It's well known that the aperture of a lens can limit resolution through diffraction, but diffraction occurs when light passes through any small gap.
Now, if we have a negative with small detail on it, then light must also be diffracted as it passes through the negative on its way to the enlarging lens. Doesn't this limit the ultimate resolution we can get in a projected image, or at least degrade the contrast of fine detail? Further, since this diffraction occurs even before the light has reached the lens; doesn't it also mean that longer focal length enlarging or projection lenses would tend to suffer from this effect more than short ones?

Logically, this effect must happen, but I've never seen it dealt with in any book or article.

-- Pete Andrews (p.l.andrews@bham.ac.uk), July 05, 2001

Answers

I'm not smart enough to answer this one, but I do know that in conventional microscopy magnification is limited by the wavelength of light. An enlarger is just a form of microscope, and it's certainly limited by the same effect. Light rays certainly must diffract around the silver grains, but maybe the effect is masked by what the aperture and paper path length contribute. As an interesting aside, it might seem possible to avoid diffraction by using a graduated aperture with no hard edge. I saw an analysis of this in an optics book and, as one might suspect, you can't cheat mother nature. It doesn't work out any better than the usual hard edge!

-- Conrad Hoffman (choffman@rpa.net), July 05, 2001.

Uhhhh, what? And here I had enough problems understanding that you need to extend the shutter speed when you close down the aperture. :P

-- Johnny Motown (johnny.motown@att.net), July 05, 2001.

Hi, Pete. I guess you've found out why some people prefer huge cameras and contact prints... You may read some good info on the matter of resolution in "Controls in B&W Photography" by Richard Henry, Focal. Good reading.

Cesar B.

-- Cesar Barreto (cesarb@infolink.com.br), July 05, 2001.

Pete,

Whether diffraction is occurring between the grains of silver or not is irrelevant. The lens interprets the negative as an object. It does not matter if the object is lit with reflected or internal or projected light. Diffraction occurs in the making of the virtual image of the object.

-- Pat Raymore (patrick.f.raymore@kp.org), July 05, 2001.

Don't really know. However, a few stray rambling thoughts...

Even the best enlarging lens is probably incapable of resolving a single grain. However, what this means is that the edge of a grain clump is probably not going to be resolved with the same level of contrast as actually exists. However, once you get to this level of resolution, you really are up against the limits of the medium i.e., one is actually breaking up the structure of the image - therefore, dare I submit that this might actually be a good thing since it smooths out the micro-tonalities? That is, it serves to fill in the spaces between the grains (by reducing the contrast) to create the illusion of seamlessness. In addition, this ignores the fact that the grains are distributed in a three dimensional emulsion space (and not in a single plane) which again complicates matters.

In the final analysis, there is nothing one can really do about light scatter in the emulsion. The enlarging lens can only take the light emerging out of the negative and attempt to focus that light.

While on the subject, could someone explain how silver grains stop light partially? Is it that they absorb some of the photons. Or is there a plausible explanation based on the wave nature of light? Cheers, DJ.

-- N Dhananjay (ndhanu@umich.edu), July 05, 2001.

Hi DJ. Silver grains stop light partially? I'm not sure what you mean there. The silver-halide crystals are transparent before they're developed, allowing some light to pass through them, and so expose crystals deeper in the emulsion.
Once the film is developed, the tiny filaments of silver are totally opaque, and any illusion of shades of grey is caused by a 'dithering' of light and dark mixing in the eye.

You've only got to enlarge a fast film by a few diameters, or look through a grain magnifier to see that film grain can easily be resolved by an enlarging lens.
After further thought, I suppose what I'm really asking is, whether longer focus lenses give less micro-contrast than shorter ones in a projection/enlarger system.

-- Pete Andrews (p.l.andrews@bham.ac.uk), July 06, 2001.

"After further thought, I suppose what I'm really asking is, whether longer focus lenses give less micro-contrast than shorter ones in a projection/enlarger system."

- Empirical question. Should be answerable with access to a microdensitometer. If I get this right, the argument is that the lens is getting an already scattered beam with some spread function to it, and this will be worse for long focus lenses since the longer path provides more distance for an angular spread fuunction to spread. Actually, I may have this wrong but one need not hit the level of individual grains, right? Any edge will cause some dispersion around the edges - with small apertures, the problem is that the 'diffracted' PERIMETER of the beam is a significant proportion of the 'undiffracted' AREA of the beam. So, if it is the edge contrast we are interested in looking at, even a relatively large area like the step of a step wedge (or a knife edge) should suffice. That is, a microdensitometer trace of the original step wedge and the copy should reveal lower edge contrast in the copy.

Another random thought. What about interference? Each grain clump provides some amount of spread but the adjacent clumps also provide some spread. Is it possible that the 'dispersed' portions of the beam largely cancel each other out while the 'undispersed' portions are unaffected? Sort of serves to increase signal to noise ratio, in a sense.

