Nikon D1x beats Velvia
Vagabond
http://foto.no/nikon/D1X_review.html
Vagabond
Randy Howard
It looks like the comparison is between a native digital
image from the D1X and one scanned in from Velvia at
2700 DPI with a Nikon LS-2000. I think to make it
fair, the top of the line digital image should be
compared with something from the best available scanner
or even a drum scan. Also, I'm not sure that Velvia
is as good a choice for this type of comparison as
Provia 100F.
Vagabond
wrote:
> I think to make it
>fair, the top of the line digital image should be
>compared with something from the best available scanner
>or even a drum scan. Also, I'm not sure that Velvia
>is as good a choice for this type of comparison as
>Provia 100F.
I was also thinking about 100F - it has finer grain than Velvia.
Velvia looks sharper but that is because it has such high contrast.
As for the quality of the scan, Bjørn Rørslett (whose integrity I
don't question) addresses this issue in the article. But it would be
nice to have a top-end drum scan to compare with - I agree.
I wonder if there are more D1x tests to find on the net. The camera is
out and I played with one at my local camera store yesterday. Not that
I could afford it though...
Vagabond
Randy Howard
> On Fri, 8 Jun 2001 12:44:15 -0500, Randy Howard <ra...@thegateway.net>
> wrote:
> > I think to make it
> >fair, the top of the line digital image should be
> >compared with something from the best available scanner
> >or even a drum scan. Also, I'm not sure that Velvia
> >is as good a choice for this type of comparison as
> >Provia 100F.
> *snip*
>
> I was also thinking about 100F - it has finer grain than Velvia.
> Velvia looks sharper but that is because it has such high contrast.
Exactly.
> As for the quality of the scan, Bjørn Rørslett (whose integrity I
> don't question) addresses this issue in the article. But it would be
> nice to have a top-end drum scan to compare with - I agree.
I was not trying to impugn Bjørn's work at all, merely that I would
like to see the results on a higher quality scan. It seems like a bit
of an unfair playing field to compare a native digital against a
film image converted to digital, unless the image quality loss in
the transformation can be minimized as much as possible.
> I wonder if there are more D1x tests to find on the net. The camera is
> out and I played with one at my local camera store yesterday. Not that
> I could afford it though...
I've been lurking on the D1 mailing list, and the early reports seem to
be pretty favorable so far, but I'm not sure I want to be an "early
adopter" of this one. I'm not entirely sure if you can actually perform
a firmware upgrade on it yourself (as you can do with even the cheap
Kodak digital cameras).
There are also some reports of problems with flash photography, even
with the SB28-DX. Not enough data available yet to be in any way
conclusive though.
You probably don't want to be a "beta" customer for the thing at that
price point.
Paul Rubin
> I wonder if there are more D1x tests to find on the net. The camera is
> out and I played with one at my local camera store yesterday. Not that
> I could afford it though...
Yes, someone posted some from a .il site yesterday. The low ISO
images were great. There was also an ISO 3200 shot that look like
a video frame capture. Film seems to still be winning in that area,
at least against the D1x, though maybe not against say the Kodak DCS720.
Max Perl
"Vagabond" <reto...@hotmail.com> skrev i en meddelelse
news:3b210cff...@news.uib.no...
But still....if you want to project your images there are still no
alternative to
a good slide projecter. But this will also come some day.
If Fuji wants to spend more money on developing film it should still be
possible
to make big improvements in this area.
Remember that the process in making a CCD chip is an optical process :)
But probably the wawelength used to expose the material is another than
visible light........
Max
Vagabond
<max_...@post11.tele.dk> wrote:
>If Fuji wants to spend more money on developing film it should still be
>possible to make big improvements in this area.
It is several years ago now that I heard of a Fuji film developing
engineer who had (during an unguarded moment at Photokina) bragged
that the company had films in development that would have 10X the
resolution of Velvia.
Then a while after this came - viola - Provia 100F. It isn't 10X
Velvia, but it is a good start. If Fuji keeps improving things at this
rate there is hope for good old chemical film. Digital chips will meet
the law of diffraction sooner or later. I am not expert in this, but I
understand the nature of light itself will limit how many sensors you
can put on a chip of a given size. Just as the quality of the optics
is a limiting factor in how much resolution you theoretically can get
out of film. So if you want higher quality than what a 35mm chip can
provide, you have to go to larger formats. Just as with film.
Vagabond
Randy Howard
> Digital chips will meet
> the law of diffraction sooner or later. I am not expert in this, but I
> understand the nature of light itself will limit how many sensors you
> can put on a chip of a given size. Just as the quality of the optics
> is a limiting factor in how much resolution you theoretically can get
> out of film. So if you want higher quality than what a 35mm chip can
> provide, you have to go to larger formats. Just as with film.
Interesting... are there "medium format" digital cameras with very large
CCD's (yet)?
Scott Jaworski
Tony Polson
Paul Rubin
> The header should've read "Nikon D1X beats LS-2000"...
As he explained, he also examined the slides directly under 40x
magnification and stuff was showing up in the D1x images that was
simply not visible on the slides.
Mxsmanic
news:3b21381b...@news.uib.no...
> ... I understand the nature of light itself will
> limit how many sensors you can put on a chip of
> a given size.
The same fundamental limits apply to film. A full-size CCD (36x24) can
easily have better resolution than almost all film, however.
> Just as the quality of the optics is a limiting
> factor in how much resolution you theoretically
> can get out of film.
It's a limiting factor for digital, too.
> So if you want higher quality than what a 35mm
> chip can provide, you have to go to larger
> formats. Just as with film.
Yes. Except it is difficult to imaging a CCD as large as an MF film
frame, but who knows?
Mxsmanic
news:MPG.158b1b414...@news.thegateway.net...
> Interesting... are there "medium format" digital
> cameras with very large CCD's (yet)?
There are scanning backs, but I haven't heard of any 6x7-cm CCDs.
Mxsmanic
news:3B21611F...@home.com...
> The header should've read "Nikon D1X beats LS-2000"...
That would be more difficult than beating Velvia, it seems.
Meryl Arbing
other supposed film vs digital pages have made. The most basic error is that
the final comparison is NOT between the output of a digital camera and the
output of a film camera but it is comparing the output of a digital camera
with the output of a scanner!! It says nothing about the film camera!
If the page wants to claim that the latest Nikon digital SLR produces a
better picture than a scanner that is NOT the latest from Nikon that is a
debatable point since the conclusions are purely subjective. It is rather
hard to believe that the interpolated 3008x1960 image from the D1X is close
to the 3894x2594 digital image that the LS2000 delivers. Of course, the
image from the digital camera is given a few software "helping hands" that
the scanned image doesn't get!! Perhaps the 5645x3654 image from the LS4000
would be a better test of scanner to digital.
Next, if we look specifically at the experimental setup there are many
conceptual errors that bring the conclusions into serious question.
The claim is made that since the tripod didn't move, that means "the final
image magnification would be identical in all cases". Not so! Since the D1X
is still not a full frame CCD (actually 23.7mmx15.6mm) there is still a
focal length multiplier that is applied to the digital camera and not to the
film camera.
He says that he examined the slide under 40x magnification and came to the
subjective opinion that there was less detail in the slide than in the
digital image. Quite unlikely and probably based on seeing what he wanted to
see.
I doubt the results would be the same if he used a projector to view the
slide compared to a digital projection of the D1X image.
Just another failed attempt to compare apples to pumpkins!!news:3b210cff...@news.uib.no...
Paul Rubin
> I had a look at the page. Unfortunately, the page makes the same errors that
> other supposed film vs digital pages have made. The most basic error is that
> the final comparison is NOT between the output of a digital camera and the
> output of a film camera but it is comparing the output of a digital camera
> with the output of a scanner!! It says nothing about the film camera!
That's funny. I read the page and it says
These test images should provide some food for thought. It could be
argued that the setup was biased against film, because higher quality
scans obviously could be done on the RVP image. To assess this, I
carefully examined the film image at 40X magnification to locate tiny
details observed on the D1X image, and I can just say that RVP simply
had less resolution, because a number of the finest details simply did
not occur on film. No scanner of any quality can possibly extract
details which are not present! My personal conclusion is that today's
top digital cameras now compete against medium-format equipment,
having won the war with 35 mm film by a comfortable margin.
That is, he also compared the D1x image directly with looking at the
slide through a loupe.
> If the page wants to claim that the latest Nikon digital SLR produces a
> better picture than a scanner that is NOT the latest from Nikon that is a
> debatable point since the conclusions are purely subjective.
Remember, he also compared the D1x image directly with looking at the slide
through a loupe.
> It is rather hard to believe that the interpolated 3008x1960 image
> from the D1X is close to the 3894x2594 digital image that the LS2000
> delivers. Of course, the image from the digital camera is given a
> few software "helping hands" that the scanned image doesn't get!!
> Perhaps the 5645x3654 image from the LS4000 would be a better test
> of scanner to digital.
