Toralf writes:

Hi.

I'm still wondering about how good the image quality of modern digital
cameras (especially SLRs) really is, in particular how it compares
with 35mm film. I've seen many articles on the subject on the Net, but
few of them seem to give you a lot of tangible information (I want to
see the numbers, please), and I can't help feeling that tests they
refer to are usually done to prove a point, i.e. that digital cameras
are as good as 35mm, which is not the way you do proper research.

To say a few words about myself, I'm working for a company that makes
high-accuracy, large-format scanners, so I'm not particularly
impressed when I hear e.g 6 million pixels (you need to talk about
*billions* of pixels if I'm really going to listen), and the word
"interpolation" leaves a bad taste in my mouth. But this also means I
know that high resolution isn't everything, of course; parameters like
geometric precision or signal-to-noise ratio also count a lot.

Be that as it may, some of the questions I'd like to have answered are
these:

1. What is the resolution of a 35mm film anyway?

It's hard to say, as the resolution limit is different from that of a
digital sensor. Rather than a hard limit, you get less information and
more blur and noise as you increase resolution in scanning a piece of
film.

I think I read
somewhere that a colour negative is at least 3000dpi. Is that
correct?

It's in the ballpark. The data sheet I have for Provia 100F shows an
MTF of 35% or so at 50 cycles/mm. That corresponds. to 100 samples/mm,
which is 2540 dpi. You presumably know what MTF means, but for others
who may be reading, that says that if you image a sine-wave grating
onto this film at a frequency of 50 cycles per millimeter, you will
get about 35% of the contrast back from the film. But then what comes
out of the lens will be reduced in contrast from what was in the
envirnoment, so it's on the ragged edge of what you might really hope
to reproduce. The MTF for Ektachrome 100 Pro is right at 30% at 50
c/mm. Kodachrome 64 is about the same, and Kodachrome 25 is about 35%.

How about black&white?

Well, the data sheet on Kodak Technical Pan, which is an extreme upper
bound, has MTF at 50% at 180 c/mm or so. That's for ISO 25, developed
in Technidol. That's Kodak Tech Pub P-255, available at
http://www.kodak.com/global/en/professional/support/techPubs/p255/p255.jhtml?id=0.1.18.14.21.22.16&lc=en.
For Plus-X, at ISO 125, Tech Pub F-4018 (at
http://www.kodak.com/global/en/professional/support/techPubs/f4018/f4018.jhtml?id=0.1.18.14.21.20.16&lc=en)
rates it at 50 lp/mm at 1.6:1 contrast, 125 at 1000:1 (which you
aren't likely to reach it practice).

So I think we can say that scanning film at more than, say, 360
samples/mm (about 9,000 samples/inch) is pretty much useless. In most
circumstances, 100 samples/mm or 2,540 /inch is plenty. So we're in
the region of 9MP for a normal 24x36mm 35mm film frame.

(Yeah I know, a film doesn't have
pixels in exactly the same sense as a digital image, but it *is*
made up of discrete elements after all.)

But they aren't spaced on a regular grid, so things look a lot different.

2. What about the print? 300dpi?

Well, there are lots of variables involved there. One of the little
secrets that the "film-only" zealots don't seem to mention is that to
view the film, it must either be scanned, projected, or printed. I
have heard the claim that scanning digitally loses quality, but so do
the other two processes. How good are the optics in your enlarger or
slide projector? How perfectly is either one focused? I really don't
know what sort of degredation is involved, but it seems that it might
be on the same order as through the camera optics, or worse. And lens
MTF's usually aren't quoted beyond 40 cycles/mm! Check the Zeiss or
Canon Web sites if you don't believe me.

3. I know that the most common sensors are made up of individual
elements for the red, green and blue channels, arranged in a
special pattern, whose data is somehow interpolated into RGB
pixels. But what exactly does e.g. 6 megapixels mean in that
context? Does it mean that the sensor has (just) 6 million
elements, or that data from a higher number (like 18 or 24 million)
is combined into 6 million RGB pixels?

The former. Except for the Foveon sensor in the Sigma SD9/SD10, which
has about 3.5 million sites, each of which detects all three channels.

The same question more bluntly put: When Canon/Nikon/Pentax is talking
about 6MP, is that just a big a lie as the one about 10MP on Sigma
cameras? (I'm hoping not, as I think the Sigma/Foveon way of counting
really takes the cake.)

4. Can the inaccuracy associated with the above mentioned
interpolation be quantified and/or measured against e.g. the error
introduced by scanning a negative with a film-scanner? And how does
it compare with pixel interpolation in the scanning sense?

Not really. The reality is that sensing all channels at each location
is a Good Thing, other factors being equal, but the color filter array
isn't as bad as you might think. First of all, the three color
channels aren't completely uncorrelated, so the information from a red
sensel can be used to help estimate the red and green values at that
point. Second, images have some sort of spatial structure, and modern
demosaicing algorithms try to detect that to deduce missing
values. Finally, the sampling rate is higher for the green channel, to
which the human visual system is most sensitive, both in luminance and
resolution.

5. And how about those other parameters I mentioned briefly above -
like different kinds of geometric distortions, noise, flat field
bias etc.? Can those be compared with the ones of plain old film?

Yup. Lots of them are only being noticed now with digital sensors
behind the same lenses that people have used for years with film.
I suspect that it's a matter of display more than anything else;
the magnification on screen at a 1:1 pixel magnification is huge
for most cameras, and people can use the little eyedropper to see
just how much the illumination falls off in the corner of the frame.

6. And the chromic aberration effects? How serious are they these
days? And are the full-frame sensors that are actually found in
some high-end cameras now, in any way comparable to film in that
respect?

Chromatic aberration is basically in the lens. Doesn't matter
what sensor is behind it.

Well, maybe some people will say I have a somewhat critical or
conservative attitude towards digital cameras, but I actually think
you ought to be a bit sceptical when something "new and wonderful"
comes a long; new technology is too often introduced for technology's
own sake, IMO.

Oh, sure. But there are a number of photographers who are putting away
their 4.5x6 film cameras because they see better quality out of the
Kodak DCS 14 MP cameras. In a functional way, the best digitals are
pretty good, though their limitations are different from those of film
cameras.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.