 100%.
>
> It was my own intellectual curiosity that prompted me to see if something
that
> worked in a film context might also work for video -- and to my delight I
> discovered that it _did_.
?!? You had to do a test to find this out? Did you think camera
manufacturers include ND options on the filter wheel for a laugh?
> I don't need unrelated technical mumbo jumbo, or
> esotieric engineering explanations, or a @#$%^&* vectorscope (!!) to
confirm
> what my own eyes (and others') confirmed in an instant.
Vectorscope? What would you expect that to confirm in this situation?
CT
<snip>
So what you are basically in you're saying in your mass of posts is that to
you, footage shot with your VX plus an ND3 *looks* like it has more contrast
and better chroma in high light-level situations. Wow, objective stuff,
thanks for enlightening us. You didn't think - as a professional - looking
on your waveform monitor would have given you some insight into what was
going on?
You obviously also can't be bothered to listen to "off-topic technical
mumbo-jumbo" from David. Funny, to me it looked like a well-reasoned and
accurate argument against a perceptual mis-conception that you're unable to
explain and too ignorant or stuborn to let go of.
There mere fact that you think it is mumbo-jumbo is very revealing.
C2001
> >I never claimed I knew *how* this was possible (again, I'm not an
> electronics
> >engineer). I've seen the results first hand (which you, David, have NOT --
> >which gives very limited credibility to your arguments).
>
> >Have you spent more than 60 hours shooting with a VX1000?
>
> Let me get this straight -- you're assuming that I've never used an ND
> filter on a video camera and looked at the results???
It doesn't sound as if you have used a VX1000 with additional ND filtration.
You should know better than I that different cameras can vary enormously in
performanceand capability. Experience with one (or even a large number) does
not qualify you to make domatic, blanket statements about gear and
configurations with which you apparently have not acquired experience.
> 60 hours -- that's only five days of full-time work.
Now you are being cheap and petty.
> As for hours put in, I've put in about 500 hours on a Sony HDW-F900 -- but I
> suppose that doesn't count as video experience to you since it is hi-def
> video.
HiDef video *is* a radically different medium compared with estanablished
standard def NTSC.
> As for my hours with standard-def video gear, it's been on & off for
> the past twelve years so I'm not in a mood to add it up. I have shot about
> 24 hours on a Sony VX-1000, and about 100 hours on a Canon XL-1, and
> hundreds of hours on beta-SP gear. Yes, I mainly shoot film but I wouldn't
> say I'm a novice at video. Unlike you, I assume, I don't primarily earn my
> living by just shooting video, although my last two feature films have been
> in hi-def video, so it's shifting that way.
>
> >I don't question your expertise on film technology. But it's very obvious
> (even
> >from the credentials you have indicated) that your experience with video is
> >quite limited.
>
> I've practically given you all my credentials & credits, and my full name is
> at the bottom of all my posts. Perhaps you'd care to do likewise? As for
> my video work, a digital feature I shot called "Jackpot" will be released by
> Sony Pictures Classics theatrically on July 27 so maybe you can check it
> out. There is also an article on it in the newest RES Magazine.
My full name is part of my email address.
Good luck with "Jackpot".
Charles
C2001
> >*EXACTLY* my point in video context! A film stock being too fast for
> daylight
> >levels is very parallel, if not identical, to the limited ability of most
> video
> >CCDs to handle really bright light levels. So what on earth are you
> arguing
> >about? Your quote from "AC" supports my own observations 100%.
>
> That wasn't your point at all! We weren't talking about simply lowering the
> overall light LEVEL -- you were suggesting that using an ND filter altered
> the CONTRAST RANGE of the image so that both bright highlights AND dark
> shadows were compressed into a range that the CCD could handle better. John
> Seale merely used ND filters because his film stock was too fast to be able
> to record an image without them. He wasn't altering his contrast through
> use of ND filters.
I wasn't "suggesting" that use of ND filtration could alter the contrast range
for video in some circumstances (and, obviously, with some but not necessarily
all equipment), I was saying that my own experience clearly demonstrates that it
DOES. Since you have not seen the tapes which demonstrate this, you are in no
position to argue with me.
> If all you are saying is that sometimes it is too bright outside to be able
> to correctly EXPOSE an image without ND filtration, then that's rather
> obvious. But you have been suggesting that the ND filter somehow compresses
> the dynamic RANGE so that both highlight AND shadow information are now
> within a narrower range that that CCD can handle better -- not that ND
> filters lower the overall level so that you can expose the picture correctly
> because without it, you're beyond the ability of the lens to be able to stop
> down enough.
>
> >Again, you continue to argue apples and oranges. Light levels *directly*
> impact
> >on requirements for contrast range in both film and video. You are now
> >contradicting yourself and (frankly) rather pathetically grasping at straws
> >because you realize (you're no fool) you've thoughtlessly boxed yourself
> into a
> >corner from which there is no easy, face-saving escape.
>
> Overall light level is a different issue than contrast range.
Technically different but -- obviously -- related, as you (and everone else at
this NG) well know.
> Using ND
> filters affects all light levels EQUALLY -- it does not change the RATIO
> between the brightness of the highlights versus the shadows. Their relation
> to each other remains the same -- if the key was four stops brighter than
> the fill, it stays that way no matter what ND filter you are using. If a
> bright object was five stops hotter than the main subject you're exposing
> for, using an ND filter does not drop that difference down to only four
> stops.
>
> The only way that "light levels directly impact on requirements for contrast
> range in both film & video" is if you change the light level of one portion
> of the image but not another -- for example, if you added more or less fill
> light but kept the key light the same. Changing the overall light level by
> use of an ND filter does not alter the contrast range.
>
> I keep saying the same things over and over - I haven't contradicted myself
> and I haven't boxed myself in at all.
You have arrogantly stated that the result I achieved using the Tiffen ND3 with
a Sony VX1000 could have been accomplished without the filter but using a
difference exposure setting and YOU ARE DEAD WRONG. Try it yourself.
> What exactly do you do for a living?
I'm retired.
> And how come you know so much about
> shooting hi-def video as to believe that the rules for how ND filters affect
> the image are different than for standard-def video?
I never claimed expertise of any kind regarding HiDef shooting -- now you are
deliberately misrepresenting my statements, again in a desperate attempt to worm
out of a tight corner.
> I've also shot
> professional NTSC video in beta-SP, Mini-DV, and Digital Betacam -- just not
> entire features in standard-def, which is why I didn't mention it.
No amount of work means that you know everything there is to know about a
particular skill. My experience with videography (and editing [I'm a training
Avid Editor]) has taught me, if nothing else, the more you know, the more you
still have to learn.
> All I'm asking is for a logical explanation from you as to how an ND
> filter -- essentially a grey piece of glass -- changes the contrast of
> image -- the range from the hottest highlight to the darkest shadow -- and
> not just the overall illumination level.
I've already told you that I'm not an electronics engineer and don't consider
myself obligated to provide a full technical explanation.
> How does it take a range that the
> camera can't handle -- let's say that's 12 stops from bright to dark -- and
> make something less extreme, like only 11 stops? It doesn't. If you use an
> ND filter to bring down the brightness overall -- basically to reduce the
> overall exposure, just like stopping down -- then you will hold more
> information in the brightest areas but you will lose information in the
> darkest of areas.
>
> It's becoming clear to me that your using an ND.3 filter to simply knock
> down the exposure, and thus bring down everything equally and hold detail in
> bright objects and color information better -- but for some reason, you
> think you are actually increasing the latitude of the camera.
I "think" this because the resulting video very clearly demonstrates it.
Anyway, if ND filters did no more than change the exposure for a camera there
wouldn't be a great deal of use for them -- it would be easier all round simply
to manufacture cameras with a wider range of exposure.
> It's also clear that the difference between the concept of overal
> illumination level and that of contrast ratio is unclear to you, since you
> keep confusing them. ND filters affect the first but not the second.
I'm not confused at all (light levels and contrast ratios are pretty basic info
for all forms of photography). You are "confused" because you don't want to
believe or even consider something you haven't heard about before.
> I've tried to explain the same thing in as many ways as possible -- I
> suppose that IS pathetic and "grasping at straws", since clearly you're in
> no mood to learn something. I'm just wasting my time.
You haven't "explained" anything, David. You have merely ranted childishly
because someone has pointed out something you were unaware of and you don't
believe it can be true if you haven't learned it yourself.
> I just hate misinformation passed on the Internet SO much that I feel
> compelled to stamp it out when I see it, before someone else is corrupted by
> it. I can only hope that others who have been reading this exchange have
> learned something. What particular irritates me is the use of John Seale's
> name to back up your theories, when he makes no such claims in the article
> you keep pointing out. So not only are you passing along misinformation,
> you compound it by misunderstanding and misrepresenting someone else's
> comments as well.
I never misrepresented Mr. Seale's comments. I merely pointed out that reading
about his work sparked my own curiosity to experiment and, accordingly, learn
more.
What you call "misinformation" is simply your own arrogant, head-in-the-sand
ignorance and unwillingness to recognize that someone else may have discovered
something you didn't know about.
> I'm only sorry that my teaching skills are not good enough to explain this
> subject more clearly. If anyone out there has been confused, I apologize.
> If someone needs some clarification, they can contact me directly.
There are limits to what can be "taught" on the internet. Some things can only
be learned by on-hands demonstration.
You are trying to "stamp out" something you don't even have the capacity (or
will) to understand (or try to understand). Your mentality is that of a
mediocre academic and a borderline intellectual fascist ("If I don't like it or
know about it, it's no good").
I'll save you further trouble, David --
Plonk.
Charles
C2001
> "C2001" <charle...@sympatico.ca> wrote in message
>
Nothing -- I was merely joshing about some video people who are afraid to do
work of any kind without consulting one (it amuses me that vectorscope rhymes
with horoscope...).