As mentioned above two times, he also compared the D1x image directly
with looking at the slide through a loupe.
> Next, if we look specifically at the experimental setup there are
> many conceptual errors that bring the conclusions into serious
> question. The claim is made that since the tripod didn't move, that
> means "the final image magnification would be identical in all
> cases". Not so! Since the D1X is still not a full frame CCD
> (actually 23.7mmx15.6mm) there is still a focal length multiplier
> that is applied to the digital camera and not to the film camera.
He used the exact same lens for both cameras. The focal length didn't
change. The 35mm frame was larger than the D1x frame, so it had stuff
in it that the D1x frame didn't, because the magnification was the same.
> He says that he examined the slide under 40x magnification and came
> to the subjective opinion that there was less detail in the slide
> than in the digital image. Quite unlikely and probably based on
> seeing what he wanted to see.
Then again, perhaps the wishful thinking is on your part.
> I doubt the results would be the same if he used a projector to view the
> slide compared to a digital projection of the D1X image.
There's no such thing as a 5 megapixel digital projector.
> Just another failed attempt to compare apples to pumpkins!!
"A high number of people refuse to accept that film is to be eclipsed
by digital media. Let them enjoy their dream world while it lasts."
The one question I can raise is whether the F5 picture was precisely
focused. It's possible the F5 focus detector was slightly off. It
would be nice if he did some slight focus bracketing.
Graham Stewart
Well how would you suggest comparing them?
It's all good and well for you to slate this as an invalid test, but please
provide a bit of feedback on a better method.
Comparing a slide through a 40x loupe and using photoshops zoom tool seems
like a very good comparison - like for like.
As for the use of the scanner - i believe that's perfectly valid. Almost all
my images are scanned from film, before being tinted, cropped, touched up
and output to Fuji Crystal Archive paper for presentation.
I appreciate there are still people who are vehemently opposed to digital
techniques but they are becoming less populous. Since my standard image
route is expose->slide film->process->scan->touchup->print this test
provides an excellent reference. If you dont use digital in your imaging
then what does it matter that a digital camera can provide arguably better
results.
I'm confident that digital will soon surpass film in almost all respects and
as soon as it does, i'll switch.
Graham
David Kieltyka
> As he explained, he also examined the slides directly
> under 40x magnification and stuff was showing up in
> the D1x images that was simply not visible on the slides.
I've seen this sort of thing even with my Nikon CoolPix. It's
artifacting...not real image data.
-Dave-
Max Perl
"Randy Howard" <ra...@thegateway.net> skrev i en meddelelse
news:MPG.158b1b414...@news.thegateway.net...
For Hasselblad e.g. PhaseOne makes digital backs. I think the one with
highest
resolution is a 5120 X 5120 as far as I remember. The price is high....about
$30.000.
You can go higher than that. For Linhof a digital back exist with a
resolution about
8000 x 9000 (can't remember exact)......have a look at their homepage.....
Max
Max Perl
"Meryl Arbing" <mar...@sympatico.ca> skrev i en meddelelse
news:_teU6.323$ht1.1...@news20.bellglobal.com...
No...this is a big misunderstanding. The D1X just uses less of the glass.
The magnification is exactly the same. The distance from the rear lens
element to the CCD or film is the same for D1X and F5.
Mxsmanic
news:_teU6.323$ht1.1...@news20.bellglobal.com...
> Perhaps the 5645x3654 image from the LS4000
> would be a better test of scanner to digital.
With 35mm film, a 5645x3654 image is likely to just be a larger,
blurrier version of a 3800x2700 image. Most 35mm film photographs
simply don't have that much detail. In particuarly, virtually nothing
shot without a tripod, or without extremely fine-grained film, will show
much of anything beyond 2800 dpi or so. Note that 2700 dpi is 53 lp/mm,
and a great deal of ordinary photographic work yields only 40 lp/mm or
so at best, so 2700 dpi is all you need.
Mxsmanic
news:hrgU6.10200$Oy2....@typhoon.mw.mediaone.net...
> I've seen this sort of thing even with my
> Nikon CoolPix. It's artifacting...not real
> image data.
What sort of artifacts, and how do you know?
Tony Spadaro
F 100, Kodachrome 25, Tech Pan, and Sometimes TMX. Good lenses are needed
too, but quite a bit of what I've scanned has been hand held.
I won't comment on the contest between the two cameras however. I'll
wait until I can get to make my own comparisons.
--
The Camera-ist's Manifesto
a Radical approach to photography
http://home.nc.rr.com/tspadaro/
Chapel Hill artist Tony Spadaro
Dark Alley Photography - a Java site
http://www.homeusers.prestel.co.uk/magor/tony
For the Provia F - Elitechrome comparison,
The Diffraction Test,
And the loooooong view of Franklin St.
http://tspadaro.homestead.com/ArtShow.html
"Mxsmanic" <mxsm...@hotmail.com> wrote in message
news:jXnU6.166458$NK4.10...@news2.aus1.giganews.com...
David Kieltyka
> David Kieltyka <dav...@msn.com> wrote:
>
> > I've seen this sort of thing even with my
> > Nikon CoolPix. It's artifacting...not real
> > image data.
>
> What sort of artifacts, and how do you know?
Easy...I've compared digicam photos with the actual subjects photographed.
In some cases the digicam has added "detail" that isn't really there...like
crosshatch patterns on flat, monochromatic surfaces. Some sort of
interpolation error, I guess. I imagine it'll get cleared up as the
technology improves.
-Dave-
Steve
>> Interesting... are there "medium format" digital
>> cameras with very large CCD's (yet)?
>There are scanning backs, but I haven't heard of any 6x7-cm CCDs.
There are one-shot backs for medium and large format. A D1X is much
much cheaper though...
Steve
Steve
>> Perhaps the 5645x3654 image from the LS4000
>> would be a better test of scanner to digital.
>With 35mm film, a 5645x3654 image is likely to just be a larger,
>blurrier version of a 3800x2700 image. Most 35mm film photographs
>simply don't have that much detail. In particuarly, virtually nothing
>shot without a tripod, or without extremely fine-grained film, will show
>much of anything beyond 2800 dpi or so. Note that 2700 dpi is 53 lp/mm,
>and a great deal of ordinary photographic work yields only 40 lp/mm or
>so at best, so 2700 dpi is all you need.
Hence the need for Nikon to come out with a 4000 dpi scanner...?
You're quite wrong here from my experience. 4000 dpi is only starting
to approach the resolution needed to faithfully capture what's on film.
A modified Nyquists theorem holds here - "scan at twice the highest dot
frequency..."
However, digital seems to be offering an advantage that film so far
doesn't - greater tonal range. I believe that the Dicomed 4x5 system
boasts 9 stops of exposure latitude...
Steve
Mxsmanic
news:9g2git$1...@nntpa.cb.lucent.com...
> There are one-shot backs for medium
> and large format.
How do they capture the image? I find a CCD that measures inches on a
side to be difficult to believe.
Mxsmanic
> Hence the need for Nikon to come out with a
> 4000 dpi scanner...?
Some people think more DPI is better, and some believe that film has
infinite resolution that can be probed eternally with ever-higher
resolutions on scanners.
> You're quite wrong here from my experience. 4000
> dpi is only starting to approach the resolution
> needed to faithfully capture what's on film.
With Technical Pan, perhaps. Not with Royal Gold or Tri-X.
> However, digital seems to be offering an advantage
> that film so far doesn't - greater tonal range. I
> believe that the Dicomed 4x5 system boasts 9 stops
> of exposure latitude...
Digital has always been ahead in just about every respect except
resolution. CCDs that are cooled can yield 18 stops of dynamic range.
Steve
I'm not sure, but I don't think they scan. Do a search for "digital
backs contax" for instance and you'll find quite a bit.
Steve
Steve
>> Hence the need for Nikon to come out with a
>> 4000 dpi scanner...?
>Some people think more DPI is better, and some believe that film has
>infinite resolution that can be probed eternally with ever-higher
>resolutions on scanners.
>> You're quite wrong here from my experience. 4000
>> dpi is only starting to approach the resolution
>> needed to faithfully capture what's on film.
Keep in mind, I said faithfully capture whats on the film. That
assumes you captured what the film is capable of. Poor latitude kind of
screws up the wonders of film vs. digital.
>With Technical Pan, perhaps. Not with Royal Gold or Tri-X.
I think also with Velvia and Provia. I don't know about actual
resolution, but to sample the number of dye blobs on the film, one would
have to oversample by 2 at least. The analagous resolution in dye blobs
was estimated at 13 megapixels I believe. What I do know is that a
4000 dpi scan of Provia beats a 2800 dpi scan hands down. But what
Bjorn has shown is that at a mere 6 megapixels, digital seems to beat
film. Now where can I sell that F5...
>> However, digital seems to be offering an advantage
>> that film so far doesn't - greater tonal range. I
>> believe that the Dicomed 4x5 system boasts 9 stops
>> of exposure latitude...
