Film cranks
RichA
what they say is just wishful or wrong. The idea promoted by some
that film has "infinite resolution" and the results can be varied
based on increasing scan resolution is nonsense. At best, the
sharpest, finest grain 35mm film can resolve as much as a 6-8
megapixel digital. Film grains, their clumpiness, and the spaces
between them ARE the "pixels" of film and they do not afford infinite
resolution. The highest resolving power film I've ever seen was Tech
Pan (discontinued). It is possible that with the very best lenses,
modern scanning, 35mm Tech Pan could approach the resolution of a
10-12 megapixel digital, but I really doubt it.
Peter Irwin
> I've read a lot of the musings from film buffs. Problem is, most of
> what they say is just wishful or wrong. The idea promoted by some
> that film has "infinite resolution" and the results can be varied
> based on increasing scan resolution is nonsense.
I haven't seen any such thing. I've seen some wishful thinking
based on Kodak figures of 24 Mpix per 24x36mm frame, but I've never
seen anyone claim infinite resolution.
> At best, the
> sharpest, finest grain 35mm film can resolve as much as a 6-8
> megapixel digital.
It ain't quite that simple. Film has an MTF curve which falls
off somewhat gradually. How much resolution a film has depends
on the response level you choose for the cutoff, and there isn't
really a hard and fast rule.
> Film grains, their clumpiness, and the spaces
> between them ARE the "pixels" of film and they do not afford infinite
> resolution.
Actually, I think light scattering in the emulsion is usually
the biggest limiting factor in film resolution.
> The highest resolving power film I've ever seen was Tech
> Pan (discontinued).
There exist films with higher resolution than Tech Pan, but they
all very slow and not general purpose films.
Peter.
--
pir...@ktb.net
Rich
> RichA <rande...@gmail.com> wrote:
>> I've read a lot of the musings from film buffs. Problem is, most of
>> what they say is just wishful or wrong. The idea promoted by some
>> that film has "infinite resolution" and the results can be varied
>> based on increasing scan resolution is nonsense.
>
> I haven't seen any such thing. I've seen some wishful thinking
> based on Kodak figures of 24 Mpix per 24x36mm frame, but I've never
> seen anyone claim infinite resolution.
I know the argument. They scan it at 48 bits and end up with huge files
which they claim = more "information." Same argument was used by VHS
fans because a VHS tape would hold something like 160G worth of
information. Problem is, it wasn't high resolution. In film's case,
increased graininess actually results in larger file sizes when scanned,
but it doesn't mean higher resolution by any stretch.
>
>> At best, the
>> sharpest, finest grain 35mm film can resolve as much as a 6-8
>> megapixel digital.
>
> It ain't quite that simple. Film has an MTF curve which falls
> off somewhat gradually. How much resolution a film has depends
> on the response level you choose for the cutoff, and there isn't
> really a hard and fast rule.
>
>> Film grains, their clumpiness, and the spaces
>> between them ARE the "pixels" of film and they do not afford infinite
>> resolution.
>
> Actually, I think light scattering in the emulsion is usually
> the biggest limiting factor in film resolution.
>
>> The highest resolving power film I've ever seen was Tech
>> Pan (discontinued).
>
> There exist films with higher resolution than Tech Pan, but they
> all very slow and not general purpose films.
>
> Peter.
I remember Ilford back in the late 1980s had some kind of colour film (6-
12 ASA) that had Tech Pan resolution, but it was never released to the
public.
Peter Irwin
>
> I remember Ilford back in the late 1980s had some kind of colour film (6-
> 12 ASA) that had Tech Pan resolution, but it was never released to the
> public.
You can buy Ilfochrome Micrographic. It is still being made. The
Swiss Ilford company that makes Ilfochrome makes it - not the UK
concern that makes B&W film. It is slower than 1 ASA and tungsten
balanced. Resolution is at least on par with Tech Pan and it has
by far the best dye life of any colour film. I think it costs
more than 100 dollars per 100 foot roll, so it is fairly expensive.
It uses Ilfochrome chemicals which are slightly different from
the ones for paper.
Peter.
--
pir...@ktb.net
Doug Jewell
> I've read a lot of the musings from film buffs. Problem is, most of
> what they say is just wishful or wrong. The idea promoted by some
> that film has "infinite resolution" and the results can be varied
infinite resolution.
> based on increasing scan resolution is nonsense. At best, the
what most people scan it. Consumer grade scanners (and that
includes things like the Nikon and Minolta scanners with
their "4000dpi") don't come close to getting the detail that
is present in film. Most of these scanners are limited by
their own inability to accurately focus on the film. I've
seen 7MP frontier scans that get better detail than 4000dpi
(approx 20MP) Minolta scans. A proper drum scan will go
better again.