Charles
C2001
> C2001 wrote:
> > As I've indicated before, my "testing", which you regard as dubious, was
> > originated from reading about the work of one of today's formost
> > cinematographers
>
> Please explain how you test for noise levels, and what figures you get both with
> and without ND filters.
I've already stated I don't have formal testing equipment, because I don't need it.
If video noise didn't *visibly* degrade an image, it wouldn't be a problem.
Measuring and testing equipment is used in the broadcast industry supposedly to
monitor and help ensure consistent signal output. I'd have to say most engineers
appear to be asleep at the wheel, so to speak.
> > (I believe John Seale knows a lot more about photography than
> > any pro broadcast engineer [at least if typical broadcast NTSC for the past 30
> > years is a measure of quality!]).
>
> He might do, but does he know how to measure signal to noise ratio in a
> television camera?
I don't enjoy being nasty, but really DO have to point out that the "quality" (if
this word can even be used) of most color NTSC broadcast television in North America
has been a sad joke from Day One. I'm quite sure there are at least a few pro
broadcast engineers who know what they are doing and I'd like to believe you are one
of them (the fact that you post to this NG suggests you are).
Even from the same station, noise levels, black level, white level, chroma
saturation and tint may vary considerably (sometimes drastically) for broadcasts of
different episodes of the same program. Try comparing a new episode of "Frasier"
with a repeat of an earlier one, as some stations do during prime time; it's often
hard to believe that the same crew and gear have been used on the show for eight
years now.
Charles
C2001
The electronics generate the noise -- but _only_ when required (usually because of
insufficient or, conversely, too much light) to perform beyond their range.
Garbage in, garbage out -- pretty basic.
Charles
Alexander Ibrahim
That isn't as accurate a test as you may think.
For one, the new shows are broadcast from digital disk recorders or direct
from satellite rather often, while re runs may be shown from Betacam SP or
even just Betacam recordings. Some stations even use U-matic!!
Also bear in mind that different operators are dealing with the stations
broadcast all the time. The night shift guy is almost always less
experienced, sometimes it is their first gig as Engineer. They sometimes are
told to just use the settings used for the first run broadcast, despite
alterations in the media...that makes it worse than dubbing the original and
resetting levels to the dub...
Ah, I am getting a bit off topic. The point is, you have several Engineers
at work and several sources for lots of material. Then their is the
difference between local and national adverts...lots and lots of issues.
You try matching levels from 3 different productions using different
equipment, different talent in front of and behind the camera, from
different sources and on different media.
You have 30-45 seconds...GO!
Not so easy after all eh ?
They don't call it Never the Same Color because of the engineers work day to
day. Tehy call it that because it takes a lot of TIME and KNOWLEDGE to get
it right, at least one of which is always in short supply. Given both, you
_can_ get very consistent color all the time from NTSC.
It is often judged unimportant to do so, so it often does not get done.
--
Alexander Ibrahim
http://www.zenera.com
I decline unsolicited commercial email (SPAM)
If you send it to me I will invoice you $500
"C2001" <charle...@sympatico.ca> wrote in message
news:3B166AF3...@sympatico.ca...
R.J.F. Stewart
> I wasn't "suggesting" that use of ND filtration could alter the contrast range
> for video in some circumstances (and, obviously, with some but not necessarily
> all equipment), I was saying that my own experience clearly demonstrates that it
> DOES. Since you have not seen the tapes which demonstrate this, you are in no
> position to argue with me.
Your experience, verbally reported by you, doesn't "demonstrate" anything to
anybody. It simply tells us that you personally are convinced of something that
you have subjectively observed, cannot explain, and for which you have given no
measured figures. If you hope to convince anybody else, something a bit more
scientific will be needed.
I have yet to hear how the use of an optical filter can change noise level of the
electronic signal, and now that someone else has thrown contrast ratio into the
discussion, I would be quite interested to know how you think it can change that
too. The extra light scatter in a filter placed in front of the lens can raise the
black level slightly (and sometimes affect black balance too), which if not
compensated for electronically might *appear* to be a change in contrast range,
but I don't know exactly what you are seeing. ND filters are useful in controlling
depth of field, but I always prefer to use the ones in the filter wheel *behind*
the lens if possible, because of these effects.
> I've already told you that I'm not an electronics engineer and don't consider
> myself obligated to provide a full technical explanation.
*Any* suggested explanation would be better than none at all...
Rod.
R.J.F. Stewart
> I've already stated I don't have formal testing equipment, because I don't need it.
> If video noise didn't *visibly* degrade an image, it wouldn't be a problem.
>
> Measuring and testing equipment is used in the broadcast industry supposedly to
> monitor and help ensure consistent signal output. I'd have to say most engineers
> appear to be asleep at the wheel, so to speak.
Well, you don't seem to think much of engineers, you don't think much of measuring
instruments, and you don't think much of the picture quality that results from not
using them. How can the rest of us who don't have your subjective judgement ever hope
to get anything right?
Rod.
R.J.F. Stewart
> The electronics generate the noise -- but _only_ when required (usually because of
> insufficient or, conversely, too much light) to perform beyond their range.
What do you think causes the noise to vary with light level, (assuming you make no
adjustment to the electronics)?
Do you think the noise increases or decreases with an increase in light level?
Rod.
C2001
I have already told you (and others) that I'm not an electronics engineer.
Why don't you answer your own question?
Charles
C2001
> In article <3B1666FB...@sympatico.ca>, C2001 wrote:
> > I wasn't "suggesting" that use of ND filtration could alter the contrast range
> > for video in some circumstances (and, obviously, with some but not necessarily
> > all equipment), I was saying that my own experience clearly demonstrates that it
> > DOES. Since you have not seen the tapes which demonstrate this, you are in no
> > position to argue with me.
>
> Your experience, verbally reported by you, doesn't "demonstrate" anything to
> anybody. It simply tells us that you personally are convinced of something that
> you have subjectively observed, cannot explain, and for which you have given no
> measured figures. If you hope to convince anybody else, something a bit more
> scientific will be needed.
Of course there are very real limits to what can be "demonstrated" about video and
film technology at an internet NG, since few of us have the opportunity to meet in
person or work together.
Contrary to what you say, I'm not trying to "convince" anyone of anything -- merely
sharing my own experience. This is not a courtroom or an operating theatre -- I don't
consider that people are obliged to provide ironclad scientific evidence for
everything they report and discuss on the internet. The reality is it simply isn't
possible. You are merely trying to discredit me because I don't happen to invest in
measuring equipment (which I don't require [my eyes and three fully calibrated Sony
PVM monitors are sufficient for my purposes]) and I've ruffled the feathers of one of
your cronies here.
If you'd care to travel to Ottawa, Canada, and bring your own measuring equipment with
you, you are welcome to do so.
> I have yet to hear how the use of an optical filter can change noise level of the
> electronic signal, and now that someone else has thrown contrast ratio into the
> discussion, I would be quite interested to know how you think it can change that
> too. The extra light scatter in a filter placed in front of the lens can raise the
> black level slightly (and sometimes affect black balance too), which if not
> compensated for electronically might *appear* to be a change in contrast range,
> but I don't know exactly what you are seeing.
Like I just said, if you want to see you'll have to visit. In the meantime you should
not discredit something which you have *not* seen.
> ND filters are useful in controlling
> depth of field, but I always prefer to use the ones in the filter wheel *behind*
> the lens if possible, because of these effects.
I agree with you here. With the VX1000, however, there is only one level of ND
filtration built in and it is fairly limited.
> > I've already told you that I'm not an electronics engineer and don't consider
> > myself obligated to provide a full technical explanation.
>
> *Any* suggested explanation would be better than none at all...
I've already told you I have no "suggested" explanation. If I were an electronics
engineer I might be able to make the "suggestion" you seek. However, even if I did
have substantial expertise and knowledge in this field, I would not go around
bickering with people about work I have not seen myself.
Charles
C2001
> You asked us to compare different runs of the same show on the same station.
> That isn't as accurate a test as you may think.
I know -- my whole point was the hopeless lack of accuracy *and* consistency* in
imaging standards from the broadcast TV networks in the US and Canada.
> For one, the new shows are broadcast from digital disk recorders or direct
> from satellite rather often, while re runs may be shown from Betacam SP or
> even just Betacam recordings. Some stations even use U-matic!!
And some veteran engineers (eg, the ones who've been working for 30 or more
years [<g>]) are still convinced U-matic and BetaSP looks better than digital.
> Also bear in mind that different operators are dealing with the stations
> broadcast all the time. The night shift guy is almost always less
> experienced, sometimes it is their first gig as Engineer. They sometimes are
> told to just use the settings used for the first run broadcast, despite
> alterations in the media...that makes it worse than dubbing the original and
> resetting levels to the dub...
>
> Ah, I am getting a bit off topic. The point is, you have several Engineers
> at work and several sources for lots of material. Then their is the
> difference between local and national adverts...lots and lots of issues.
>
> You try matching levels from 3 different productions using different
> equipment, different talent in front of and behind the camera, from
> different sources and on different media.
>
> You have 30-45 seconds...GO!
>
> Not so easy after all eh ?
Since we're getting somewhat off-topic, I'd like to add I question whether
mainstream HDTV telecasting will ever suceed as long as broadcast technical
standards remain so lax and generally erratic.
> They don't call it Never the Same Color because of the engineers work day to
> day. They call it that because it takes a lot of TIME and KNOWLEDGE to get
> it right, at least one of which is always in short supply. Given both, you
> _can_ get very consistent color all the time from NTSC.
I'm well aware that it's doable. It's just too bad that most networks, stations
and their technical staff very obviously don't give a *****!
> It is often judged unimportant to do so, so it often does not get done.
And all of this (which I'm well aware of) is the main reason I'm unable to take
too seriously arguments of *any* kind from a "pro broadcast engineer" (even more
so one claiming 33 years' experience).
Regarding "Frasier", I was referring to CanWest Global showing a new episode,
then a repeat from earlier during the same season, so the differences should not
have been as extreme as, say, comparing a new episode just downloaded with a
ratty, third or fourth generation, hacked-to-death tape used for syndication
from an episode produced several years previously.