>Digital has always been ahead in just about every respect except
>resolution. CCDs that are cooled can yield 18 stops of dynamic range.
Nice....so I keep my camera on ice before using it?
Steve
Randy Howard
> What I do know is that a
> 4000 dpi scan of Provia beats a 2800 dpi scan hands down.
Correct, clearly in defiance of Mxsmanic' bogus original claim.
> But what Bjorn has shown is that at a mere 6 megapixels, digital
> seems to beat film.
Actually, what he has shown that a 2700dpi scan of Velvia (not
the finest grain film available) may not stack up against the
D1X. Given that we don't have the original bricks to compare
with, it is difficult to be sure that artifacting isn't
involved. Even taking his word for it (I doubt he's making
things up anyway), we still would need to see a 4000dpi
scan of a better film, perhaps Provia, to get a better
idea of how they compare. Also, most production shots
to be blown up, say for a magazine cover are drum scanned,
that would be a better test for absolute picture quality.
> Now where can I sell that F5...
I think it's still a bit early yet...
Steve
>> What I do know is that a
>> 4000 dpi scan of Provia beats a 2800 dpi scan hands down.
>Correct, clearly in defiance of Mxsmanic' bogus original claim.
>> But what Bjorn has shown is that at a mere 6 megapixels, digital
>> seems to beat film.
>Actually, what he has shown that a 2700dpi scan of Velvia (not
>the finest grain film available) may not stack up against the
>D1X. Given that we don't have the original bricks to compare
>with, it is difficult to be sure that artifacting isn't
I think Bjorn is very careful and I'm beginning to think the end of film
in the volume we know is sooner than we all thought. Damn, if they then
come out with a 12 megapixel camera....
>involved. Even taking his word for it (I doubt he's making
>things up anyway), we still would need to see a 4000dpi
Remember, no matter the film, the lattitude of digital is going to allow
for many more details that film (transparancies) can't catch.
>scan of a better film, perhaps Provia, to get a better
>idea of how they compare. Also, most production shots
While Provia is less grainy, Velvia still wins the resolution contest,
making it the perfect film for this comparison.
>to be blown up, say for a magazine cover are drum scanned,
>that would be a better test for absolute picture quality.
But if there are details that film just isn't catching, it doesn't
matter. And digital is grainless - a very neat fact indeed.
Steve
Mxsmanic
> But what Bjorn has shown is that at a mere
> 6 megapixels, digital seems to beat film.
It's hard for the human eye to see much more than six megapixels, which
is why images with 20 megapixels don't look that much better.
> Nice....so I keep my camera on ice before
> using it?
I suppose a Peltier cooler or something could be built in. I don't
expect to see that in ordinary digital cameras any time soon, however.
Mxsmanic
> Correct, clearly in defiance of Mxsmanic' bogus
> original claim.
How is the original image photographed? Handheld? With what kind of
lens?
> Also, most production shots to be blown up,
> say for a magazine cover are drum scanned ...
Which is massive overkill, because no magazine cover can use either the
dynamic range or the resolution of a drum scan. A 13x19" tabloid cover,
for example, requires only about 2900x2000 pixels with a 150-line screen
(and that is a very high screen frequency), and the dynamic range will
be considerably less than 100:1.
Randy Howard
> >Actually, what he has shown that a 2700dpi scan of Velvia (not
> >the finest grain film available) may not stack up against the
> >D1X. Given that we don't have the original bricks to compare
> >with, it is difficult to be sure that artifacting isn't
>
> I think Bjorn is very careful and I'm beginning to think the end of film
> in the volume we know is sooner than we all thought. Damn, if they then
> come out with a 12 megapixel camera....
I certainly don't contest this. I just think I'll want a film
camera until film is no longer available, or I die, whichever
comes first. Anyone make an infrared CCD camera yet?
> Remember, no matter the film, the lattitude of digital is going
> to allow for many more details that film (transparancies) can't catch.
Perhaps, I don't argue that digital is going to be very, very good.
I'm just not sure it's already surpassed film (at least with current
cameras selling < $5000. Even so, we need to see it in camera bodies
prices in the $1000 range before it's really "here".
> While Provia is less grainy, Velvia still wins the resolution contest,
> making it the perfect film for this comparison.
How is a coarser grain film able to provide more resolution? Most
people think it *looks* sharper, based upon the popped colors.
> But if there are details that film just isn't catching, it doesn't
> matter.
True. I don't find that many situations where film is not more
than capable of capturing the details. A look at the work of
the world's best film photographers would seem to imply the
same thing.
> And digital is grainless - a very neat fact indeed.
Oh really? I've seen some EXTREMELY grainy looking digital photos.
They have "grain" any time they are displayed above the native
resolution of the CCD they were captured on. They also show
something analogous to grain when color-depth limitations can
be clearly seen. Try a sunset picture on a digital camera, very
few of them, even with heavy post-processing look as smooth
in color variations as film.
Paul Rubin
> I certainly don't contest this. I just think I'll want a film
> camera until film is no longer available, or I die, whichever
> comes first. Anyone make an infrared CCD camera yet?
Yes, CCD's are sensitive down into the infrared and require an IR Cut
filter to keep infrared out. Most consumer digicams work down into
the infrared. You can take infrared photos by putting an infrared
filter over the lens. Sony video cameras even have a built-in switch
("Night Shot") to slide out the IR Cut filter and activate some
built-in infrared LED's so you can shoot video in total darkness. I
don't know whether you can also shoot infrared stills with those
cameras.
> Perhaps, I don't argue that digital is going to be very, very good.
> I'm just not sure it's already surpassed film (at least with current
> cameras selling < $5000. Even so, we need to see it in camera bodies
> prices in the $1000 range before it's really "here".
At the amateur level, that sounds about right. At the professional
level the D1x quickly pays for itself in film savings alone.
> Oh really? I've seen some EXTREMELY grainy looking digital photos.
> They have "grain" any time they are displayed above the native
> resolution of the CCD they were captured on. They also show
> something analogous to grain when color-depth limitations can
> be clearly seen. Try a sunset picture on a digital camera, very
> few of them, even with heavy post-processing look as smooth
> in color variations as film.
That's CCD noise, which is analagous to grain.
VT
<ra...@thegateway.net> wrote:
>In article <9g3ek0$6...@nntpa.cb.lucent.com>, s...@randomc.com says...
>> While Provia is less grainy, Velvia still wins the resolution contest,
>> making it the perfect film for this comparison.
>
>How is a coarser grain film able to provide more resolution? Most
>people think it *looks* sharper, based upon the popped colors.
>
Film resolution isn't simply just limted by grain - there are many
other factors -
Here are the actual specs
Grain Resolution
Film Name RMS Lo Hi Sharp
Fuji Provia 100F 8 60 140 15 RDP3
Fuji 100F batch >012 8 60 140 15 RDP3
Fuji Velvia 50 9 80 160 25 RVP
So despite the fact Provia 100F is finer grained than Velvia - it has
a lower resolution rating at both the low contrast (1:1.6) and high
contrast (1:1000) - also Velvia is sharper (higher acutance)
Bookmark this useful film characteristics site:
--
Vincent
vtVi...@Prodigy.Net
VT
wrote:
>Which is massive overkill, because no magazine cover can use either the
>dynamic range or the resolution of a drum scan. A 13x19" tabloid cover,
>for example, requires only about 2900x2000 pixels with a 150-line screen
>(and that is a very high screen frequency)
I think that figure of the pixel count may be wrong
The general rule is, I believe, 2x the screen line resolution - so for
a 150-line screen the ppi required is
2x150 = 300ppi
or one can get away with 1.5x the screen line so that's still
150 x1.5 = 225ppi
So for a 13x19" tabloid cover (at 150-line screen) one actually needs
ideally (2x ie: 300ppi)
3900 x 5700 = 22.23Mp
or even at 1.5x the line screen (225ppi) this is still
2925 x 4275 = 12.5Mp
--
Vincent
vtVi...@Prodigy.Net
VT
wrote:
>Digital has always been ahead in just about every respect except
>resolution. CCDs that are cooled can yield 18 stops of dynamic range.
I'm really impressed with that figure -
Can you please point me to any reference that gives that kind of
dynamic range in practical photographic usage?
My understanding so far under most practical applications most
consumer grade digital cameras yield at maximum 8-stops dynamic range
- even that may be dubious due to noise and lower order bit loss.
It is true that there are now digi-cams that boast 12-bit or even
14-bit internal image processing (per color) - but that is wider
processing - _not_ necessarily the actual dynamic range of the
CCD/CMOS used.
I am not aware of any commercially available consumer digi-cam that
yields more than 8-bits/color -ie: 8-stop dynamic range.
I would really pleased to hear otherwise.
Note: Film can easily yield a dynamic range of 10 stops.