> sharpest, finest grain 35mm film can resolve as much as a 6-8
> megapixel digital.
400) has a resolving power that is about on par with 8MP
when properly scanned. Lower ISO, and generally better
quality films are significantly better. My estimate for
Velvia 50 is somewhere in the vicinity of 15-20MP.
>Film grains, their clumpiness, and the spaces
> between them ARE the "pixels" of film and they do not afford infinite
> resolution.
True that they are a long way from "infinite" resolution.
But at the same time, you can't directly compare them to
pixels - the variable size of grain, and randomness of it
means that it is better at recording irregular shapes - eg
diagonal lines. Film's diagonal resolution is the same as
it's horizontal & vertical resolution, whereas digital's
diagonal resolution is only 1/SQRT(2) of it's
horizontal/vertial resolution. Since few things in reality
are perfectly aligned vertically or horizontally, this
immediately gives film an advantage.
>
The highest resolving power film I've ever seen was Tech
> Pan (discontinued). It is possible that with the very best lenses,
> modern scanning, 35mm Tech Pan could approach the resolution of a
> 10-12 megapixel digital, but I really doubt it.
I don't know how you are doing wrong, but techpan is capable
of 200 lp/mm (admittedly at reduced contrast) - which would
require 138MP to reproduce in digital or 200MP to reproduce
on a diagonal. Now I'm not saying that a 35mm frame of
techpan is equal to 200MP - since no lens can do that, but
it is definately able to record far more than 10-12MP.
>
> http://www.kenrockwell.com/tech/real-raw.htm
Much of rockwell's stuff is crap, but he does make some good
points about being able to rescan film to get a better
quality image.
bugbear
> True that they are a long way from "infinite" resolution. But at the
> same time, you can't directly compare them to pixels - the variable size
> of grain, and randomness of it means that it is better at recording
> irregular shapes - eg diagonal lines. Film's diagonal resolution is the
> same as it's horizontal & vertical resolution, whereas digital's
> diagonal resolution is only 1/SQRT(2) of it's horizontal/vertial
> resolution. Since few things in reality are perfectly aligned vertically
> or horizontally, this immediately gives film an advantage.
Of 40% at best, on your numbers.
BugBear
Daniel
news:b731abf3-c527-45a9...@l39g2000yqn.googlegroups.com...
This old chestnut. Again! ;-)
Comparing scanned 35mm film to a 35mm DSLR seems a bit of a non entity in
reality. If you scan your own film using a desktop film scanner, although
the image maybe 24MP, comparing it to a 24MP image straight out of a DSLR is
night and day image quality wise, let alone the time involved. A 24MP scan
is still OK and I love looking at old street shots where you see things in
the images you have never noticed before (like dodgy looking geezers in the
sidelines watching your every move!).
Some may argue the fact that drum scans are the mutts nuts, but lets face
it, unless you're selling your photos the cost of a drum scan of even a
single frame is bloody expensive. I have yet to find a bulk scanning
service that is cost effective.
So, even if a best quality neg/slide, scanned on a drum scanner was on par
to say a 5DII, it's still a non entity in real life because of the extra
time and cost involved. Even then, you still loose the ability to adjust
the film speed on-the-fly, you still need to change the film every 36
exposures and the when you get into higher ISO's, well the gap gets even
greater image quality wise. Also, scanned images even at 16 bits/channel
don't respond to post processing as well as a RAW image from a DSLR.
What I find strange is that film scanners never really seem to evolve much.
I'm also surprised that the likes of India haven't picked up on the film
scanning industry with their lower labour costs, afterall, they must be
scanning motion picture film already. Maybe I've just not come across them.
Bruce
>
>What I find strange is that film scanners never really seem to evolve much.
No need.
When you can already scan at, say, 4000 ppi with every pixel position
being scanned for red, green and blue, you have a 21.4 MP image* that
would need 64 million pixels to reproduce with digital RGB, or all of
86 million pixels in the almost universal Bayer pattern (RGGB).
Camera and sensor manufacturers (such as Sony) are deliberately being
dishonest when they claim their cameras have 24 megapixel sensors when
in fact they have 6 million pixel sensors with four (RGGB) receptors
contributing to each multicolour pixel.
In contrast, flat screen TV manufacturers (such as Sony) are being
scrupulously honest when they state that their 1920 x 1080P HDTV panels
provide a 2.07 MP image from 6.21 million photodiodes.
