Digital is analogue, film is binary….
Filed in Film & film cameras - December 11, 2007You’ll have heard that. The digital part is right – that’s why sensors have an analogue to digital converter attached. But the film part? I’ve always been slightly uneasy about it. I mean, as typically expressed:
Grain particles are binary. An individual film grain can only be either black or not-black, on or off, exposed or not exposed.
it sounds plausible and when you magnify a mono neg you can see the grains. So, true?
Well, maybe not:
Film emulsions are generally Ag/Br/I atoms combined into crystals from about 1 – 10 microns in size. They contain millions of atoms and many sensitivity specks which consist of sulfur and gold.
When film develops, it can form anywhere from 3 silver metal atoms minimum up to the entire grain, and grains can be stacked, and therefore the dynamic range of density is analogue in nature and virtually infinite. For practical purposes, it ranges from 0.1 – 3.0 density units in a normal negative B&W film.
I can’t verify that information, but it has a genuine ring about it, and the blog that it is on Photo Utopia is an informed and thoughtful sort of place.
Does it matter? Probably not, but I like to know stuff, not simply play follow-my-leader.
That’s correct. The film grains clump during development and their size is very much a continuous function of exposure and development. Not binary at all.
Aha!, I just noticed the source of your first “binary” statement is a website not at all known for its insights into the basic physics and chemistry of photography. I have caught other howlers there in the past.
scott
December 11, 2007 @ 2:21 pm
I like that!
December 11, 2007 @ 2:39 pm
The common misconception here is that what we call grain on the prints is, well, not grain at all. Since the silver grains are opaque, they stop the light and is we saw actual grain, it would be white dots, not black dots.
So, what we call grain are the holes between the actual grains. Each so-called grain can indeed have many grey values.
I had seen the sentence you quoted and I don’t think it is true.
The explanation about the grain coming from the holes comes from Barry Thornton.
December 11, 2007 @ 4:02 pm
From the quote:
“When film develops, it can form anywhere from 3 silver metal atoms minimum up to the entire grain, and grains can be stacked, and therefore the dynamic range of density is analogue in nature and virtually infinite”
This sounds like an argument based on an ideal film, ideally exposed. In other words, pretend for a moment that you’re God, and can manually arrange the silver atoms however you like, in grains of a size that suits you best, with whatever spacing between, or stacking of, the grains gives you the best effect. By implication, you have an unlimited amount of time to fiddle around with this procedure until you’re happy with the results. In that case, the range of densities is, indeed, ‘virtually infinite’.
I think it would be much more convincing, as an argument, to use actual measurements from actual pieces of exposed film. Otherwise the argument is not realistic, it’s just rhetorical. It’s a nice idea, but we don’t know how close to this idea we can actually get, with the materials at our disposal. It appears at first sight to be grounded in reality (because it talks about the low-level entities involved) but it isn’t the whole story, because it omits the degree of control needed to achieve this ‘virtually infinite’ range.
BTW, I do have a lot of sympathy for the argument, despite using a digital camera.
Just my two pen’orth.
December 12, 2007 @ 12:22 am
Alan,
I disagree. The argument that film is binary is, itself, a binary argument. It is either true or not true. If it is even a little bit wrong it is wholly wrong. Actual films may not reach the perfection suggested on Photo Utopia, but the point isn’t that every film is ideal. The point is that no film is binary.
December 13, 2007 @ 7:39 pm
scott,
On the source – I agree. MR puts himself forward as an educator and also has a big audience. This means that things which appear on his site get widely disseminated. All the more reason to publish corrections when he is wrong (which, before anybody gets het up about this, isn’t to suggest that he is always, or even often, wrong)
December 13, 2007 @ 7:43 pm
All,
Apologies for my delay in responding. Blame an unexpected period of no internet access.
December 13, 2007 @ 7:44 pm
Alan,
Remember that this article is meant to redress the wholly incorrect one posted on LL.
Your quote from my blog ignored a very important qualifier that is normal negative films have a density of 0.1- 3.0 and the values of density between those points is infinite, that is it is governed by how many photons hit the grains.
If one photon hits a grain then the ag atom moves to the sensitivity speck and a very small 3 atom structure is produced, but if many photons strike the grain then millions of ag atoms attach to the tens of thousands of sensitivity specks, the resulting structure is rather like a wire wool pad, that is it transmits little light.
There are a couple of problems with making the process (well more but these are the big ones)
The energised ag atoms need to find the specks and film manufacturers coat the sensitivity specks (sulpher) with gold in order to stop the ag atoms ‘roaming’.
The other problem is the activity of the developer, if this is too great then grains that should be a certain tone say mid grey will be developed to completion, that is they will turn black causing uneven tone.
This is why Tmax developers don’t always give good results with cubic emulsions like HP5
So my blog uses actual scientific facts rather than the total fud espoused by MR in his binary grain–40 grains ‘clumping’ to make a mid-tone I’d love to how they move though the emulsion to create mid tone as MR puts it ‘like an ink-jet dithers’ I wonder who the chump is?
December 18, 2007 @ 5:09 pm
May I add that digital is not exactly analogue, as the sensor is just a photon counter. The count is converted into the charge, that is voltage – and then converted back to the count. The typical conversion rate to get good S/N ratio across the whole dynamic range is 6-10 photon counts to 1 output count.
With film, overexposing regular b/w film by 2 stops and developing it in D-23 diluted 1:1 at 15C for 15 mins results in what one might mistake for an infinite dynamic range – so huge it is :)
December 21, 2007 @ 2:44 am
Binary or not is only part of the equation. It’s also a matter of bit depth.
May 21, 2008 @ 1:58 pm