--
Vincent
vtVi...@Prodigy.Net
Annika1980
>But what
>Bjorn has shown is that at a mere 6 megapixels, digital seems to beat
>film. Now where can I sell that F5...
Toys R' Us?
EDGY01
lens blown up to 20 inches across. Was printed on a Kodak LED poster printer
(chemical). It was then that I finally realized that film is over. Digital is
it. When Nikon releases the D2 with a CMOS imager of 24mm by 36mm
dimensions,--I'm jumping in with both feet. If the D1X is that much better
than the D1,--wow.
dan Lindsay
santa barbara, ca
Paul Rubin
Is that a genuine rumor, that the D2 will have 24x36mm CMOS? Or
just a guess?
VT
>About 18 months ago I had a TIFF image that I took with my D1 and 105mm f/2.5
>lens blown up to 20 inches across. Was printed on a Kodak LED poster printer
>(chemical). It was then that I finally realized that film is over. Digital is
>it.
With all due respect (and I do think the Nikon D1 is a very fine
digital SLR) printing a Nikon D1 image - which is all of 2012x1324
pixels to 20" across is using in about 100ppi.
At 100ppi an image is very likely to be pixellized on anything
resembling any detail - or if upsampled/interpolated will be pretty
soft -
Since 100ppi is the equivalent of about 2lp/mm on paper -
This is 1/4 what Leica considers quality (ie: 8lp/mm) and even for
most mortals this is still 1/3 the requirement of the generally
accepted 6lp/mm on paper. 4lp/mm on paper is pretty marginal but
acceptable on larger sized prints that do not require close scrutiny -
even then the Nikon D1 can only deliver 1/2 that resolution.
I think that the Nikon D1 may have replaced film for some - but for
many this kind of resolution may be less than adequate.
--
Vincent
vtVi...@Prodigy.Net
EDGY01
believer. I could shoot a section of the print with my 950 to show you...
Dan
EDGY01
just a guess?
A: That's my WISH. I strongly suspect that it will be full frame 35mm but I
don't think they'll be ready with CMOS yet. Will probably be CCD. CMOS takes
significantly less power to operate, and that's the other big issue with the
current generation of digital cameras,--power consumption.
Dan Lindsay
SB, Cal.
Paul Rubin
> At 100ppi an image is very likely to be pixellized on anything
> resembling any detail - or if upsampled/interpolated will be pretty
> soft -
Please look at
http://home.wanadoo.nl/~nibbit/foto.jpg
and tell me if you still think that. What do you think its original
resolution was?
Dr_K_H
and also
4in x 5 in
and also
6cM x 6cM
and so on....
VT
>No matter what you calculate, I had NO pixellation on that print. It made me a
>believer. I could shoot a section of the print with my 950 to show you...
Thank you for the offer - why a 950?
What happened to your Nikon D1?
Is it possible to post the original image that you had printed to 20"
across somewhere and possbly a photo of a section that has some fine
detail.
I am still surprised that at 100ppi (no matter how "gpod" the pixels)
is still 100ppi any diagonal or non-vertical non-horitontal detail
will show pixellation - unless the image has no detail to begin with.
Printing on a good professional printer would probably involve
upsampling/interpolation by the printing system - those would avoid
pixllation but the image would tend to be soft'ish - again this is all
caveat by the amount of detail the image subject had in the first
place.
The 2012x1324 (2.6Mp) of the Nikon D1 is the equivalent to 28 lp/mm at
the (35mm) film plane..... think about it - this is worse than the
44lp/mm of a Kodak MAX disposable camera as tested by Pop Photo. I am
not saying that a disposable camera can match the Nikon D1 - but this
shows where digital is currently at......
Some may assert that digital has replaced film -
but digital enthusiasts have been proclaiming that since the 1
mega-pixel digi-cam days, so when 2Mp arrived - 1mp can no longer
replace film but 2Mp could - so when 3Mp arrived - you get the
picture..... (pixellized of course :-)
However digital is still coming fast and strong with 6-Mpixel
full-frame SLRs announced by both Contax and Pentax (although I am a
bit concerned at their continued non-delivery of production models).
So it would not surprise me if Nikon and Canon are not too far
behind......
I think by the time 7 to 8 mega-pixel full-frame SLRs become
commonplace (ie: the ability to deliver 300ppi for 10x8 prints), the
film replacement argument would become more valid -
in the meantime, it just gives us something to talk about..........
--
Vincent
vtVi...@Prodigy.Net
Randy Howard
says...
> > Perhaps, I don't argue that digital is going to be very, very good.
> > I'm just not sure it's already surpassed film (at least with current
> > cameras selling < $5000. Even so, we need to see it in camera bodies
> > prices in the $1000 range before it's really "here".
>
> At the amateur level, that sounds about right. At the professional
> level the D1x quickly pays for itself in film savings alone.
I wonder about that. First, you need to spring for a pocketfull of
compact flash cards, large ones. Something like the Lexar 12X units,
which are not cheap, or a microdrive which eats batteries and by
reports from D1 users, not very reliable under heavy use.
Then, your computer needs a new LARGE hard drive, a big monitor,
a lot of memory and a very high quality printer and a ton of
ink for it.
Then, you need to spring for photoshop, and whatever add-ins you
want/need for it. Then, you get to spend many hours learning the
software, what works, what doesn't with various images, colorspaces,
ad nauseam. Then, whatever that process is, how many hours does
it take you to post-process, sort, crop, 36 shots? Then, you
archive both the original, and modifications, then archive them
either on a hard drive, or recordable media, preferably both.
It shouldn't take long to have you questioning if your time
alone isn't worth a lot more than a few bucks for film and
processing.
> > Oh really? I've seen some EXTREMELY grainy looking digital photos.
> > They have "grain" any time they are displayed above the native
> > resolution of the CCD they were captured on. They also show
> > something analogous to grain when color-depth limitations can
> > be clearly seen. Try a sunset picture on a digital camera, very
> > few of them, even with heavy post-processing look as smooth
> > in color variations as film.
>
> That's CCD noise, which is analagous to grain.
Precisely why I was confused when you said digital photos have no
grain, semantics aside.
Mxsmanic
news:3b25a1f1...@news.prodigy.net...
> At 100ppi an image is very likely to be
> pixellized on anything resembling any detail
> - or if upsampled/interpolated will be pretty
> soft -
Whether or not that is important depends on the viewing distance. For a
picture on a wall, it's fine.
> ... even then the Nikon D1 can only deliver
> 1/2 that resolution.
There is much more to image quality than resolution, despite the
obession of many photographers with it, and if the image quality is
superlative in other aspects, as it is likely to be with the D1x,
resolution isn't that important.
Mxsmanic
news:7xvgm2t...@ruckus.brouhaha.com...
> Please look at
>
> http://home.wanadoo.nl/~nibbit/foto.jpg
>
> and tell me if you still think that.
It's pretty soft.
> What do you think its original
> resolution was?
About 1024x768, I'd guess. The softness and some artifacts would seem
to indicate that the image on the page has been interpolated from a
smaller original image.
VT
wrote:
I looked and on the screen, it looks fine for a screen image -
although like any image displayed on the screen on close scrutiny one
can see the actual screen pixels that make up the image (ie:
pixellation) the original image was 1600x1200.......
However I don't think that is what you meant - so I downloaded the
photo - then cropped a section of it at 532 x 401 and printed that
section out at 100ppi (ie: to 5.32 x 4" size) with an Epson Stylus
Photo 780 using Red River Premium Glossy paper using all the photo
settings of the printer (ie: 1440x720dpi - microweave on, high-speed
off, ICM)
The resultant print would have been a section of a 16x12" print from
that file, ie: at 100ppi -
The image looks fine - but with what I would consider overall softness
- and on close scrutiny especially at edge between the background and
the right side of the child's face there are some obvious edge
artifacts (which could be jpeg artifacts or sharpening by the camera)
- remember this would have been a large - 16x12" print from the
original 1600x1200 image - that is 100ppi or about 2lp/mm printed on
paper.
Note there was _no_ fine detail in the subject to speak of, the hat 's
detail is basically coarse and the pattern is high contrast but not
exactly hard edge or fine detail - the wire fence in the background
cannot be judged because they are basically out of focus images - that
leaves the child's face and eyes - none of the features on the
equivalent of 16x12' print would have been considered particularly
sharp - I would say they were actually soft - that's not saying it
wasn't a "nice photo" - just that there wasn't enough detail that
demands any higher resolution to be pleasing
- but the image printed to that size (@ pixel density of 100ppi) was
basically soft.
--
Vincent
vtVi...@Prodigy.Net
Paul Rubin
> > http://home.wanadoo.nl/~nibbit/foto.jpg
> >
> > and tell me if you still think that.
>
> It's pretty soft.
For sure, it's not as good as a D1 picture or Leica scan, but it's
viewable.
> > What do you think its original resolution was?
>
> About 1024x768, I'd guess. The softness and some artifacts would seem
> to indicate that the image on the page has been interpolated from a
> smaller original image.