Thanks for your informed, polite and non-confrontational comments and
clarifications.
Charles
C2001
Please see my response to Mr. Ibrahim's post. It's common knowledge that technical
standards at the major broadcast networks are a joke. At best everything sinks to the
lowest common denominator.
I'm not saying there are not at least a few pro broadcast engineers who know what they
are doing and what they are talking about -- only that, based on the quality of their
work as delivered by the networks and even the best digital satellite and cable servers,
they don't exactly command unquestioning respect.
I'm also well aware of the uses of measuring equipment -- I'm just saying you don't have
to follow them slavishly to know whether or not a video image is obviously over or
underexposed, over or undersaturated, too noisy etc.
Are you now telling me you can't be sure if a video image is too noisy if you haven't
consulted your measuring equipment? If so, this doesn't say much for yor 33 years'
experience on the job.
Let me put it this way: when I sold my old Hi8 camera five years ago and bought the
VX1000 I certainly did not require measuring equipment to realize that the VXK had far
superior S/N numbers.
Charles
Matt Sandstrom
> I'm retired.
thank you!
/matt
R.J.F. Stewart
> > What do you think causes the noise to vary with light level, (assuming you make no
> > adjustment to the electronics)?
> >
> > Do you think the noise increases or decreases with an increase in light level?
>
> I have already told you (and others) that I'm not an electronics engineer.
>
> Why don't you answer your own question?
I can't think why the noise level should vary with light level, and I've no way of
knowing why you think it would, since you haven't said.
Rod.
R.J.F. Stewart
> > I have yet to hear how the use of an optical filter can change noise level of the
> > electronic signal, and now that someone else has thrown contrast ratio into the
> > discussion, I would be quite interested to know how you think it can change that
> > too. The extra light scatter in a filter placed in front of the lens can raise the
> > black level slightly (and sometimes affect black balance too), which if not
> > compensated for electronically might *appear* to be a change in contrast range,
> > but I don't know exactly what you are seeing.
>
> Like I just said, if you want to see you'll have to visit. In the meantime you should
> not discredit something which you have *not* seen.
I think the reason I haven't seen it is that it doesn't happen.
> I've already told you I have no "suggested" explanation. If I were an electronics
> engineer I might be able to make the "suggestion" you seek. However, even if I did
> have substantial expertise and knowledge in this field, I would not go around
> bickering with people about work I have not seen myself.
If you were an electronic engineer you would know that noise is a random voltage
variation that is superimposed on the signal, and will not vary if you don't make any
change to the amplifier gain. Not only that, if the signal level is maintained - as it
normally is between reasonable limits - by controlling the amount of light falling on
the CCD, there is no possible way the CCD and subsequent circuitry could "know" whether
this had been done by ND filters or iris adjustment, whatever you may think you've seen.
I'm baffled that you think it could.
Rod.
C2001
> there is no possible way the CCD and subsequent circuitry could "know" whether this
> had been done by ND filters or iris adjustment, whatever you may think you've seen.
Rod,
I *never* stated circuitry could "know" anything (let alone discernment of whether or not
light levels reaching a CCD were the result of iris adjustments or use of ND filters
[built-in or optical]). Apparently you are (mis)interpreting other statements of mine as
meaning something in engineering contexts which -- frankly -- I wouldn't even presume to do
(nor be interested in doing).
> I'm baffled that you think it could.
You people appear to be self-brainwashed. Now you are putting words into my mouth which I
never stated (or even thought).
All I said was that, after reading about the cinematography for "The English Patient" in AC,
and noting John Seale's use of varying levels of ND filters to accommodate varying amounts
of excess light/glare at different times of the day, I was curious to see if using an ND
filter on my VX1000, in addition to the camera's built-in ND filter, might reduce glare in
dazzle-bright and/or high-contrast scenarios.
I discovered that doing so indeed _did_ reduce glare. I ALSO discovered that doing so
increased shadow detail -- which, in tandem with reduced glare, certainly _suggests_ that
the additional use of ND filtration was somehow increasing the contrast range of the camera
(optics or electronics).
Anyone who thinks I'm making this up (or imagining it) is welcome to visit and see the
related tapes themselves. If there is 'some other' explanation for what I'm inclined to
perceive as an increase in contrast range, I'm quite willing to consider it. But I don't
appreciate presumptuous, rude, boorish, misquote-prone, and puffed-up know-it-alls (even
those who may indeed know quite a bit) immediately assuming that something *they* haven't
experienced must be "inaccurate" or "wrong".
Moreover, being a "pro broadcast engineer" (for 33 years!) is _very_ dubious grounds to make
such pronouncements, when one considers the abysmal technical quality of most broadcast TV
in North America (and I don't think many people _anywhere_ will argue with me on this...).
Frankly, if I had spent 33 years working as a pro broadcast engineer in North America, I'd
go out of my way to keep quiet about it.
Charles
R.J.F. Stewart
> Moreover, being a "pro broadcast engineer" (for 33 years!) is _very_ dubious grounds to make
> such pronouncements, when one considers the abysmal technical quality of most broadcast TV
> in North America (and I don't think many people _anywhere_ will argue with me on this...).
>
> Frankly, if I had spent 33 years working as a pro broadcast engineer in North America, I'd
> go out of my way to keep quiet about it.
I haven't. I have been to North America, but I don't live there. I live and
work in England. I wouldn't presume to pontificate about the quality of
your pictures, not having much experience of what they are like,
but even though I say so myself, I think ours are rather good.
Rod.
C2001
Thanks for clarification.
C.
Frank Wylie
wrote:
>Let me put it this way: when I sold my old Hi8 camera five years ago and bought the
>VX1000 I certainly did not require measuring equipment to realize that the VXK had far
>superior S/N numbers.
I must say that in all the years I have been reading Usenet posts
(near 20), this thread takes the cake.
I have never laughed so hard in all my life!
Mr. Mullen, don't waste your time; I know you have better things to
do and advice to give to others who may actually put it to good use.
C2001
> C2001 wrote:
> >Let me put it this way: when I sold my old Hi8 camera five years ago and bought the
> >VX1000 I certainly did not require measuring equipment to realize that the VXK had far
> >superior S/N numbers.
>
> I must say that in all the years I have been reading Usenet posts
> (near 20), this thread takes the cake.
I should mention I've been on a strict gluten-free diet for more than six years.
Fortunately, there is a fabulous German bakery in my city which has two cakes that don't
use wheat flour.
> I have never laughed so hard in all my life!
You may or may not be aware there is some sort of award for funniest post of the year at
Usenet, though it might be a challenge to pick the best nominee from the thread.
> Mr. Mullen, don't waste your time; I know you have better things to
> do and advice to give to others who may actually put it to good use.
He should start with a shrink. (He meets all the DSM-IV criteria to support a diagnosis of
advanced Narcissistic Personality Disorder.)
Charles
gh
Wylie) wrote:
I agree. I have also been amusing myself reading this thread. It is
farcical and quickly ran into an embarrasing cul-de-sac of arrogant
pseudo-knowledge. David, you surely have better things to discuss.
There are those of us, like you, who know the tools of the trade and
how they really work. It's pretty basic how a neutral density filter
works: it is simply one control over the amount of light entering a
lens, not a control over the contrast range of the light entering the
lens, no matter whether that lens is attached to a motion picture, still
or video camera of any type. So let a few homevideo hambones use them as
if they are low contrast filters or something and whack off to each
other about this great new thing they are "experimenting" with while the
rest of us chat about something interesting.
GH
--
R.J.F. Stewart
> It's pretty basic how a neutral density filter
> works: it is simply one control over the amount of light entering a
> lens, not a control over the contrast range of the light entering the
> lens, no matter whether that lens is attached to a motion picture, still
> or video camera of any type. So let a few homevideo hambones use them as
> if they are low contrast filters or something
Maybe this is where the misunderstanding has come from. Maybe the cheaper
homevideo filters *do* act like "low-contrast" filters due to extra light
scattering? As I've pointed out, even the professional ones add some extra
flare if they are not adequately shaded, and this can affect black level
and black balance. The effect is sometimes enough to see on a picture
monitor, and nearly always visible on one of those waveform monitors or
vectorscopes that Mr. C2001 seems to think are unnecessary. (This is one of
the reasons for using instruments in television, of course - to see
potentially troublesome effects before they become troublesome).
Rod.
CT
> CT wrote:
>
>
> Nothing -- I was merely joshing about some video people who are afraid to
do
> work of any kind without consulting one (it amuses me that vectorscope
rhymes
> with horoscope...).
>
>
You'll also need a waveform monitor.
CT
Then why don't you stop attempting to answer questions that are beyond your
level of knowledge and experience?
CT
"R.J.F. Stewart" <escap...@08002go.com> wrote in message
news:VA.0000004...@08002go.com...
I can only think that although for a particular level of gain the noise
level should remain constant, the noise will be more visible on a monitor
under certain conditions.
It's like listening to an analogue audio recording. Do you notice the noise
in the loud parts or the quiet parts? For video I'd suggest that an ND
filter might possibly reduce the_perceived_level of noise by pushing it down
toward the blacks. The noise will still be there, but less obvious to a
human eye watching a monitor. The same could also be said for noise in
areas of the image approaching and exceeding maximum white level.
CT
I'd agree Rod, except for the ever increasing amount of crap originated on
VX1000s and similar ;)
C2001
I suggest you re-read the thread: I never attempted to answer questions about
the engineering aspects of video concerning which I have little knowledge and
claim no expertise.
Because certain other ego-inflated pundits at this NG were skeptical of my
reported results using ND filters (all I did was report my own experience in
using them), *THEY* demanded technical explanations which I responsibly declined
to provide since I already knew I lacked the knowledge to attempt such.
Had I attempted some sort of bumbling scientific explanation, *then* an arrogant
pig like Mr. Mullen _would_ have justifiably had something about which to
complain.
