Wow - read this re: Film versus Digital debate!

[rishij]

Andre: "My Nikon ED 5000's 4000 dpi 16bit scans of 35mm color negative film are clearly superior in terms of dynamic range and resolution than the 14bit RAW images from my Nikon D300."

 

D300 is 12MP, though not full-frame, so some softness introduced by current optics due to the crop factor sensor (I know I'm gonna regret quoting Ken Rockwell here, but: http://www.kenrockwell.com/tech/full-frame-advantage.htm).

 

So 12MP could actually be better, if you were using a 5D, for example.

 

But I'm surprised you say the LS-5000 scans are 'clearly superior'... I mean, it's been a toss up for me (see my posts above comparing the Imacon scans of 35mm film to a 10MP image from a Rebel XTi).

 

Regardless, I think you may find my thread comparing the Nikon LS-9000 to the Imacon interesting:

http://www.photo.net/digital-darkroom-forum/00RTYM

 

I guess, for you, 35mm film would be even more superior than 'clearly superior' if you were scanning on an Imacon and getting even more of the information that was on the film to begin with :)

 

Rishi

[bernardwest]

Hey Rishi, good to see you back. What is your thinking on Hugh's example, and my interpretation of it?

[rishij]

Mark Smith: "I'm NOT suggesting that a single grain is many shades of grey far from it, what I am suggesting is that the filamentary structure of grain is proportional in density to the light that strikes it. Many grains stacked though several layers, overlapping or not, all attenuating light proportional to the silver content create the ILLUSION of tone, just as clumping is an OPTICAL effect, not a physical on, graininess is also OPTICAL."

 

I feel like above, DLT & Mark Smith are arguing the same thing. Except DLT would add that the SIZE of the final resulting filamentary structure, per sensitivity spot that was 'exposed' or had at least 3 reduced silver atoms, also plays a large role in determining macroscopic tonality... such as in an ink jet printing process where the size of dots can be varied.

 

Vijay: "Like I said before, nobody denies that the state of metallic silver can either be present or absent - i.e, one of two states at the end, but the grain itself is not a binary entity. It doesn't transition rapidly from state to state, and it can and does have intermediate, continuous, stable values of silver in it, proportional to exposure."

 

Sure, but the more interesting question is: how many such 'intermediate, continuous, stable values' does one silver grain potentially have? It's, actually, a chemically very interesting question in terms of kinetics of reduction based on amount of initial reduced silver (at various sensitivity sites).

 

Vijay: "40-60 grains; which occupy such large area that it would yield too low a resolution number for film." No... I've already stated above that if a 35mm frame has approx. 10 billion silver crystals, and we allow, not 40-60, but 500 grains to represent one tone, then we get a 20 megapixel image. Which is well near/above subjective numbers for the resolution of film.

 

Vijay: regarding your "Apparent graininess larger than actual grain with "halftone" grain" image... NICE WORK. I've read this explanation a number of times before and your picture shows this concept very clearly. Well done.

 

Vijay: "Exactly Benny - but what is the scale of that clumping? Subgranular or supragranular? The answer will point to the resolution of film, and that will give us a meaningful way to compare film with digital." Well put. That is exactly the point of this entire debate. I actually think, now, that if it's really the density and extent of filamentary structures that grow out from sensitivity centers (that have the threshold of 3 reduced silver atoms), we shouldn't even be looking at this as a 'subgranular' or 'supragranular' argument... because, in the end, it probably isn't even dependent upon the *grains* themselves, but the filamentous silver structures that can grow from the sensitivity centers of grains. Meaning, it's possible that the filamentous structures don't all grow from sensitivity centers to make the entire grain maximally dense... that's not their 'purpose'. Their 'purpose' is to just grow, by the steady reduction of silver halides, as electrons are channeled from the developer, through the conductive cluster of >3 metallic silver atoms at an exposed sensitivity center (since metallic silver is more conductive than ionic silver due to the valence band being fill which leaves the conduction band free to accept and conduct electrons), to the next and the next silver halide. In this view, grains are just there to present sensitivity centers and to provide the raw materials (silver halides)... but it's the growth from sensitivity centers that determines the macroscopic tonality.

 

Mark Smith: "but light can't pass through a gap that could be say 0.2µm? LOL" -- Damn, man, do you realize that visible light starts at 400nm, 0.4µm? Which is bigger than a 0.2µm gap? So, uh, you just make yourself look really stupid.

 

DLT: You argue that the amount of development per crystal controls the amount of final posited silver. OK, I can buy that. But on a grain to grain basis, the amount of development *could* determine how DENSE the resulting grain is (the density of black filaments resulting within the grain). This *could*, I think, give the illusion of tone per grain... though, I would suspect, not much of a range of tones. Because really what would be making up the tones of the entire image, viewed at low magnification, would be how many of these filamentous structures, and how big they are of course, grow from sensitivity centers containing reduced silver metal atoms.

 

In this case, then, it's not even the grains that make up the final image. It's the filamentous structures growing out from sensitivity centers! Overlapping of these dense filamentous structures then gives the appearance of macroscopic grain, as Vijay's illustration so nicely pointed out.

 

Mark: "No that's not true. what is going on at 'sub microscopic level' is vital grain is filamentary not solid black. those filaments TRANSMIT light through layers, which in turn add to the effect of tonality." -- I could buy that. But I bet the effect on tonality is very small... much smaller than the effect caused by the total density of sensitivity centers which gave rise to filamentous structure growth upon development.

 

Mark: "If grain has a filamentary structure light CAN pass through, after all thats what we do with negative during opto mechanical reproduction, we shine light through..." -- Invalid argument, since the light that passes through most likely passes through the clear film in between the grains.