It was actually taken with a Sony TRV900 camcorder (probably a PAL
model), resolution pretty near 640x480 more or less. Interpolated
upwards with Altamira Genuine Fractals, printed, and scanned from the
print.
Chris Brown
>
>I think also with Velvia and Provia. I don't know about actual
>resolution, but to sample the number of dye blobs on the film, one would
>have to oversample by 2 at least. The analagous resolution in dye blobs
>was estimated at 13 megapixels I believe. What I do know is that a
>4000 dpi scan of Provia beats a 2800 dpi scan hands down. But what
>Bjorn has shown is that at a mere 6 megapixels, digital seems to beat
>film. Now where can I sell that F5...
This is the crux - when scanning negatives or slides you are sampling a
sample; creating a copy of a copy, as it were. Not only that, but the
technology you are using to make the second copy is rather different to the
technoplogy used to create the first one, and in order to faithfully
reproduce the image information, you need to faithfully capture the vaguries
of the film medium. This means that you have to sample at a much higher
resolution than would be needed if you were just sampling the image
directly, as digital cameras do.
Putting it another way, the digital capture process throws away certain
information. The film capture process also throws away certain types of
information, and it's not the same kind of information. In order to minimise
losses, the scanner needs to operate at a resolution which is overkill for
the amount of image detail.
--
/* _ */main(int k,char**n){char*i=k&1?"+L*;99,RU[,RUo+BeKAA+BECACJ+CAACA"
/* / ` */"CD+LBCACJ*":1[n],j,l=!k,m;do for(m=*i-48,j=l?m/k:m%k;m>>7?k=1<<m+
/* | */8,!l&&puts(&l)**&l:j--;printf(" \0_/"+l));while((l^=3)||l[++i]);
/* \_,hris Brown -- All opinions expressed are probably wrong. */return 0;}
Chris Brown
Randy Howard <ra...@thegateway.net> wrote:
>
>Then, your computer needs a new LARGE hard drive, a big monitor,
>a lot of memory and a very high quality printer and a ton of
>ink for it.
Only if you want to print large numbers of your own photos. Most prints from
my D30 are done at my local photo lab. Some of them I print myself at home,
and most of them I never print at all.
Chris Brown
Randy Howard <ra...@thegateway.net> wrote:
>Oh really? I've seen some EXTREMELY grainy looking digital photos.
>They have "grain" any time they are displayed above the native
>resolution of the CCD they were captured on.
That's not grain, it's sensor noise, whcih can look very similar.
There are many factors which affect the amount of noise present, but perhaps
the most important ones are:
Exposure time - longer exposures produce more noise. For long exposure work,
film is probably the better choice. I've had good results from a D30 at 30
seconds at 100 ISO, but beyond that is starting to push it.
Sensor size - smaller sensors are more prone to noise. Digital SLRs have
much physically larger sensors and are less prone to noise, which often
allows them to produce images which are better than digital point and shoot
cameras with nominally greater resolution.
YMMV.
Chris Brown
VT <vtVi...@prodigy.Net> wrote:
>Note there was _no_ fine detail in the subject to speak of, the hat 's
>detail is basically coarse and the pattern is high contrast but not
>exactly hard edge or fine detail - the wire fence in the background
>cannot be judged because they are basically out of focus images - that
>leaves the child's face and eyes - none of the features on the
>equivalent of 16x12' print would have been considered particularly
>sharp - I would say they were actually soft - that's not saying it
>wasn't a "nice photo"
So there we have it - a 0.3 megapixel image which Vincent, an arch sceptic
by all appearances ;-), says produces a 16*12 which although wasn't
"particularly sharp", was still probably a "nice photo".
Consumer digicams of the present sample at ten times the resolution of that
picture.
Dag
>
>>> What I do know is that a
>>> 4000 dpi scan of Provia beats a 2800 dpi scan hands down.
>
>
>>> seems to beat film.
>
>>the finest grain film available) may not stack up against the
>>D1X. Given that we don't have the original bricks to compare
>>with, it is difficult to be sure that artifacting isn't
>
>I think Bjorn is very careful and I'm beginning to think the end of film
>in the volume we know is sooner than we all thought. Damn, if they then
>come out with a 12 megapixel camera....
>
Film still has one BIG advantage over digital.
I can take my dad's old Pentax SLR, load it with the latest greatest
film and it will take just as good pictures as a brand new SLR. On the
other hand a 5 year old digital camera is obosolete and nothing can
bring it back to life.
The day someone starts selling digital cameras with exchangeable CCDs
I'll be the first in line. Untill then I'm really not sure that a high
end digital camera is a good long term investment.
Dag
Tony Polson
Please define what you mean by "a genuine rumor".
It sounds like an oxymoron to me.
;-)
--
Tony Polson
Steve
>At the amateur level, that sounds about right. At the professional
>level the D1x quickly pays for itself in film savings alone.
>> Oh really? I've seen some EXTREMELY grainy looking digital photos.
>> They have "grain" any time they are displayed above the native
>> resolution of the CCD they were captured on. They also show
>> something analogous to grain when color-depth limitations can
>> be clearly seen. Try a sunset picture on a digital camera, very
>> few of them, even with heavy post-processing look as smooth
>> in color variations as film.
>That's CCD noise, which is analagous to grain.
Which is well under control in the D1X apparently.
Steve
>How is a coarser grain film able to provide more resolution? Most
>people think it *looks* sharper, based upon the popped colors.
There is a loose correlation between resolution and grain. But the two
are not equal.
>> But if there are details that film just isn't catching, it doesn't
>> matter.
>> And digital is grainless - a very neat fact indeed.
>Oh really? I've seen some EXTREMELY grainy looking digital photos.
>They have "grain" any time they are displayed above the native
>resolution of the CCD they were captured on. They also show
>something analogous to grain when color-depth limitations can
>be clearly seen. Try a sunset picture on a digital camera, very
>few of them, even with heavy post-processing look as smooth
>in color variations as film.
Yes, you see something analogous to grain, but even the D1 did a neat
job of blue skys. In the traditional places where you see grain digital
looks good.
But I'm with you - not ready for digital yet.
Steve
Steve
>> But what Bjorn has shown is that at a mere
>> 6 megapixels, digital seems to beat film.
>It's hard for the human eye to see much more than six megapixels, which
>is why images with 20 megapixels don't look that much better.
Its the enlargement capabilities that matter most to me...
Steve
Steve
>> Correct, clearly in defiance of Mxsmanic' bogus
>> original claim.
>How is the original image photographed? Handheld? With what kind of
>lens?
>> Also, most production shots to be blown up,
>> say for a magazine cover are drum scanned ...
>Which is massive overkill, because no magazine cover can use either the
Not at all. If you've ever done it you'd know.
>dynamic range or the resolution of a drum scan. A 13x19" tabloid cover,
>for example, requires only about 2900x2000 pixels with a 150-line screen
>(and that is a very high screen frequency), and the dynamic range will
>be considerably less than 100:1.
If you scan at 150 lines for that spread, you will lose detail. You
still need to scan at a much higher bit depth to capture the details and
then screen them at 150. Analog to digital is not a 1:1 relationship.
If you were doing digital from the start, you could get away with that.
Steve
Steve
Steve
>>But what
>>Bjorn has shown is that at a mere 6 megapixels, digital seems to beat
>>film. Now where can I sell that F5...
>Toys R' Us?
I'm not embarassed to say I go there with my kids. However, you should
know better, they only carry EOS1Vs...
Steve
VT
<cpb...@ntlworld.no_uce_please.com> wrote:
>In article <3b25c145...@news.prodigy.net>,
>VT <vtVi...@prodigy.Net> wrote:
>>equivalent of 16x12' print would have been considered particularly
>>sharp - I would say they were actually soft - that's not saying it
>>wasn't a "nice photo"
>
>So there we have it - a 0.3 megapixel image which Vincent, an arch sceptic
>by all appearances ;-), says produces a 16*12 which although wasn't
>"particularly sharp", was still probably a "nice photo".
>
I will take that "standing count" -
and yes, I am very impressed with the fact that the original was a
640x480 image, from a camcorder no less. This does say a lot about
how well algorithms such as Genuine Fractals can work on even
apparently small amounts of information.
However my use of 'that's not saying it wasn't a "nice photo"' was
merely a comment about the composition etc - so that Paul did not feel
that I tearing the photo apart - having clarified that, I am not
backing away from Chris's interpretation either - please re-read my
opening paragraph.
>Consumer digicams of the present sample at ten times the resolution of that
>picture.
>
Just to keep things straight -
- current consumer 3Mp (2048 x 1536) digi-cams do have 10x the number
of pixels over a 0.3Mp -
- but they do not have 10x the resolution -
The increase in resolution is closer to 3.2x
So in perspective -
- 0.3Mp digi-cam has the (35mm film) equivalent of 9 to 10 lp/mm
resolution at the film plane
- 3Mp digi-cam has the (35mm film) equivalent of 28 to 32 lp/mm at
the film plane.