C.
C2001
My Avid Xpress has both -- but I don't need them to "confirm" the results of
filtration in post.
C.
C2001
Really?
I happen to know the BBC loves them.
C.
CT
> >
> > You'll also need a waveform monitor.
>
> My Avid Xpress has both -- but I don't need them to "confirm" the results
of
> filtration in post.
>
To be honest, you might as well rely on your eyes as to trust that to give
you a reasonable impression of your final output ;)
The reason we use external, calibrated test equipment is to obtain a
fully-objective assessment of a signal. My eyes are 20/20, my assessment
abilities well tuned, but I would avoid working on an edit without proper
test equipment because I know that the eye can be fooled.
CT
> CT wrote:
>
> > >
> > I'd agree Rod, except for the ever increasing amount of crap originated
on
> > VX1000s and similar ;)
>
> Really?
>
> I happen to know the BBC loves them.
>
>
Really, for what? Low-budget current affairs when the talent is operating
the camera?
Using DV _supposedly_runs contrary to the BBC's technical policy,
unless justified for artistic or practical reasons.
Still, what are you going to do as low budget crap pushes 5k per half hour?
C2001
> "C2001" <charle...@sympatico.ca> wrote in message
>
> > CT wrote:
> > > >
> > > I'd agree Rod, except for the ever increasing amount of crap originated
> > > on VX1000s and similar ;)
> >
> > Really?
> >
> > I happen to know the BBC loves them.
>
> Really, for what? Low-budget current affairs when the talent is operating
> the camera?
Perhaps it was incorrect for me to say "loves", but I'm aware of videographers
from several NGs (mainly rec.video.production [where, incidentally, no one has
ever questioned my reported use of ND filters]) who work for the BBC and have
used the VX1000, often with a widescreen anamorphic adaptor, rather than the
crop/stretch widescreen mode available in PAL models of the camera, for
documentaries.
I'm not in a position to confirm exactly "where" in the BBC's 'pecking order'
such programming falls, but I do know it exists.
> Using DV _supposedly_runs contrary to the BBC's technical policy,
> unless justified for artistic or practical reasons.
"Supposedly"?
Have you verified this with a vectorscope and waveform monitor???
(Sorry -- couldn't resist. [<g>)
> Still, what are you going to do as low budget crap pushes 5k per half hour?
I don't work for the BBC.
C.
C2001
> "C2001" <charle...@sympatico.ca> wrote in message
>
> > > You'll also need a waveform monitor.
> >
> > My Avid Xpress has both -- but I don't need them to "confirm" the results
> > of filtration in post.
>
> To be honest, you might as well rely on your eyes as to trust that to give
> you a reasonable impression of your final output ;)
>
> The reason we use external, calibrated test equipment is to obtain a
> fully-objective assessment of a signal. My eyes are 20/20, my assessment
> abilities well tuned, but I would avoid working on an edit without proper
> test equipment because I know that the eye can be fooled.
You are correct (obviously) -- though I continue to be appalled at the lack of
calibration for many (usually very expensive) industrial monitors used in pro
and broadcast edit suites.
Out-of-the-box, most Sony monitors (just like their consumer bretheren [at
least in Canada and the US]) are *way* overdriven with white levels almost
blinding. If you are working with a high-quality monitor (I have three PVMs)
it's still mandatory to have them fully calibrated. Not only will they be more
watchable (especially for long edit sessions) but, when properly calibrated be
a trained technician (in America, at least, ISF-certified is usually best), are
far _less_ likely to fool your eyes.
No one can judge the technical quality of video (*or*, film transfered to
video) if they are watching tapes on an improperly adjusted monitor. Still, I
would never rely exclusively on test equipment to assess black level, white
level, or even chroma saturation.
C.
David Mullen
>I can only think that although for a particular level of gain the noise
>level should remain constant, the noise will be more visible on a monitor
>under certain conditions.
Perhaps -- film grain works that way. It's more visible in midtones than
blacks or white, so low contrast scenes will show off the grain more than
high contrast scenes. But when I've tested high levels of video gain in the
past, it seemed evenly visble across the frame.
I had this argument with someone who was convinced you could not shoot a
low-key night scene in video without getting noise -- this was because they
were using a DV camcorder on auto-exposure and of course, as the image
became darker, it started boosting the signal above 0 db once maximum lens
aperture was reached.
There can be a lot of misunderstanding from people using consumer equipment
because so much is internal or automatic -- the f/stop, the gain settings,
etc.
Under or overexposure per se does not increase noise during image capture /
recording if you don't alter the manual gain setting -- when noise comes in
later if you need to make color-corrections to the image and thus try to
bring down bright areas or bring up shadow information. Hence why some
people think dark scenes are noisier -- it's often because they don't let
them stay dark, but underexpose them too much and then lift them back up
slightly, either by boosting the camera signal or later brightening them in
color-correcting.
David Mullen
Sswitaj
writes:
>Nothing -- I was merely joshing about some video people who are afraid to do
>work of any kind without consulting one
I for one, hate to do video work without a waveform monitor handy, whether on
the set, or in the telecine bay.
Though I'm not a slave to the 'scope, I damn well spot check once in a while,
just to check what I'm actually printing to tape.
IMHO, this is no different, or less important than regualarly checking the
printing lights to see where you're tending with your negative exposure.
But then again, I have a background as, what was the term, oh yeah, "one of
those hopeless engineers"
- Steve
Sswitaj
Mullen" <dav...@earthlink.net> writes:
>The CCD doesn't know that an ND filter is being used -- for example, whether
>or not the camera has an ND.9 filter and shoots at f/2.8 or no ND filter and
>shoots at f/8, the amount of light reaching the CCD is the same. So why
>would the same amount of light reaching the CCD now suddenly be easier for
>the CCD to handle???
Hmmm...
Interesting (and heated) discussion.
I feel like playing Devils' Advocate (despite Charles' pretty arrogant
attitude, which could use some adjustment).
Assuming there's really something going on with his camera that we don't
normally see in film, some of the possible culprits might be....
- Is the iris ring on the VX100 accessable? How far does it stop down? Maybe
he's overexposed without a filter because he's past his minimum aperature
without ND.
- Maybe the ND filter is of low quality and is acting a bit like a
diffusion/defraction filter (like a light black promist).
- Alternately, maybe it's a plastic filter (like a Cokin). These are notorious
for acting like polarizers, and it could be affecting the specular highlights.
- Maybe the ND filter has a cutoff band which reduces near IR transmission from
specular higlights, therefore preventing point saturation in the detector.
- Maybe the lens has really non-linear sharpness characteristics with respect
to aperature. Maybe it's cheap optics and opening up 2 stops softens the image
appreciably.
- Maybe opening up those two stops produces a less-collimated output beam from
the lens exit pupil. This could have some effect on the way the light is read
at the somewhat 3-dimensional ccd surface.
Of course, in all fairness...
- maybe it's all psycological.
My $.02.
- Steve
C2001
I'm a firm believer in covering all the bases. But my eyes are the first and final
judge of image quality.
If I tend to be a tad skeptical about heavy reliance on vectorscopes and waveform
monitors it's because I've been in too many edit suites which had *terrible*,
overdriven and (obviously) never properly calibrated video monitors which made a
mockery of all the other elaborate and pricey gear at hand.
Charles
C2001
> >The CCD doesn't know that an ND filter is being used -- for example, whether
> >or not the camera has an ND.9 filter and shoots at f/2.8 or no ND filter and
> >shoots at f/8, the amount of light reaching the CCD is the same. So why
> >would the same amount of light reaching the CCD now suddenly be easier for
> >the CCD to handle???
>
> Hmmm...
>
> Interesting (and heated) discussion.
>
> I feel like playing Devils' Advocate (despite Charles' pretty arrogant
> attitude, which could use some adjustment).
Um...., just remember it takes _two_ to tango: I was *only* arrogant in dealing
with a certain boorish person who repeatedly [and, frankly, inexplicably] misquoted
me and misrepresented my comments with cavalier and generally ill-considered
personal and intellectual arrogance. (I shudder to think what would happen to this
person if he ever has to be a witness in court. Anyway, I've killfiled him so it's
no longer a concern for me.)
> Assuming there's really something going on with his camera that we don't
> normally see in film, some of the possible culprits might be....
>
> - Is the iris ring on the VX100 accessable? How far does it stop down? Maybe
> he's overexposed without a filter because he's past his minimum aperature
> without ND.
The iris is controlled by a small wheel on the side of the camera (I never use the
'auto' mode [which tends to overexpose everything, at least outdoors, even with the
camera's built-in ND filter activated]). The range is f2, through f11, with
options for gain adjustments, as well as the camera's built-in ND filter. As I
mentioned in one of my posts, in bright outdoor light, using the additional Tiffen
ND3 on top of the built-in ND, I found almost invariably the two exposure settings
that worked were f8 orf9.6 (f11 was visibly underexposed with crushed blacks,
unless there might be some artistic reason to use this look).
The camera's exposure range is _not_ at the level of Sony's more expensive DVCAM
and DigiBeta units. The 14X Canon lens on my DSR300 covers f1.4 to f16, and of
course has much more elaborate parameters for virtually all standard adjustments
(and this is by no means the best lens available for the 300).
And I'd concur that the camera does overexpose (not horribly, but visibly) in the
brighest outdoor light, using only the built-in ND filter and with iris at f11.
Indeed, it was my nagging dissatisfaction with this excess exposure and my reading
in "AC" about John Seale's use of ND filters for the desert shots in "The English
Patient" which prompted me to try the additional ND filter with the VX1000.
I was *not* seeking enhanced shadow detail in tandem with tamed glare (which I
[perhaps simplistically?] perceived as an increase in contrast range), but the
reality is this is what the Tiffen _also_ gave me.
> - Maybe the ND filter is of low quality and is acting a bit like a
> diffusion/defraction filter (like a light black promist).