I have nothing against Sony here. All the digital camera and sensor
manufacturers are guilty of the same wilful deceit. :-(
* based on 24 x 36mm at 4000 pixels per inch.
nospam
<n...@nospam.net> wrote:
> When you can already scan at, say, 4000 ppi with every pixel position
> being scanned for red, green and blue, you have a 21.4 MP image* that
> would need 64 million pixels to reproduce with digital RGB, or all of
> 86 million pixels in the almost universal Bayer pattern (RGGB).
not even remotely close to being correct.
> Camera and sensor manufacturers (such as Sony) are deliberately being
> dishonest when they claim their cameras have 24 megapixel sensors when
> in fact they have 6 million pixel sensors with four (RGGB) receptors
> contributing to each multicolour pixel.
that's not how bayer works.
whisky-dave
"Rich" <no...@nowhere.com> wrote in message
news:nbSdnXjLbqs-fAHU...@giganews.com...
>
> I remember Ilford back in the late 1980s had some kind of colour film (6-
> 12 ASA) that had Tech Pan resolution, but it was never released to the
> public.
Any ides what sort of 'film' is/was usder for micro-fiche (SP)
you know the sort that was used in libraries around that time for newspaper
scanning etc...
Peter Irwin
>
> "Rich" <no...@nowhere.com> wrote in message
> news:nbSdnXjLbqs-fAHU...@giganews.com...
>
>> I remember Ilford back in the late 1980s had some kind of colour film (6-
>> 12 ASA) that had Tech Pan resolution, but it was never released to the
>> public.
>
> Any ides what sort of 'film' is/was usder for micro-fiche (SP)
>
More like under 1 ASA, but definitely still available.
Peter.
--
pir...@ktb.net
spam...@ncf.ca
"whisky-dave" <whisk...@final.front.ear> wrote in message
news:gnjjar$al8$1@qmul...
Fuji also makes one, various companys like H&W Control Pan repackaged it
with a developer very similar to Technidol that made it practical for
pictorial uses. Today you can buy a version of this as Bluefire Police (made
by Adox)
http://www.frugalphotographer.com
http://www.adox.net/Products.htm
When Kodak Tech Pan became popular, Leica in the magazine claimed even their
lenses could not resolve to that film capabilities.
Don Stauffer
I tend to agree with much that you say, but there ARE grainless films,
such as those used in microfilm and some technical process photography.
There are folks who use these in regular photogaphy by altering
developers and development. For instance, folks who use Kodalith in
normal contrast range work.
Even the Tech Pan has ways of considerably increasing res by magic soups.
Still, that was extreme stuff and a labor of love. I tried it a few
times (my wife stuck with it longer). We both have gone to digital :-)
tmo...@wildblue.net
When Kodak brought out the DSC 460 in 1995 or so, they were telling
folks this was the digital equivalent of general purpose film.
Marketing yes but I have also seen 460 files at 16x20 looking better
than film enlargements. The best color films were about 120lpm in high
contrast drop to low contrast and you have 60-80lpm, this is for Kodak
EPY, or Kodachrome 25, You can squeeze out 200lpm out of tech pan or
150lpm out of PanX, or PanF. But it comes down to not many lenses can
support that resolution. All the res data for films is in Kodak
literature, they were very good at publishing data. I have a color
film book from the early '90s. Again there are high and low contrast
responses of film, which extends to shutter speed. Come to think of it
I don't know if this extends to digital. I'll still say digital
enlarges better, whether it is because it retains edges better or it
just a more pleasing look it just does.
Have to get my *ss in gear to bring out my 4x5, that is where film
shines.
Tom
Pat
You can contort number and equations and other things that you may or
may no know about all day, but in the end the only thing that matters
is what you can see.
Go watch an old movie in a theater where the screen is huge. Go
ahead, and site in the front row. The picture quality is very good
and it's using 1/2 frame 35mm technology. No go and get a 6-8 mp
digital and project it to the same size and see if the quality is
remotely comparable (I won't even make you use a 3-4 mp to even the
field). This is the only test that matters.
Nomen Nescio
"Pat" <gro...@artisticphotography.us> wrote in message
news:0a91b8be-f98f-4986...@s24g2000vbp.googlegroups.com...
>
> Go watch an old movie in a theater where the screen is huge. Go
> ahead, and site in the front row. The picture quality is very good
> and it's using 1/2 frame 35mm technology. No go and get a 6-8 mp
> digital and project it to the same size and see if the quality is
> remotely comparable (I won't even make you use a 3-4 mp to even the
> field). This is the only test that matters.
>
project it as a still frame. You'l see how crappy movie film really is!