As you see a 3Mp digi-cam is getting close to a film camera for
resolution -
but unforunately that's just not a particularly good film camera -
BUT upsampling/interpolation algorithms can give much bigger
enlargements/ prints than would be first perceived using resolution
figures alone - however that is at a cost of image softness verses the
other image degrading effects when enlarging film beyond its
acceptable resolution limits.
(I do realize image "quality" is much, much more than just resolution)
Please bear in mind that film can and is often scanned (using
professional drum scanners of 5000dpi) therefore all the
upsampling/interpolation algorithms (such as Genuine Fractals) can
also be applied to these scanned film images -- to result in even
bigger enlargements - or enlargements from much smaller cropped
sections, than would be possible using the traditional analog/light
enlargers.
So yes, digital is here - but it applies much more in post capture
processing - as for image capture it is close but no cigar yet - but
as always YMMV
--
Vincent
vtVi...@Prodigy.Net
CJMorgan59
>With all due respect (and I do think the Nikon D1 is a very fine
>digital SLR) printing a Nikon D1 image - which is all of 2012x1324
>pixels to 20" across is using in about 100ppi.
>
>At 100ppi an image is very likely to be pixellized on anything
>resembling any detail - or if upsampled/interpolated will be pretty
>soft -
>
>Since 100ppi is the equivalent of about 2lp/mm on paper -
>
>This is 1/4 what Leica considers quality (ie: 8lp/mm) and even for
>most mortals this is still 1/3 the requirement of the generally
>accepted 6lp/mm on paper. 4lp/mm on paper is pretty marginal but
>acceptable on larger sized prints that do not require close scrutiny -
>even then the Nikon D1 can only deliver 1/2 that resolution.
>
>I think that the Nikon D1 may have replaced film for some - but for
>many this kind of resolution may be less than adequate.
What's often overlooked in such discussions is that when we go from film to
print, there's a loss of information (from one generation to the next) which
doesn't occur the same way when working with a direct digital file. While I
don't at all challenge the numbers you present, seeing is nevertheless still
believing.
I've only worked with one digital camera so far, and this was 3.3 megapixel.
Images from this, worked in Photoshop and then printed out to 12x18, were quite
impressive. I imagine with a Nikon D1 the result would even be superior to
this, and enough so that were you to look at a final image, I don't believe you
couldn't say with any certainty whether it came from a digital image or 35mm
film.
That's the difference between shooting targets where we're counting lines and
actually making visual images; the difference between theory and practice; the
difference between spending our life with a magnifying glass and making
photographs which we use.
CJ
Randy Howard
says...
> The day someone starts selling digital cameras with exchangeable CCDs
> I'll be the first in line. Untill then I'm really not sure that a high
> end digital camera is a good long term investment.
Well, if Nikon would make a digital back for the F5, they could
upgrade it over the years, and still stick with the same body.
I suspect that the back would cost about as much as most camera
bodies though.
VT
>That's the difference between shooting targets where we're counting lines and
>actually making visual images; the difference between theory and practice; the
>difference between spending our life with a magnifying glass and making
>photographs which we use.
>
I understand what you are saying -
but perhaps you are missing what was trying to be said here......
OK - if I said such and such image was really high quality -
why would or _should_ anyone believe me? -
Unfortunately we are all just using words to describe a visual
experience - and anyone's say-so is just that -
hearsay.
It is useful to be able to quantify things to explain why certain
things work certain ways - even 35mm film images have some limitation
to which they can be enlarged to - even if we scan them then use the
digital domain to process and print them, so that basically they can
enjoy the loss-free environment of "pure digital" - so what is that
limit?
Well one can usually work this out from facts and figures (numerous
books and article have been written, and college level courses exist
all about this - and the accumulated years of actual practical
expeience from the accredited authors/lecturers probably can be
measured in person-centuries) -
Photography is both a science and an art - to deny one is
delliberately not wanting to see the complete picture (excuse the pun
:-)
Perhaps there is "NO limit" since one may tolerate much higher degree
of softness - or the subject simply does not demand the high
definition. But that does nothing to invalidate figures that apply in
the majority of cases, neither does it deny the simple fact there will
always be exceptions and the fact YMMV.
So if someone says it's not sharp - how does one overcome the
hearsay/opinion?
Perhaps some facts and supporting figures might go some way to help
explain why?.....
Just because one uses figures to try to explain what one expects to be
able to see is no indication of their life-style or experience -
just as it would be inappropriate/impertinent for me to comment about
any poster who expresses an opinion, substainated or not..
--
Vincent
vtVi...@Prodigy.Net
Randy Howard
cpb...@ntlworld.no_uce_please.com says...
> In article <MPG.158f72961...@news.thegateway.net>,
> Randy Howard <ra...@thegateway.net> wrote:
> >
> >Then, your computer needs a new LARGE hard drive, a big monitor,
> >a lot of memory and a very high quality printer and a ton of
> >ink for it.
>
> Only if you want to print large numbers of your own photos. Most prints from
> my D30 are done at my local photo lab. Some of them I print myself at home,
> and most of them I never print at all.
This response applies only to the printer and ink. All the rest,
including the points you did not include in your reply from the
original post still apply.
Joseph S. Wisniewski
>
> Q: Is that a genuine rumor, that the D2 will have 24x36mm CMOS? Or
> just a guess?
>
> A: That's my WISH. I strongly suspect that it will be full frame 35mm but I
> don't think they'll be ready with CMOS yet. Will probably be CCD. CMOS takes
> significantly less power to operate, and that's the other big issue with the
> current generation of digital cameras,--power consumption.
Biggest power consumption is still the display (and most of that is
backlight). The sensor is a fairly small part (except if there's a
Peltier cooler. The DSP is a bit more of a drain.
Now, if they hooded the display, and maybe added some optics and an
eyepiece, so you could view it always at the optimum angle, and not need
much backlight power, the end result would be like a high eyepoint SLR
prism, but electronic.
DOF preview, without the display going dim at small apertures. A B&W,
IR, or UV finder, as needed.
The user interface is definitely the next thing that needs to be
revamped in digital cameras.
My wish is an easily removable IR filter.
Ciao!
Joe
David Chien
play, all the time.
www.silverace.com/dottyspotty/
But, for a picky person who wants both quality image and resolution of fine details,
we've still got a ways to go to at least 8MP before I'll be happy, but certainly
there are/can also be lots of happy 5MP D1x users today who don't need any more.
In any case, for the rest of us with MP envy, there's the 4.1MP Sony S85 for $799
that'll get us between a D1 and D1x for the meantime.
d =)
ccf...@attglobal.net
Chih Fang
"VT" <vtVi...@prodigy.Net> 撰寫於郵件
news:3b25c145...@news.prodigy.net...
Paul Rubin
> The image looks fine - but with what I would consider overall softness
> - and on close scrutiny especially at edge between the background and
> the right side of the child's face there are some obvious edge
> artifacts (which could be jpeg artifacts or sharpening by the camera)
> - remember this would have been a large - 16x12" print from the
> original 1600x1200 image - that is 100ppi or about 2lp/mm printed on paper.
Still, the point of the example was to show how useable images don't
need as much resolution as most people think. As mentioned in other
posts, the original of that image was taken with a Sony camcorder and
had only about 0.3 megapixels. It was scaled up to high resolution
with Altamira Genuine Fractals, printed on a printer, and then scanned
back in with a scanner. The results aren't exactly in the D1x class,
but I'd have expected them to look like complete crap, and they don't.
Mxsmanic
> Still, the point of the example was to show how
> useable images don't need as much resolution as
> most people think.
This is certainly true, and in fact, even among the most famous
photographs in the world, very few indeed show particularly high
resolution. That reality is rarely acknowledged in this newsgroup,
however.
Mxsmanic
news:slrn9ibtj...@Omally.animagicnet.no...
> The day someone starts selling digital cameras with
> exchangeable CCDs I'll be the first in line. Untill
> then I'm really not sure that a high end digital
> camera is a good long term investment.
I tend to agree. One of the great dangers of digital photography--one
that few users have noticed thus far--is that it may lead to a "upgrade
every six months" syndrome similar to the one that has plagued the
personal-computer industry for the past two decades. The only way to
get a better picture will be to buy a new camera body (and the old one
will be worthless, so you won't be able to resell it). This prospect
makes digital-camera manufacturers drool, but it should be scaring
consumers. Unfortunately, many of them have not been down this path
with computers, and so they don't see it coming.
I want interchangeable sensors measuring 24x36 mm, just like film,
otherwise digital will not have completely arrived, as far as I'm
concerned.
BTW, Nikon had increased sales of 35mm film equipment last year,
although none of the other major players did, so this may have been an
anomaly.
Mxsmanic
news:MPG.15904715a...@news.thegateway.net...
> Well, if Nikon would make a digital back for
> the F5, they could upgrade it over the years,
> and still stick with the same body.