I don't have sufficient experience with a large range of filters (types, or
manufacturers) to say much, though this is the first time anyone within my
acquaintance has ever suggested Tiffen filters might be "low quality"!
(If Tiffens are good enough for Vilmos Zsigmond, I think they're good enough for me
at least.)
<g>
> - Alternately, maybe it's a plastic filter (like a Cokin). These are notorious
> for acting like polarizers, and it could be affecting the specular highlights.
As I said, it's Tiffen not Cokin and I'm pretty sure it's glass, not plastic.
> - Maybe the ND filter has a cutoff band which reduces near IR transmission from
> specular higlights, therefore preventing point saturation in the detector.
I'm out of my depth on this point -- and happy to acknowledge such.
> - Maybe the lens has really non-linear sharpness characteristics with respect
> to aperature. Maybe it's cheap optics and opening up 2 stops softens the image
> appreciably.
There has been a lot of debate (*and* misinformation) about the lens on the VX1000
since the camera was introduced in 1995. Because it was a prosumer unit, many
commentators (and video-bashing film-only snobs [in case anyone wonders, I happen
to love film _and_ video formats]) erroneously assumed the lens was "cheap plastic"
-- which it isn't. My dealers have stated it is lead glass, and probably
manufactured for Sony by Canon. (Sony doesn't advertize this, since Canon happens
to be a rival camera manufacturer; OTOH, Sony loves to crow about the genuinely
dubious 'Carl Zeiss' lenses [they aren't made by Zeiss] used in a few of its
upper-mid priced consumer handycams.)
Of course, even from Canon you get what you pay for and I wouldn't expect the lens
built for the VX1000 to be anywhere near as good as the best ones manufactured for
industrial cameras, since industrial lenses themselves often cost more than the
entire price of the VX1000. (Though I _can_ mention that my primary Avid trainer,
who is also a pro videographer who works mainly with BetaSP, was bowled over by the
technical quality of the best work I've done with the VX1000 [he had never used the
1000 or, I'm fairly sure, ever previously seen tapes shot with one]; at risk of
getting way off-topic, while the VX1000 doesn't have the range of the best
industrial units, its base quality [especially color accuracy which is nothing
short of staggering IMO] is good enough that most videographers who have used it
extensively agree it can be used for B-roll in tandem with BetaSP or DVCAM.)
It's certainly to be expected that even the best prosumer cameras such as the
VX1000 and its main Canon rival, the XL1, will have limitations and, perhaps,
abberations which may generate unpredictable and (seemingly) unlikely results,
compared with industrial cameras (never mind the best film ones).
> - Maybe opening up those two stops produces a less-collimated output beam from
> the lens exit pupil. This could have some effect on the way the light is read
> at the somewhat 3-dimensional ccd surface.
Thanks for sending me to my Websters to learn what "collimate" means!
> Of course, in all fairness...
>
> - maybe it's all psychological.
No -- as I said, anyone who wishes to visit may do so and -- be assured -- they
will see the difference between the two original tapes I discussed, as well as a
lot of subsequent work I've done.
In the meantime, I'd say anyone who owns or has access to a VX1000 (IIRC, the
person I've plonked uses [or has used] one), should buy a Tiffen ND3 (they aren't
particularly expensive) and use it in the same conditions (very bright outdoor sun)
and settings (f8 or f9.6, with the camera's built-in ND filter also activated) I
have described -- and see for yourselves.
> My $.02.
Thanks for your constructive, open-minded and very polite comments. It's people
like you who make Usenet worthwhile.
Charles
R.J.F. Stewart
> I'm a firm believer in covering all the bases. But my eyes are the first and final
> judge of image quality.
Your eyes can only judge what is on the particular screen available to you, under the
lighting conditions in which you have to use it. If you only ever have to work in a
properly lit control room with properly adjusted broadcast quality monitors, then you
are very lucky.
> If I tend to be a tad skeptical about heavy reliance on vectorscopes and waveform
> monitors it's because I've been in too many edit suites which had *terrible*,
> overdriven and (obviously) never properly calibrated video monitors which made a
> mockery of all the other elaborate and pricey gear at hand.
If the picture monitors cannot always be relied upon, then there is all the more
reason, I would have thought, to know how to use the instruments.
Rod.
C2001
> C2001 wrote:
> > I'm a firm believer in covering all the bases. But my eyes are the first and final
> > judge of image quality.
>
> Your eyes can only judge what is on the particular screen available to you, under the
> lighting conditions in which you have to use it. If you only ever have to work in a
> properly lit control room with properly adjusted broadcast quality monitors, then you
> are very lucky.
Which is one of the major reasons I have my own edit suite (in my home) with proper
lighting (usually very low), three fully calibrated PVMs, and the walls painted with the
lighest of four ISF-approved gray paints from Benjamin Moore. And yes -- I know I'm very
lucky to have such a setup under my own roof.
I also have the G3, drive tower and PCI extender hard-wired to an adjacent room, because
their noise makes sound editing virtually impossible, unless you turn the volume to
migraine-inducing levels, or else use headphones, which many editors like, but I don't
because [1] the sound they deliver is markedly different from almost any kind of speaker
setup and [2] I might not hear other sounds (phone etc.) I wouldn't wish to miss -- even
when the interruption is an inconvenience.
If I were asked or expected to work with facilities I considered seriously sub-standard,
I'd do so _only_ on condition that I would not be given a credit.
> > If I tend to be a tad skeptical about heavy reliance on vectorscopes and waveform
> > monitors it's because I've been in too many edit suites which had *terrible*,
> > overdriven and (obviously) never properly calibrated video monitors which made a
> > mockery of all the other elaborate and pricey gear at hand.
>
> If the picture monitors cannot always be relied upon, then there is all the more
> reason, I would have thought, to know how to use the instruments.
I agree (and, yes, this did occur to me when I wrote the above) -- but it's still a very
sad commentary on overall broadcast technical standards (at least in America) that most
of the telecasting is technologically, at best, wildly erratic and, at worst,
unwatchable. And this is why I was not exactly 'intimidated' by someone from this arena.
While I'm not convinced PAL is so totally superior to NTSC as even many North Americans
believe (I have a real problem with the flicker, and the tint problems long associated
with NTSC don't really exist with the digital formats), my impression is that TV
broadcasting in Europe is much more tightly controlled than what Americans and Canadians
have been seeing for the past 30 years.
Thanks for a reasonable, intelligent and civilized post.
Charles
CT
> "R.J.F. Stewart" wrote:
>
> If I were asked or expected to work with facilities I considered seriously
sub-standard,
> I'd do so _only_ on condition that I would not be given a credit.
>
RFLMAO
This thread just gets funnier and funnier.
CT
> Sswitaj wrote:
> Um...., just remember it takes _two_ to tango: I was *only* arrogant in
dealing
> with a certain boorish person who repeatedly [and, frankly, inexplicably]
misquoted
> me and misrepresented my comments with cavalier and generally
ill-considered
> personal and intellectual arrogance. (I shudder to think what would
happen to this
> person if he ever has to be a witness in court. Anyway, I've killfiled
him so it's
> no longer a concern for me.)
>
Your loss.
C2001
> I wrote:
>
> > "R.J.F. Stewart" wrote:
> > > C2001 wrote:
> > > > I use Tiffen's ND3 with the Sony VX1000, in addition to the camera's
> > > > built-in ND filter (I think the XL1 has a built-in ND also). The results
> > > > are *spectacular* and I have never experienced difficulty with accurate
> > > > color balance using the ND3.
> > > >
> > > > That extra dose of ND _really_ stretches the camera's contrast range,
> > > > reduces noise and (IMO) enhances color acuracy and related subtlety.
> > >
> > > Can you explain how the use of an ND filter can reduce noise in a
> > > television camera, and have you patented the method yet? As a television
> > > engineer for 33 years, I'd be very interested to know this.
> >
> > I'm not an electronics engineer, but I'd speculate that the reduction of noise
> > (and it's not my imagination; I did back-to-back testing) is simply because
> > smoothing the contrast ratio reduces stress on the image sensors. Most
> > videographers are well aware that too little *and* too much light result in
> > video noise.
> >
> > On dazzle-bright days, I found that using the Tiffen HD3 wiht the VX1000
> > created a markedly more film-like look, simply because the camera could handle
> > contrast so much more effectively -- the really bright spots never burned yet
> > the shadow detail was also tremendous.
>
> Neglected to mention that I started experimenting with additional ND filtration
> after reading about John Seale's use of ND filters for "The English Patient" in
> American Cinematographer.
I'm reviewing this thread (guess why? -- <g> ) and wish to remind people that in
this initial reference to Mr. Seale's work for "The English Patient" I did *not*
state he had claimed use of ND filters could increase contrast range -- as at least
one person thoughtlessly and arrogantly claimed I had done.
Just thought I'd clear that up.
Charles
C2001
> David Mullen wrote:
>
> > >I'm not an electronics engineer, but I'd speculate that the reduction of
> > noise
> > >(and it's not my imagination; I did back-to-back testing) is simply because
> > >smoothing the contrast ratio reduces stress on the image sensors. Most
> > >videographers are well aware that too little *and* too much light result in
> > >video noise.
> > >
> > >On dazzle-bright days, I found that using the Tiffen HD3 wiht the VX1000
> > >created a markedly more film-like look, simply because the camera could
> > handle
> > >contrast so much more effectively -- the really bright spots never burned
> > yet
> > >the shadow detail was also tremendous.
> >
> > I'm sorry, but this sounds a little like psuedo-science here. Think about
> > what happens in the camera when using an ND filter.
> >
> > The amount of light coming OUT of the rear of the lens, past the lens iris,
> > and finally hitting the CCD's -- whether you used an ND.90 and shot at an
> > f/2.8 or no ND filter and shot at f/8 -- IS THE SAME. The intensity is the
> > same and the brightness of highlights & shadows is the same. The contrast
> > is the same, except for the fact that out-of-focus areas will be lower in
> > contrast due to the blending of bright & dark -- and the wider aperture
> > version allows more of the frame to be out-of-focus. There also might be
> > more lens flare when shooting wide open that will reduce some of the black
> > levels.