It is an optical illusion, the same one that makes a movie move that
deceives the brain into thinking that. Your test is based on no actual
experience.
Steve
>news:0a91b8be-f98f-4986...@s24g2000vbp.googlegroups.com...
Have a peek at this:
http://fwd.five.tv/gadget-show/blog/programme-4-weblinks
The interesting bit begins about 13 minutes into the video where they
compare the results from a Nikon D700 with those from a Nikon F5. The
results are displayed a tad bigger than a cinema screen.
John McWilliams
> Have a peek at this:
> http://fwd.five.tv/gadget-show/blog/programme-4-weblinks
>
> The interesting bit begins about 13 minutes into the video where they
> compare the results from a Nikon D700 with those from a Nikon F5. The
> results are displayed a tad bigger than a cinema screen.
Yikes! How many £ s spent on all the hoopla? Infotainment at its
glorious best.
All to demo what we already knew.
Interesting that the photog. holds his camera upside down for verticals.
Must never have used a monopod.
--
John McWiliams
Rich
$815e...@news.qwest.net:
Tech Pan at 25 ASA out of Technidol developer had very fine grain, but
all films have grain, it's how they work. Until someone somewhere comes
up with a way to render images without making them up out of little
pixels/grain, or Kodak, whomever figures out how to make nano-sized
grains sensitive enough to record light, we'll have film grain.
Rich
That's not so bad, price-wise. About $7.00/36 if you bulk load it.
Rich
news:gnjr36$9tq$1...@theodyn.ncf.ca:
It's possible. Film loses about 1/2 its resolution by the time it gets
printed.
OG
"Steve" <st...@DELETEMEFIRSTzord.co.uk> wrote in message
news:D62dnVSMzOS3FgDU...@bt.com...
> >"Pat" <gro...@artisticphotography.us> wrote in message
> >news:0a91b8be-f98f-4986...@s24g2000vbp.googlegroups.com...
>>On Feb 18, 10:16 pm, RichA <rander3...@gmail.com> wrote:
>>> I've read a lot of the musings from film buffs. Problem is, most of
>>> what they say is just wishful or wrong. The idea promoted by some
>>> that film has "infinite resolution" and the results can be varied
>>> based on increasing scan resolution is nonsense. At best, the
>>> sharpest, finest grain 35mm film can resolve as much as a 6-8
>>> megapixel digital. Film grains, their clumpiness, and the spaces
>>> between them ARE the "pixels" of film and they do not afford infinite
>>> resolution. The highest resolving power film I've ever seen was Tech
>>> Pan (discontinued). It is possible that with the very best lenses,
>>> modern scanning, 35mm Tech Pan could approach the resolution of a
>>> 10-12 megapixel digital, but I really doubt it.
>>>
>>> http://www.kenrockwell.com/tech/real-raw.htm
>>
> Have a peek at this:
> http://fwd.five.tv/gadget-show/blog/programme-4-weblinks
>
> The interesting bit begins about 13 minutes into the video where they
> compare the results from a Nikon D700 with those from a Nikon F5. The
> results are displayed a tad bigger than a cinema screen.
>
Who would use ISO 400 film if they were interested in testing resolution? I
would be more convinced if they had used an ISO100 film for comparison.
Bruce
>Who would use ISO 400 film if they were interested in testing resolution?
Someone who wanted to "demonstrate" that digital was better than film?
> I would be more convinced if they had used an ISO100 film for comparison.
But that wouldn't have given the answer they wanted. ;-)
.
Grimly Curmudgeon
drugs began to take hold. I remember "OG" <ow...@gwynnefamily.org.uk>
saying something like:
>> Have a peek at this:
>> http://fwd.five.tv/gadget-show/blog/programme-4-weblinks
>>
>> The interesting bit begins about 13 minutes into the video where they
>> compare the results from a Nikon D700 with those from a Nikon F5. The
>> results are displayed a tad bigger than a cinema screen.
I saw that. Load of bollocks, made by geeks.
>
>Who would use ISO 400 film if they were interested in testing resolution? I
>would be more convinced if they had used an ISO100 film for comparison.
Yep. Testing the best finest-grained film available against the best
digital would be more appropriate.
I fail to see the point of the idiots who endlessly wrangle about film
v. digital - each has its place and eventually digital will really
outclass film. But so what? Just as there are still people who delight
in using ancient technolgies just for the fun of it, there will always
be people who like and want to use film, if only sometimes.
Grimly Curmudgeon
something like:
> At best, the
>sharpest, finest grain 35mm film can resolve as much as a 6-8
>megapixel digital.
Cobblers.
Ken Rockwell, oh yeah, that says it all.