Why would they do that, if they can force photographers to buy a whole
new body instead? That's the idea that is currently making digicam
manufacturers salivate. Imagine selling F5-equivalents to photographers
_every year_, instead of every _twenty_ years! It would provide the
same cash cow that PCs have provided for years. To the consumer's
detriment, of course ... but who cares about the consumers?
Mxsmanic
news:9g55ki$9...@nntpb.cb.lucent.com...
> Its the enlargement capabilities that matter
> most to me...
Enlargement has no effect. Only relative viewing distance matters. A
4x6-foot poster print viewed at six feet need have no greater resolution
than a 4x6-inch snapshot viewed at six inches.
Mxsmanic
> Not at all. If you've ever done it you'd know.
I have, and it is.
> If you scan at 150 lines for that spread,
> you will lose detail.
You will lose detail when you go to press, so it doesn't really matter.
Jeff Cook
> > most to me...
>
> Enlargement has no effect. Only relative viewing distance matters. A
> 4x6-foot poster print viewed at six feet need have no greater resolution
> than a 4x6-inch snapshot viewed at six inches.
Of course it has effect. Size is not the only thing involved; pixel
interpolation and similar factors that go into how and when you enlarge
something have profound effect on any sizing issues.
--
Jeff Cook
je...@cookstudios.com
http://www.cookstudios.com
Washington DC area
Mxsmanic
news:3B27261B...@cookstudios.com...
> Of course it has effect.
No, it doesn't. Resolution is a matter of the actual angle subtended by
a detail in the visual field. As long as that remains constant,
resolution remains constant. If the viewing distance changes in
proportion to changes in the image size, the size of details in the
visual field remains constant.
> Size is not the only thing involved; pixel
> interpolation and similar factors that go
> into how and when you enlarge something have
> profound effect on any sizing issues.
Pixels need not be interpolated for enlargements, and even if they are,
unless the interpolation is very clumsy, it should not diminish visible
resolution.
Steve
>> Well, if Nikon would make a digital back for
>> the F5, they could upgrade it over the years,
>> and still stick with the same body.
>Why would they do that, if they can force photographers to buy a whole
>new body instead? That's the idea that is currently making digicam
>manufacturers salivate. Imagine selling F5-equivalents to photographers
>_every year_, instead of every _twenty_ years! It would provide the
>same cash cow that PCs have provided for years. To the consumer's
>detriment, of course ... but who cares about the consumers?
Its quite nausiating actually. Come out with a D1 and a few months
later come out with a D1x. Next the D2....its way out of line in my
opinion. They need to modularize this thing so you can upgrade your
body for maybe $1000 or less.
Steve
Steve
>> >Then, your computer needs a new LARGE hard drive, a big monitor,
>> >a lot of memory and a very high quality printer and a ton of
>> >ink for it.
>> Only if you want to print large numbers of your own photos. Most prints from
>> my D30 are done at my local photo lab. Some of them I print myself at home,
>> and most of them I never print at all.
>This response applies only to the printer and ink. All the rest,
>including the points you did not include in your reply from the
>original post still apply.
But if you scan your images and do your own digital retouching and
inkjet printing, you've already spent that money. High quality digital
for the original image is just the elimination of several lossy steps.
Steve
Steve
>> Its the enlargement capabilities that matter
>> most to me...
>Enlargement has no effect. Only relative viewing distance matters. A
>4x6-foot poster print viewed at six feet need have no greater resolution
>than a 4x6-inch snapshot viewed at six inches.
Enlargement does have an effect. For some reason you want a 4 x 6 foot
poster compared to a 4 x 6 inch snapshot. But within a certain range,
viewing distance is about the same for most folks. Many of us do get
right up on an 8 x 10 or 11 x 17. And the 4x6 inch snapshot can't even
contain the detail that an 8 x 10 enlargement can. So for me, holding
detail out to 8 x 10 is essential, out to 11x17 preferred.
Steve
Steve
>> Of course it has effect.
>>No, it doesn't. Resolution is a matter of the actual angle subtended
>resolution remains constant. If the viewing distance changes in
>proportion to changes in the image size, the size of details in the
>visual field remains constant.
If those details are acutally there, and if you actually force your
viewing audience to stay at the proper distance. I find this argument
to be quite useless. When a fine art photographer displays his stuff,
I don't see any tape or rope keeping the viewers at the proper distance!
Hell, most get as close as they can to marvel at the luscious detail in
the print. Personally, I get as close as I can to any picture, and also
view it from a distance. If you aren't holding the detail, you aren't
holding the detail.
Everybody stand back! Please view the prints at their proper distance!
We don't want you to know that we enlarged them 1200% !!
>> Size is not the only thing involved; pixel
>> interpolation and similar factors that go
>> into how and when you enlarge something have
>> profound effect on any sizing issues.
>Pixels need not be interpolated for enlargements, and even if they are,
>unless the interpolation is very clumsy, it should not diminish visible
>resolution.
They have to be if your above your original image size. And I don't
usually like the look at that point.
Steve
Steve
>> Not at all. If you've ever done it you'd know.
>I have, and it is.
>> If you scan at 150 lines for that spread,
>> you will lose detail.
>You will lose detail when you go to press, so it doesn't really matter.
Sure does. The fact that you do lose it is more reasont to contain as
much as possible when going to press. This applied back in the old
copy camera and negative days. Its not like you lose everything you
gained by oversampling when you go to press. If you're doing large
spreads like that, scanning at 4000 dpi makes a huge difference in the
final output.
Steve
Mxsmanic
> They need to modularize this thing so you can
> upgrade your body for maybe $1000 or less.
I bet they won't. It'll be the whole PC story over again. Already
digital cameras are selling better than PCs; I believe digicams are the
number-one consumer electronics item right now in terms of growth.
VT
wrote:
> in fact, even among the most famous
>photographs in the world, very few indeed show particularly high
>resolution. That reality is rarely acknowledged in this newsgroup,
>however.
>
I'll acknowledge this in spades -
For example Robert Capa's famous photo of the landing at Omaha Beach
with the trooper lying in the water -
http://www.rpn.ch/csc/branches/hist-sec/images/rc04.jpg
shows a large degree of movement/blur and almost no sharpness to speak
of - yet this is one of the all time great photographs (and one of my
favorites)
However on the other hand many of Ansel Adams' photographs have
definition that is probably still unattainable even by the best of the
modern 35mm systems and films (and current consumer digital isn't even
a consideration).
As people will say it's the subject/composition and the message the
photo is trying to convey that matters.
Sometimes resolution/sharpness matters - other times it can actually
be a distraction.
But let us ask this simple question - given a notable scene - possibly
a once in a lifetime shot - if you had the choice -
would you rather take the photo on a well known brand 35mm
_disposable_ camera
or something like a Leica, Nikon or Canon?
[I know the argument that any camera is better than nothing to take
that once in a lifetime shot - but that is _not_ the question being
asked.....]
--
Vincent
vtVi...@Prodigy.Net
Steve
That's what makes me angry. However, its conceivable that a
professional level camera could be different. But they aren't doing it
with $30,000 scan backs and one-shot backs. Why would they do it with
Nikon D series cameras?
But by the same token, if you finally get your 8 MP camera, do you need
more?
Steve
Chris Brown
>
>Its quite nausiating actually. Come out with a D1 and a few months
>later come out with a D1x. Next the D2....its way out of line in my
>opinion. They need to modularize this thing so you can upgrade your
>body for maybe $1000 or less.
It's shocking really. I'm glad that Canon have not yet announced their new
digital EOS because I'm sure that the moment they do, my D30 will sieze up
on the spot and never be capable of taking another photo.
--
/* _ */main(int k,char**n){char*i=k&1?"+L*;99,RU[,RUo+BeKAA+BECACJ+CAACA"
/* / ` */"CD+LBCACJ*":1[n],j,l=!k,m;do for(m=*i-48,j=l?m/k:m%k;m>>7?k=1<<m+
/* | */8,!l&&puts(&l)**&l:j--;printf(" \0_/"+l));while((l^=3)||l[++i]);
/* \_,hris Brown -- All opinions expressed are probably wrong. */return 0;}
Mxsmanic
news:3b277ecd...@news.prodigy.net...
> However on the other hand many of Ansel Adams'
> photographs have definition that is probably
> still unattainable even by the best of the
> modern 35mm systems and films ...
Ansel Adams' entire body of work is _based_ on technical perfection.
That's his claim to fame. If it weren't for that, most of his work
would just be another set of standard landscape photographs. You marvel
at the technical tour de force in each of his photos, and far less at
the composition or content of the image.
Look at HCB, and you see just the opposite. A lot of his work is badly
focused and exposed. But you marvel at the content and composition of
his images, and not at the technical quality (which is often mediocre).