> >
> > Lower depth-of-field in video also mimics the depth-of-field of 35mm more,
> > which is generally lower due to the longer focal lengths necessary -- hence
> > the greater "film look". But this is not because contrast is lowered.
> >
> > The dynamic range of the video image is NOT increased by use of ND filters,
> > unless without ND filters, you cannot expose the image properly. However,
> > lower depth-of-field might create a more out-of-focus background, which will
> > LOOK lower in contrast. But the in-focus subject has the same contrast and
> > the dynamic range permitted by the CD's & recording format is not changed by
> > the use of ND filters, other than some possible loss of black levels from
> > lens flare from shooting wide open on some lenses.
> >
> > ND filters are not magic -- they just cut the overall amount of light. They
> > don't compress or extend contrast. The only issue is how much the image is
> > changed by being able to shoot at wider lens apertures, and how that affects
> > PERCEIVED contrast. But on the in-focus subject, you won't find that bright
> > highlights are suddenly less clipped or that shadow detail is suddenly
> > extended through the use of ND filters, assuming that you are still exposing
> > for the same IRE level on the subject (in other words, you aren't actually
> > underexposing highlights in one version versus another).
> >
> > Pola filters used as ND filters are another issue, since they can affect
> > glare off of highlights...
>
> David, at one point in your post you say the light coming out of the rear of the
> lens is the same _with or without an ND filter_, *then* a few paragraphs later
> you state that ND filters "cut the overall amount of light".
>
> They may not directly compress or extend the contrast of the light going through
> the lens, but the way they work definitely makes it easier for the CCD to handle
> high-contrast and/or dazzle-bright outdoor light -- and I have the A/B
> comparison tapes to prove this.
>
> Polarizers (depending on how they are implemented) may reduce reflection, but
> this is not necessarily the same as glare -- ie, a reflection is not
> _automatically_ glare; OTOH glare is usually the result of a surface that is
> reflecting a high level of ambient (usually sun) light. Also, polarizers may
> screw up color accuracy and/or consistency.
>
> When polarizers are used to reduce or eliminate reflections the reason usually
> is to eliminate an unwanted reflection rather than to eliminate glare.
>
> As I mentioned, I decided to experiement with additional ND filtration after
> reading about the use of ND filters for the desert sequences in "The English
> Patient".
I'd say my initial response to DM is decidedly diplomatic and conciliatory --
certainly not "arrogant" as one or two people have tried to suggest.
Please note that I readily acknowledged, after reading his post, that ND filters (as
he stated) may not directly affect contrast range, but in my particular experience
(which I described in further detail in subsequent posts, in an attempt to clarify
my own conclusions and the reactions of others) they were certainly doing
_something_ which suggested this. I was, and still an, open to intelligent,
non-confrontational dialog on the subject.
I'd agree that a reduction in depth-of-field (one of the primary effects of using ND
filtration) may create a softer focus in portions of the image which, in turn, could
misleadingly suggest a reduction in visible video noise. In retrospect, I'd
acknowledge it may have been simplistic for me to conclude that the ND filter was
actually reducing noise. Nevertheless, however one wishes to explain or qualify the
results, the tapes I shot _without_ the Tiffen ND3 were often 'grittier' as well as
more glary in very bright areas looking than those I did with the Tiffen ND3.
As I stated in several posts, I'm not an electronics engineer and never claimed to
know or understand precisely how every aspect of video acquisition functions.
Charles
David Marks
> Had I attempted some sort of bumbling scientific explanation, *then* an arrogant
> pig like Mr. Mullen _would_ have justifiably had something about which to
> complain.
>
You, C2001, are a moron. David Mullen has more knowledge in one of his fingernails
(and is more than willing to share said information) than you will ever hope to
acquire. Get a clue.
Dave Marks
C2001
I never questioned DM's base knowledge -- only his boorish bad manners and, worse,
inexcusable predilection for loading the dice of an argument further toward his favor
by flagrantly and dishonestly misquoting and misrepresenting someone else's
observations and comments. If he is as secure in his 'knowledge' as he would like to
believe, there would be no need for him to do this.
And I must frankly question how much _anyone_ can learn from even the most informed
person, when that person resorts to DM's type of tawdry manipulation and general
chicanery.
Charles
Sswitaj
writes:
>Because certain other ego-inflated pundits at this NG were skeptical of my
>reported results using ND filters (all I did was report my own experience in
>using them),
No, you reported your own experience, then you told people with a decade of
practical experience that they were wrong when they realted *their* results
with ND filters.
After all, they only told you that ND filters don't do that particular thing,
and you have some other, unexplored effect at work. Hardly fightin' words.
> an arrogant pig like Mr. Mullen _would_ have justifiably had something about
> which to complain.
I'll discuss filters with you all day long, but I won't tolerate someone
badmouthing David Mullen.
Far from being "an arrogant pig" David has been a long standing voice of
tolerant reason and, importantly, good advice in this group.
Then again, unlike you, I have never used a VX1000. Like David, the only thing
I've got to work with is a pitiful 20 years of experience with still, video,
motion picture, and airborne infrared photography, so what would we know
anyhow?
- Steve
derek
> David Mullen wrote:
>> It IS simple. An ND.9 filter cuts out three stops of light. If you adjust
>> by opening up the iris by three stops, the same amount of light passes to
>> the CCD or film, unless further altered by shutter speed. What else do
>> you think an ND filter does beside cut the overall level of light?
>
> It doesn't perhaps *directly* do anything more -- BUT, because it cuts down
> overall light level, it enables alternative flexibility in other aspects of
the image capture process.
I know you've tried to explain this, but ... it cuts down the 'overall
light level', which, in order to prevent the image being underexposed,
the iris is opened to return to the correct exposure level, and the same
amount of light reaches the CCD. As the ND filter acts uniformly for
different wavelengths, and the optical system is adjusted to provide the
same illumination of the imaging device as if the ND filter had not been
used, and since the ND filter doesn't change contrast, I'm struggling to
follow the technical argument. As someone who also shoots on video, I'd
like to believe you, but all I can think of now is that I'll see it when I
believe it. Last weekend I was at the premiere of a feature I cut last
year and I shot some interviews/background on DB and with the director and
his family. It was a beautifully crisp, clear sunny day, used an ND
filter (instead of the camera's built in filter wheel) and shot mostly at
about f3.5 - 4, with camera gain set to -3db. I just couldn't see the
improvements you refer to, but I'd happily try another test to see if I
can reproduce the results you describe. I certainly couldn't see more
shadow detail - this was tested by removing the ND filter and using a
smaller aperture instead. Dynamic contrast control circuit was turned off
throughout and edge enhancement was 5 per cent below factory default.
Like I say, I'm open minded about this, but haven't been able to reproduce
your results. I'm speculating here but if you are using automatic gain
control and switch to a low gain setting of -3 to -6, the combined effect
of lower noise with a wider aperture to compensate for ND filtration will
produce a more pleasing picture with moderately better DoF than typical
video, but otherwise I can't see what might be happening.
Question: does the VX1000 have a non-controllable AGC circuit that kicks
in automatically? If so, this might modify the foot and knee of the gamma
curve.
>> It sounds clearly to me that you used an ND.3 filter without compensating
>> for the one stop light loss, and therefore brought down your highlights to a
>> lower level, thus allowing you to hold more info in the brightest areas and
>> hold chroma better. Shadow detail would be darker as well, but since video
>> has better latitude for underexposure, compensation might have not been
>> necessary. This is a common trick -- a slight overall underexposure -- for
>> video photography. It doesn't necessarily mean that the image LOOKS
>> underexposed.
This is the only explanation that makes sense to me.
regards,
Derek
>> If you think that a mild ND.3 filter (only ONE STOP of light loss!) can
>> dramatically change the latitude and color reproduction of a video camera,
>> you really don't understand the basics of photography or exposure.
>
> Now you are being condescending and offensive.
I don't think David's being either of those; in the context of this
argument it's a valid observation. While you may be pleased with the
results of minimising gain and shooting with a wider aperture, everything
David and others have explained is right on the mark. No condescension or
offense intended.
> All the ND filter did was enable settings that enabled the camera to produce
> video that very clearly had more contrast range.
This brings us back to the point that much of how an image is perceived is
in the way the brain operates, not the way the camera/projection/display
operates. You say the image 'very clearly had more contrast range',
however I suspect that in this instance David's para. above re. minimal
video underexposure in a high a contrast environment is what's at play
here.
Anyway, you are getting results that please you and those you are working
with, so well done.
derek
> I've practically given you all my credentials & credits, and my full name is
> at the bottom of all my posts. Perhaps you'd care to do likewise? As for
> my video work, a digital feature I shot called "Jackpot" will be released by
> Sony Pictures Classics theatrically on July 27 so maybe you can check it
> out. There is also an article on it in the newest RES Magazine.
David, I presume this was transferred to 35mm? If you've seen what the
projected image is like, I'd like to hear your comments. I'm not sure if
it would be picked up for theatrical distribution outside US.
regards,
Derek
C2001
> C2001 writes:
>
> >Because certain other ego-inflated pundits at this NG were skeptical of my
> >reported results using ND filters (all I did was report my own experience in
> >using them),
>
> No, you reported your own experience, then you told people with a decade of
> practical experience that they were wrong when they realted *their* results
> with ND filters.
No, Steve, it was the opposite: I reported my experience and other people told me
I couldn't possibly be "right" since my experience and findings didn't conform 100%
to their understanding of video electronics.
> After all, they only told you that ND filters don't do that particular thing,
> and you have some other, unexplored effect at work. Hardly fightin' words.
Steve, I think you should reread your *own* post in which you speculated on
alternate explanations for the results I described -- at the time you seemed to
take my findings at least somewhat seriously.