Not to hijack the thread but I'm still trying to puzzle out the light stopping ability of silver grains. I understand the greater depth of exposure in the emulsion - I presume what you mean by the 'dithering' is that there is a smaller area where a straight line path through the emulsion does not hit a grain of silver. That is, one still has a high contrast bunch of dots at the easel but the eye cannot resolve these dots and merges them and the white base into a grey. The reason I was wondering about this is because I was studying some negs on a light table. When I switched off the light, I noticed a partial reversal of the tones - it looked like the silver was actually partially reflecting light - stray thought was whether that was a contribution to how tonalities were created, because this selective reflection should reduce the intensity of light that got out on the other side.

Cheers, DJ.

-- N Dhananjay (ndhanu@umich.edu), July 06, 2001.

DJ, even though you are not "hijacking the hread", I think you might be interested in reading chapter 31, "The Origin of the Refractive Index", from the Feynman Lectures on Physics, Volume I (1963). It discusses several effects in terms of electric fields in a simplified model; the effects include, among others: refractive index, dispersion and absorption. They all relate to your question about partial transmission.

To give the general tone of the discussion, I excerpted a paragraph from 31-1: he has been talking about how electric fields in a glass plate drive electrons back and forth; these electrons constitute new radiators:

>> [from Feynman] Incidentally, you should notice that there is another effect caused by the motion of the charges in the plate. These charges will also radiate waves back towards the source S. This backward-going field is the light we see reflected from the surfaces of transparent materials. It does not come from just the surface. The backward radiation comes from everywhere in the interior, but it turns out that the total effect is equivalent to a reflection from the surfaces. These reflection effects are beyond our approximation at the moment because we shall be limited to a calculation for a material with an index so close to 1 that very little light is reflected. <<

To Pete: I think your original question is better answered by Conrad's original contention that resolution limits at the enlarging stage is equivalent to the situation with a microscope and that is where the best literature would come from. Note that the numerical aperture (along with wavelength of light) sets the resolution limit so it should not rely strictly on focal length. Numerical aperture is similar to an effective f-number but is expressed in terms of an solid angle. The light source behind the target (negative) must be larger than the lens' numerical aperture or it (the source) will limit resolution.

My disclaimer: Don't understand this well; just enough to be dangerous.

-- Bill C (bcarriel@cpicorp.com), July 06, 2001.

Have you ever looked thru a dirty window at night at a distant light? Did you see a colored diffraction pattern? Our front window doesn't stay clean and I often see these patterns. If I flog my memory back to college physics I remember something about the angle a particular wavelength is diffracted off the zero order is a function of the wavelength and the pitch of the grating. Exposed and developed film is like my dirty window and is a 2 dimensional diffraction grating. Could it be that the light that is not absorbed or reflected by the negative is diffracted off the optical axis so far that it misses the lens pupil?

If memory also serves me correctly, it's not the 'grains' that actually block the light. The grains are clumps of the silver crystals and the actual crystals are much smaller; perhaps on the order of the wavelength of visible light. Films such as Kodachrome 200 have very large grain but are sharper than would be expected from the size of the grain.

I wonder if this is the explanation for the Callier effect? ie. greater than expected observed densities for particular densities when using a collimated light source.

Now having thought about this, I'm not so sure anymore. If it were a strong phenomenon, every B&W negative would have an observable color spectrum. Oh well,

Cheers,

Duane

-- Duane K (dkucheran@creo.com), July 06, 2001.

Thanks guys.
This thread seems to have ground to a halt, after going off at more of a tangent than a diffracted light ray! At least you've confirmed my suspicion that there isn't much info on any diffraction effect in negatives.

-- Pete Andrews (p.l.andrews@bham.ac.uk), July 20, 2001.

Hello Pete,

to answer your question, no, diffraction in negatives does not limit the final resolution of a print in any way. The thing that distinguishes a lens aperture from a negative is that whereas (in theory) light rays are supposed to go straight through the aperture, not changing their direction in any way, they ARE supposed to scatter when going through a negative. Scattering (off silver particles or dyes, doesn't matter) alters the direction of the light rays from the source randomly and in all directions; diffraction will only add to the scattering. The scattered light is then collected by the lens, which refocuses all rays coming from a single point on the negative (no matter in hich direction they were going after being scattered or diffracted) onto a single point on the paper; thus the diffraction off the varying density pattern on is automatically corrected and accounted for.

Note that I'm assuming the size of detail on the negative is much bigger than the wavelength of the light, i.e. the light will scatter and diffract on it. Microscopes (as mentioned by a previous poster) have the opposite problem - when things get too small, they will neither scatter nor difract the light that passes through them, thus leaving the light unaffected by the passage through them, and so the light carries no information about the object, i.e. the object is invisible in the microscope.

Once again, the difference is that the aperture is a part of the imaging system that's NOT supposed to scatter light in random direction, whereas the negative is the imaged object and IS suposed to scatter light in random directions.

-- Peter Langfelder (plangfel@insti.physics.sunysb.edu), March 17, 2002.