If technical image quality were all that anyone cared about, painting
would have disappeared within a few years of the invention of
photography, because photography is a far more accurate representation
of the real world than any painter can achieve, and so anyone who wants
technical quality above all else would simply discard painting without a
second thought.
> But let us ask this simple question - given a notable
> scene - possibly a once in a lifetime shot - if you
> had the choice - would you rather take the photo on a
> well known brand 35mm _disposable_ camera or something
> like a Leica, Nikon or Canon?
I'd rather use the latter, on a tripod in perfect natural light and in
great weather with exactly the appropriate film. But if it's a
once-in-a-lifetime shot, having the perfect equipment is not going to
improve the shot very much over just having it in the first place.
Nobody turned down that tourist's shot of a Concorde going down in
flames, and I think he used a disposable.
Mxsmanic
news:6s18g9...@gecko.localdomain...
> I'm glad that Canon have not yet announced their
> new digital EOS because I'm sure that the moment
> they do, my D30 will sieze up on the spot and
> never be capable of taking another photo.
If it's going to seize up, it better do so now, because you haven't a
ghost of a chance of getting it serviced two years from now.
Then again, there probably isn't much that is servicable inside; the
cheapest option is to replace it. Of course, if you have to replace it,
and it's no longer on the market, you are forced to upgrade--just as
things have traditionally been in the PC world. If you think that's a
good thing, wait and see, and you'll change your mind.
Mxsmanic
> But they aren't doing it with $30,000 scan
> backs and one-shot backs. Why would they
> do it with Nikon D series cameras?
Because 35mm SLRs have never had interchangeable backs.
> But by the same token, if you finally get
> your 8 MP camera, do you need more?
Not more, but different. What if I want to shoot infrared? What if I
want ISO 3200, instead of ISO 100? What if I want the look of Tri-X?
What if I want Tech Pan?
VT
wrote:
>"VT" <vtVi...@prodigy.Net> wrote in message
>news:3b277ecd...@news.prodigy.net...
>
>> However on the other hand many of Ansel Adams'
>> photographs have definition that is probably
>> still unattainable even by the best of the
>> modern 35mm systems and films ...
>
>Ansel Adams' entire body of work is _based_ on technical perfection.
>That's his claim to fame. If it weren't for that, most of his work
>would just be another set of standard landscape photographs. You marvel
>at the technical tour de force in each of his photos, and far less at
>the composition or content of the image.
Sorry I have to respectfully disagree with that - it is not just the
technical quality - I do believe it has a lot more to do with his
artistic eye -
- just to use your own kind of argument/analogy -
Otherwise any person with good technical ability would be able to
produce a masterpiece from a mundane landscape -
with today's technically excellent equipment there would be a lot of
"Ansel Adams" out there
- there aren't there is/was only one -
I think what you are trying to say in defence of your previous post -
insults the art of Ansel Adams, regardless of one's personal taste -
some respect is due to an acknowledged master of photography -
that respect should not be at a cost of merely winning some argument
on the internet......
>
>Look at HCB, and you see just the opposite. A lot of his work is badly
>focused and exposed. But you marvel at the content and composition of
>his images, and not at the technical quality (which is often mediocre).
>
Again an exaggeration and rather disrespectful to a master of
photography -
Henri Cartier-Bresson - was hardly anywhere inept at photography - he
took many shots of "decisive moments" (in fact, that term was probably
invented for him) - sometimes there just was not time to do all the
things necessary to get a sharp well exposed shot - but I think most
of the time static and sharp would have been unsuitable for the
subject/composition/mood.
Almost any barely competent photographer knows how to set their lens
at the hyperfocal distance and pre-set the shutter speed & aperture to
the prevailing light conditions to get a pretty well exposed and sharp
shot -
I think Henri Cartier-Bresson knew what he was doing and did things
deliberately for the mood/expression of the photograph - very few I
would think were accidents -
that's why they are the acknowledged masterpieces that they are.
>If technical image quality were all that anyone cared about, painting
>would have disappeared within a few years of the invention of
>photography, because photography is a far more accurate representation
>of the real world than any painter can achieve, and so anyone who wants
>technical quality above all else would simply discard painting without a
>second thought.
>
That way too naive an outlook -
as you have pointed out there are many photographs that are not tack
sharp - but neither are they lifeless or static - the acknowledged
masterpieces are so because they convey the moment and artistic
expression.
But using competent and equipment of good technical quality is not the
incompatible with artistic expression and taking good photographs -
Most of the great photographers used top equipment of their days -
they just did not use technical quality for technical quality's sake -
and most of us already know this here - it's not the equipment but the
person behind it.
>> But let us ask this simple question - given a notable
>> scene - possibly a once in a lifetime shot - if you
>> had the choice - would you rather take the photo on a
>> well known brand 35mm _disposable_ camera or something
>> like a Leica, Nikon or Canon?
>
>I'd rather use the latter, on a tripod in perfect natural light and in
>great weather with exactly the appropriate film.
Thank you for answering the question.
>But if it's a
>once-in-a-lifetime shot, having the perfect equipment is not going to
>improve the shot very much over just having it in the first place.
Then why wouldn't you use a simple disposable, since the choice
appears to be inmaterial to you??
> On Wed, 13 Jun 2001 15:27:14 GMT, vtVi...@prodigy.Net (VT) wrote:
>>[I know the argument that any camera is better than nothing to take
>>that once in a lifetime shot - but that is _not_ the question being
>>asked.....]
>Nobody turned down that tourist's shot of a Concorde going down in
>flames, and I think he used a disposable.
>
Thank you also for answering the question that was not being
asked...... :-)
--
Vincent
vtVi...@Prodigy.Net
VT
wrote:
>
>Because 35mm SLRs have never had interchangeable backs.
>
Huh?
Yes, they have......
System SLRs can interchange backs to -
- bulk film backs
- data backs.
- multi-control backs
- NPC Polaroid backs
Just as an example there are 3 pages of listed interchangable backs
for 35mm SLRs currently sold by B&H - this is not even counting past
SLR models -
http://www02.bhphotovideo.com/default.sph/FrameWork.class?FNC=CatalogActivator__Acatalog_html___CatID=285___SID=E7497671DF0
--
Vincent
vtVi...@Prodigy.Net
Mxsmanic
> Almost any barely competent photographer knows
> how to set their lens at the hyperfocal distance
> ...
And a competent photographer knows also that even with the lens set to
the hyperfocal distance, the only part of the image that is actually
sharp is the part upon which the lens is focused. Depth of field works
only for a standard viewing distance. I'm surprised that with so much
talk of resolution being required beyond that necessary for a standard
viewing distance, there exists simultaneously such a widespread belief
in sharpness within the depth of field.
> Then why wouldn't you use a simple disposable,
> since the choice appears to be inmaterial to you??
I would use whatever might be available at the time.
Bud Cook
http://www.free-photons.de/contarex.htm
--
Bud Cook
Remove X to Reply
"VT" <vtVi...@prodigy.Net> wrote in message
EDGY01
Just sent you a file (7.5MB). Remember that the printer you use makes a
significant difference. That printer used costs in the tens of thousands of
dollars.
Dan Lindsay
Santa Barbara
VT
wrote:
>
>And a competent photographer knows also that even with the lens set to
>the hyperfocal distance, the only part of the image that is actually
>sharp is the part upon which the lens is focused. Depth of field works
>only for a standard viewing distance. I'm surprised that with so much
>talk of resolution being required beyond that necessary for a standard
>viewing distance, there exists simultaneously such a widespread belief
>in sharpness within the depth of field.
It is true that there is only one true focussed point - and at that
distance the image is the "sharpest".
Pardon the pun, but this shows pretty shallow understanding of depth
of field.
Please do ask any competent street photographer how zone focussing and
depth of field works for them.
Standard Depth of field tables are not the only way to use DoF -
you can re-calculate depth of field to any acceptable resolution you
want, and work to those standards.
Standard DoF tables usually use a limiting circle of confusion of
0.001" or 0.025mm - this about 20 lp/mm - which for most practical
usage is pretty sharp -
One can easily use a higher resolution of say 40lp/mm ie: half the
size for the limiting circle of confusion and sharp, one merely has to
re-calulated the new hyperfocal distance for this required resolution
You can even use 80lp/mm if you wish, which many lenses can't even
achieve at their focussed point. - you'd be surpised how useful the
hyperfocal distance remains, even with this ridiculous
demand..........
--
Vincent
vtVi...@Prodigy.Net
Steve
>> But they aren't doing it with $30,000 scan
>> backs and one-shot backs. Why would they
>> do it with Nikon D series cameras?
>Because 35mm SLRs have never had interchangeable backs.
>> But by the same token, if you finally get
>> your 8 MP camera, do you need more?
>Not more, but different. What if I want to shoot infrared? What if I
>want ISO 3200, instead of ISO 100? What if I want the look of Tri-X?
>What if I want Tech Pan?
Tech Pan - use Tech Pan. ISO 3200 - when do you use that? Tri-X -
simulate it in Pshop.
Steve