Why the switcheroo?
Unlike certain other posters to this NG, I'm not saying my experience is the final
authority on anything. Even if I were infintely more experienced and knowledgable
than I am at this time, I would *never* make or imply such a claim.
> > an arrogant pig like Mr. Mullen _would_ have justifiably had something about
> > which to complain.
>
> I'll discuss filters with you all day long, but I won't tolerate someone
> badmouthing David Mullen.
>
> Far from being "an arrogant pig" David has been a long standing voice of
> tolerant reason and, importantly, good advice in this group.
There is nothing "reasonable" or "tolerant" about a man who knowingly and willfully
misquotes and misrepresents the comments of others -- which is precisely what he
did repeatedly regarding my posts. Please reread the thread.
I don't question his base knowledge -- but his ethics and argument skills are, at
best, sleazy.
> Then again, unlike you, I have never used a VX1000. Like David, the only thing
> I've got to work with is a pitiful 20 years of experience with still, video,
> motion picture, and airborne infrared photography, so what would we know
> anyhow?
Steve, I never questioned your base expertise or that of anyone else at this NG. I
tried to respond to you offline (to avoid further debate on a thread which I'm
fairly sure everyone is tired of [I know I am...]), but my email was returned (I
even zapped the "NSPM" in your e-address).
IIRC, in one of the last posts from DM that I read he asked people to respond
offline if they had further comments on this thread.
I'm now making the same request -- at least we seem to agree on that.
Charles
derek
> And some veteran engineers (eg, the ones who've been working for 30 or more
> years [<g>]) are still convinced U-matic and BetaSP looks better than digital.
U-matic low-band is only marginally better than Video Hi-8, although some
of the Hi-band tube cameras produced better skin tone than many other
forms of video. I don't know anyone who considers SP better than DB
(sometimes/often better than DV, however) and find it extraordinary that
there are engineers consider U-matic superior to digital video.
derek
> Had I attempted some sort of bumbling scientific explanation, *then* an
arrogant
> pig like Mr. Mullen _would_ have justifiably had something about which to
> complain.
Charles, I'm not going to engage in a flame war, but, for the record,
David is very experienced, has been posting to this and a few other NGs
for several years and has always been professional, technically reliable,
and generous with his advice and help. It might help your case to consider
him in that light.
regards,
Derek
David Mullen
>projected image is like, I'd like to hear your comments.
We cropped the 1920 x 1080 picture to 2.35 : 1 and output this to a 35mm
anamorphic negative, so we effectively were using only 1920 x 810. We did
this mainly because we wanted the wider aspect ratio, plus comparison tests
showed the loss of resolution to be minimal, if undetectable. I think the
resolution difference between the 1.85 and 2.35 test got smoothed over in
the up-rezzing to a 2K data file before recording out to 35mm (I believe
this is the way E-Film handles the information.)
Overall, I think the sharpness seems halfway between Super-16 and 35mm - but
fairly grainless and "clean", which gives the illusion of improved
sharpness.
Clipping of bright areas is probably the most tell-tale artifact, but it
wasn't too bad in general.
Main technical problem, for me, was variations in focus due to a variety of
problems: some back-focus drift, some shots that I used heavier diffusion
and now regret it, some loss of resolution when the Canon zoom lenses were
shot wide-open, and one shot where the focus is simply off (otherwise, my
focus-puller did an excellent job.) It's a problem that I tried to take
more seriously in the second 24P HD feature I shot.
Color-wise, I did the color-correction very quickly -- in only six hours for
a two hour movie -- and I would have made further adjustments having now
seen the output to film. In an ideal world, I would have only corrected
Reel One, output it to film & screened it, and then gone back and
re-corrected it and then the following reels, having some idea in my head as
to how it looked on film.
The film has a lot of flashforwards & flashbacks, so it has some visual
jumps in style. When it was softer & hazier in some scenes, it had a sort
of 1970's road picture feeling, which was fine.
Suffice to say, I learned a lot.
David Mullen
C2001
> C2001 wrote:
> > DM wrote:
> >> It IS simple. An ND.9 filter cuts out three stops of light. If you adjust
> >> by opening up the iris by three stops, the same amount of light passes to
> >> the CCD or film, unless further altered by shutter speed. What else do
> >> you think an ND filter does beside cut the overall level of light?
> >
> > It doesn't perhaps *directly* do anything more -- BUT, because it cuts down
> > overall light level, it enables alternative flexibility in other aspects of
> the image capture process.
>
> I know you've tried to explain this, but ... it cuts down the 'overall
> light level', which, in order to prevent the image being underexposed,
> the iris is opened to return to the correct exposure level, and the same
> amount of light reaches the CCD. As the ND filter acts uniformly for
> different wavelengths, and the optical system is adjusted to provide the
> same illumination of the imaging device as if the ND filter had not been
> used, and since the ND filter doesn't change contrast, I'm struggling to
> follow the technical argument.
At this point so am I!
I'd like to just remind anyone who isn't 100% fed up with this thread, that it was
never my intention to provide a comprehensive explanation of my experience with the
Tiffen ND3 in either optical or electronic terms. Obviously, some of my statements
(some of which I acknowledge may have been simplistic) were perceived as
explanations and -- lo! -- WW3 was launched.
> As someone who also shoots on video, I'd like to believe you,
> but all I can think of now is that I'll see it when I believe it.
Since I don't have the option of showing video on the internet I appreciate your
point of view, without reservation.
> Last weekend I was at the premiere of a feature I cut last
> year and I shot some interviews/background on DB and with the director and
> his family. It was a beautifully crisp, clear sunny day, used an ND
> filter (instead of the camera's built in filter wheel) and shot mostly at
> about f3.5 - 4, with camera gain set to -3db. I just couldn't see the
> improvements you refer to, but I'd happily try another test to see if I
> can reproduce the results you describe.
I assume by "DB" you mean Digital Betacam. Obviously, you were using a camera that
is vastly superior to the VX1000, so I don't think this is an ideal frame of
reference. OTOH, since I've never used DigiBeta (not for lack of wanting...) I
don't wish to question your observations on your own experience with the format.
I can mention that quite a few people who use the VX1000 as one of their
major/regular acquisition cameras find that the camera's built-in ND filter _and_
the -3db gain setting to be fully satisfactory in reducing glare in very bright
outdoor light scenarios. I've tried this and consider it an excellent option,
though I prefer the results I get with the Tiffen ND3 in addition to the camera's
built-in ND filter -- but this is a strictly subjective preference (and I won't get
into the Viet Nam of attempting to explain what the difference is [I've learned a
few things in the last couple of days here......]).
> I certainly couldn't see more
> shadow detail - this was tested by removing the ND filter and using a
> smaller aperture instead. Dynamic contrast control circuit was turned off
> throughout and edge enhancement was 5 per cent below factory default.
> Like I say, I'm open minded about this, but haven't been able to reproduce
> your results.
As I've just said, DigiBeta is not the same as the VX1000. My understanding (based
only on readong reports from others, both in trade magazines and on the internet)
of DigiBeta is that one of its primary strengths is that its contrast range is
superior to most other formats (including even the best DVCAM and analog Beta) and
is even considred by some users to rival film. (But I'm only referencing comments
of others and have no opinion of my own regarding gear and/or configurations I
haven't used.)
> I'm speculating here but if you are using automatic gain
> control and switch to a low gain setting of -3 to -6, the combined effect
> of lower noise with a wider aperture to compensate for ND filtration will
> produce a more pleasing picture with moderately better DoF than typical
> video, but otherwise I can't see what might be happening.
>
> Question: does the VX1000 have a non-controllable AGC circuit that kicks
> in automatically? If so, this might modify the foot and knee of the gamma
> curve.
To answer your questions, I don't ever use the auto gain circuit on my VX1000 as it
tends to result in excess noise with anything other than the brighest
illumination. Sometimes I use the -3db option. Even in the lowest-light night
conditions I never use the highest gain levels since they result in unacceptable
noise.
> >> It sounds clearly to me that you used an ND.3 filter without compensating
> >> for the one stop light loss, and therefore brought down your highlights to a
> >> lower level, thus allowing you to hold more info in the brightest areas and
> >> hold chroma better. Shadow detail would be darker as well, but since video
> >> has better latitude for underexposure, compensation might have not been
> >> necessary. This is a common trick -- a slight overall underexposure -- for
> >> video photography. It doesn't necessarily mean that the image LOOKS
> >> underexposed.
>
> This is the only explanation that makes sense to me.
May well be correct -- from my perspective I'd have to experiment with numerous
different cameras to attempt to arrive at conclusions. But, as I've said from the
start of this thread, my primary interest as a videographer is getting the best
results possible. I also like to know as much as possible about the related
technology, but I'm not an electronics engineer and never aspired to be one.
The one point I can repeat is that the exposure range of the VX1000 _is_ limited
compared with the best industrial lenses, and this might explain (as a few people
have suggested) a phenomenon which I was able to ID using the 1000, but which might
not be noted on a more advanced unit.
> >> If you think that a mild ND.3 filter (only ONE STOP of light loss!) can
> >> dramatically change the latitude and color reproduction of a video camera,
> >> you really don't understand the basics of photography or exposure.
> >
> > Now you are being condescending and offensive.
> I don't think David's being either of those; in the context of this
> argument it's a valid observation. While you may be pleased with the
> results of minimising gain and shooting with a wider aperture, everything
> David and others have explained is right on the mark. No condescension or
> offense intended.
Perhaps not consciously intended, but certainly implied from at least one person.
> > All the ND filter did was enable settings that enabled the camera to produce
> > video that very clearly had more contrast range.
>
> This brings us back to the point that much of how an image is perceived is
> in the way the brain operates, not the way the camera/projection/display
> operates. You say the image 'very clearly had more contrast range',
> however I suspect that in this instance David's para. above re. minimal
> video underexposure in a high a contrast environment is what's at play
> here.
I'm glad to acknowledge it may have been careless for me to state "very clearly had
more contrast range" -- "definitely appeared to have more etc..." would have been a
better choice of words. The point is, however, the _difference_ simply jumped out
at me from the TV screen; numerous other video and photographic pros (including at
least two with several decades' experience) with whom I _have_ been able to share
back-to-back comparisons agreed with me (and these were not people who simply
wished to be polite).
> Anyway, you are getting results that please you and those you are working
> with, so well done.
Thanks for being polite.
Charles
C2001
I'm quite aware of his credentials and his history at Usenet.
Neither, however, preclude bad manners and/or debating skills which depend
heavily on flagrant misquoting and otherwise misrepresenting the statements of
others. This was far more offensive to me than his obviously condescending
"holier-than-thou" attitude, which I don't really care about (I'm old enough
and mature enough not to expect everyone I deal with [in person or on the
internet] to like me, and I don't really care though, given the choice, I
prefer to be on good terms with people).
I never questioned DM's experience and/or base knowledge of film and video
photography.
Charles
derek
> I happen to know the BBC loves them.
That's a rather misleading statement. They (VX1000) have proved
useful/'interesting' in some environments, and for some genres, but do not
meet the specs required of high-end work.
C2001
should have said "likes".
C.
David Mullen
usually try to refrain from, but one has to understand that I'm very
passionate about the whole subject of cinematography, almost like it's a
"calling", and I feel very obligated to make sure that any beginners reading
these posts don't get an inaccurate or incomplete explanation of
photographic phenomenon.
There have been a lot of wive's tales over the century about cinematographic
techniques, some of which have been promoted by Oscar-nominated
cinematographers (not John Seale, by the way, who's a very practical man).
It's because cinematography is both an art and a science that it invites
rather sloppy psuedo-scientific explanations by us laymen on how things work
(I include myself as a layman, not being a video engineer nor film stock
designer nor lab person) -- that doesn't mean that the techniques are wrong
or that the results aren't good, but that the understanding of WHY these
techniques are working is flawed.
And I believe that students should understand the WHY's, not just the HOW's.
For example, everyone can get good results by rating their color negative
stocks slightly slower than normal (overexpose them slightly) but they
should understand WHY they are doing it, rather than just doing it because
it works for some DP's. Because if they understand the process better, they
can make variations in it, or they can try different techniques to get
different results.
I have absolutely no doubt that Charles was getting better results by using
his ND filter; I just felt it was important to explain to him more clearly
what an ND filter does and DOESN'T do, because I felt he was overstating its
abilities to improve the latitude & color & gain in the video image. That
doesn't mean that he isn't getting good results with his ND filtering
technique, but in the long run, a videographer will have MORE control over
his imagemaking if he understands more clearly WHY it was working for him.
I have been on a huge learning curve this year as I shoot three digital
features back-to-back after spending the majority of my professional time
shooting 35mm film. As I understand this process better, I'm trying to pass
this hard-won knowledge onto those who are just beginning to shoot.
Anyway, I apologize for any bad feelings caused by myself being overly
agressive in correcting someone's statements, although I also stand by every
technical argument I made regarding the properties of ND filters in image
capture.
David Mullen
C2001
Thank you very much David, for your apology and for your -- obviously -- very
carefully considered explanation of precisely where your views and reactions
were coming from. Notwithstanding my own harsh words, I never questioned your
indisputable expertise and knowledge, nor the details about ND filters which you
discussed in several of your posts.
It's unfortunate that my reaction to yours went off the deep end and I'm more
than pleased to be able apologize to you unequivocally, and to others at this
group, some of whom I realize have been genuinely distressed by this thread.
I completely agree that to maximize one's skills it's best to understand
completely why things work the way they do. Sometimes this is not quite as easy
as any of us would like. I'm obviously a comparative novice to the world of pro
videography and editing and often find myself overwhelmed by the gigantic
amounts still to be learned.
Your comment about your current learning curve shooting three digital features
back-to-back certainly hit home for me in many ways.
But the bottom line is that film and video techologies are improving at a pace
that is as exhilarating as it is challenging. For me, it provides something
genuinely thrilling to look forward to as I get older.
Amen.
Charles
R.J.F. Stewart
> No, Steve, it was the opposite: I reported my experience and other people told me
> I couldn't possibly be "right" since my experience and findings didn't conform 100%
> to their understanding of video electronics.
I think we are referring to *knowledge* of electronics, something which you
cheerfully admit you do not have, as if it gave your subjective opinion greater
credence simply because an opinion cannot be refuted. I don't doubt that you can see
things on your screen that you personally interpret in a particular way, but some
things that can be scientifically measured and verified really happen in the real
world and some things don't.
Rod.
David Mullen
rather than a negative one. We're all hear to learn from each other, and
that is only possible through open communication. As you have seen, I'm
prone to debating a subject to death sometimes, which doesn't always lead to
good feelings all around but instead, can kill the process of communication,
not promote it.
But after awhile, I remembered my mother's advice when me & my sister would
be arguing as children, that sometimes the person who stops arguing first is
the only winner (being Japanese, she tends to favor avoidance of
confrontation I guess.) If only I had remembered it a few posts earlier...
Let the Great ND3 Filter Debate of 2001 be officially ended (though I'm sure
most people nodded off a long time ago anyway.)
David Mullen
C2001
I understand this and, as I expect you've noted, have buried the hatchet.
Charles
C2001
Thanks again, David.
Best regards,
Charles
R.J.F. Stewart
> I also like to know as much as possible about the related
> technology, but I'm not an electronics engineer and never aspired to be one.
I hesitate to jump in again, but if you're working with television equipment
isn't this a bit of a contradiction?
Rod.
CT
Should have said "will tolerate suppliers using if justified"
C2001
> C2001 wrote:
> > derek wrote:
> > > > I happen to know the BBC loves them.
> > >
> > > That's a rather misleading statement. They (VX1000) have proved
> > > useful/'interesting' in some environments, and for some genres, but do
> > > not meet the specs required of high-end work.
> >
> > should have said "likes".
>
> Should have said "will tolerate suppliers using if justified"
<g>
C.
C2001
> C2001 wrote:
> > I also like to know as much as possible about the related technology,
> > but I'm not an electronics engineer and never aspired to be one.
>
> I hesitate to jump in again, but if you're working with television equipment
> isn't this a bit of a contradiction?
Not necessarily -- I'd say it depends on context.
Being an independent videographer/editor is not the same as working full-time for
a network. And there are *many* permutations within both these categories so,
frankly, I'd prefer to avoid making further statements which, inevitably, will to
some extent be generalizations and as such potentially subject to further
misunderstanding and contradiction from others (which, IINM, is what caused this
thread to become as acrimonious as it did for a while).
If you check out my response this morning to David's post to me, you'll see that I
agreed unequivocally with his statement that it's preferable for videographers to
know and undertsand as much as possible about their equipment's technology.
Speaking for myself, I can state only that I can learn only so much at a time (I'm
not a teenager and my health and energy are limited).
Believe it or not, I _would_ like to learn more about the engineering side of
video but at the moment I'm still learning the more advanced functions of Avid
software (which is comparable to studying a foreign language) as well as my DSR300
-- not to mention I recently completed two harrowing months editing a project that
was entirely in French which I hadn't used for years (I told the client, a work
colleague from more than a decade ago, that she should use a Francophone editor in
Montreal, but she insisted she wanted me mainly because she considered I could
bring an artistic and intellectual perspective to the project which not all
veteran editors could equal).
Without going into the entire story of my life for the past three years, I feel
I'm _already_ probably trying to do too much at present -- but that's for me to
sort out, on my own.
Charles
Rob Mora
> problems: some back-focus drift, some shots that I used heavier diffusion
> and now regret it, some loss of resolution when the Canon zoom lenses were
> shot wide-open, and one shot where the focus is simply off (otherwise, my
> focus-puller did an excellent job.) It's a problem that I tried to take
> more seriously in the second 24P HD feature I shot.
Hi, David
You may have posted this info in another thread, but I'm very curious
anyway. Hope you don't mind:
Did you use the Panavision-modified Sony camera, or the stock version?
Which lens(ses) did you use? -- Speaking of which, am I correct in
assuming that the Sony F900 has no servo motor control for these new
HD lenses?? (I was at this year's Sundance and got to play around w/
the F900 at the Digital Center, but forgot to ask this)
I always enjoy your posts. Thanks so much for taking time out of your
schedule to share your knowledge!
Rob
Matt Sandstrom
> I reported my experience and other people told me
> I couldn't possibly be "right" since my experience and findings didn't conform 100%
> to their understanding of video electronics.
nobody ever questioned your experience. all people did was to tell you
that your *interpretation* ("nd filters increase the shadow detail of
vx1000 cameras") was wrong and that there must be something else going
on that you probably wasn't able to see and articulate due to your
limited knowledge and experience.
/matt
David Mullen
It was the standard F900, not the Panavised one.
>Which lens(ses) did you use? -- Speaking of which, am I correct in
>assuming that the Sony F900 has no servo motor control for these new
>HD lenses??
I used the Canon 5.5-50mm and 7.8-144mm HD zooms. These lenses come in both
"cine" style (straight lens) and ENG style (standard built-on zoom motor &
exposure control.) I used the cine style on the first feature, requiring a
separate zoom motor to go on the iris rods, and the ENG-style one on my last
feature. There are advantages to both -- mainly that when you go handheld,
the ENG-style is easier to hold because of the grip/zoom motor combo. But
the cine style is probably lighter.
David Mullen
C2001
I'm not going to comment further since pretty well everyone has had their say and -- more
important -- David and I have both apologized to each other and buried the hatchet. In
retrospect, I'd say that what happened was a severe misunderstanding which unfortunately
hit raw nerves for both of us.
Obviously, had I suspected that a fleeting and (I do realize) ambiguously worded,
reference to my experience with the Tiffen ND3 could generate such an unpleasant debate,
I would never had made the post.
Charles