 

DLT: "But we do not see any individual grains between black and white in a microscope view, which means individual grains block all light." -- As much as your arguments make the most sense in this thread, I don't know if I buy this. It could be that a couple of such filamentary structures must be stacked on top of one another to block all the incoming light, compared to just one such filamentary structure on one layer. Think about it -- this'd increase the tonal range recordable. And one would think that if you shine light bright enough, some of it might make it through the filamentary structure, right?

 

DLT: "LOL! If they added to the effect of tonality then that would be observable at 400x!" -- I understand what you're trying to say, but if that 400x magnification really is just a magnification of the catchlight in the eye, then it's possible that it is 'blown out', and so it's just trying to represent pure white, which'd register as Dmax on the negative. In which case, most of the grains would be black. Perhaps we can find a 400x enlargement from a gray section of a negative?

 

That's all I can do for the moment. I'm spent.

Rishi

 

[j_sevigny]

With all due respect to the original poster, who came to this forum with a legitimate question, I'd like to nominate this thread as the most ridiculous in the history of photo.net

[hugh_templin]

Bernie, I'll explain how I got the image of the pixels which will help interpret it. I cropped that area of the original TIFF in Photoshop, changed it to 8 bit, and stored it as a bitmap (BMP). I then opened it in Microsoft Paint. I clicked on View, Zoom, and Large Size. This displayed the pixels as in the sample in my post. To create a picture of it, I pressed the PrtScn button on my keyboard which placed a copy of my screen as a Bitmap on the Clipboard and I paisted it back into a graphic program and cropped it again. Voila! My interpretation of what I am displaying is that the actual pixels which have been captured in the scan are shown as little squares which are easy to see but if you display the picture on a screen or print it , they will be miniscule dots. If you upscale a digital picture, as you did in Photoshop, the pixels are spread out in interpolation but you lose resolution since it blurrs the original pixels into the surrounding areas. What I was trying to show was that the slide resolved the heels of the people down to one pixel wide. If the grains of the film were bigger than that, then you would see clumps of similar colour several pixels wide. Whether this means that it is equivalent to a 23 Megapixel camera or not, I am still amazed at the detail the picture carries when you look at the original size if the slide. Back to you for discussion.

Hugh<div></div>

[James G. Dainis]

J. S. Look at it this way, if a person is having his corn flakes in the morning and suddenly thinks, "Say, I wonder if film is analog or binary (or maybe linear)", there is one site on which he can find the definitive answer. Or not.

[hugh_templin]

Bernie, you will notice that the detail in the enlarged portion is almost dead centre which is the sweet spot of the

lens. This applies to DSLRs as well but we are talking film resolution so this is the best that it will get. Here it is

again for comparison. Hugh<div></div>

[john_castronovo]

LIFE is digital: ignoring all the subtleties, you're either ALIVE or you're DEAD, but that tells us about as much

about life as this discussion has about the difference between film and digital capture. To me, one mimics life

and the other one is but this should be argued by philosophers and not scientists.

[benny_spinoza]

John....Ever heard of Schrodinger's cat?

[rich_evans]

Holy cow - this is one of the longest threads I've ever seen here on pnet......interesting though...........

 

Again, allow me add some potentially relavent comments, and not to split hairs but:

 

(Preface - I have been a research microscopist with a major medical company for the past 30 years, I currently manage a state-of-the-art microanalysis lab with Scanning Electron Microscopes and any number of different types of infrared and light microscopes - most of my work is in materials, not biologicals, so take this from where it comes)

 

bernie quote: "...Well that is exactly what a microscope does, and exactly what an ELECTRON microspoce (sic) doesn't..."

 

Do not allow yourself to be convinced about anything by using any of the referenced or attached micrographs (I don't care if they came from Kodak) without knowing the specific details of how the images were obtained. I can make things look very different than what they really are. There are multiples of detectors (electron and x-ray) and operating parameters that control the final resulting image(s).

 

Daniel quote "...You can't see anything on the film without a light behind the film. The microscope has a light behind (underneath) the film. You turn that light on to observe the film..."

 

Another blanket assumption - transmitted light can be either brightfield, darkfield, polarized, any of a variety of interference techniques - the list goes on, and each will give the sample a totally different appearance. The same can be said about incident light microscopy. Point is we simply don't know enough about how any of these images were generated to make blanket statements.

 

Any solid material at the atomic level is not 'opaque' to light - sputter-coat a 20 angstrom film of gold onto glass and tell me what you see.

 

The rest of your theories may be spot on, but don't make assumptions using any of these micrographs.

 

--Rich

 

 

 

 

 

...

[vijay_nebhrajani]

Sounds like Rishi and I are finally converging. I should like to point out a few things.

 

Rishi: "Sure, but the more interesting question is: how many such 'intermediate, continuous, stable values' does

one silver grain potentially have? It's, actually, a chemically very interesting question in terms of kinetics of

reduction based on amount of initial reduced silver (at various sensitivity sites)."

 

If grain can't be binary, it can't be ternary, quaternary etc. By induction, it has to have a continuous range of

values. The "kinetics of reduction" can't be a discrete process with thresholds either.

 

Rishi: "No... I've already stated above that if a 35mm frame has approx. 10 billion silver crystals, and we

allow, not 40-60, but 500 grains to represent one tone, then we get a 20 megapixel image."

 

Only if these were "grayscale" crystals. Binary crystals, because the black ones stacked in the third dimension

can't create tone, so you're talking something like 49 5 micron crystals in the x-y dimensions - resulting in 35

micron sized "pixels", resulting in about 16 cycles/mm. See my calculation earlier in this thread.

 

Rishi: "Damn, man, do you realize that visible light starts at 400nm, 0.4µm? Which is bigger than a 0.2µm gap?

So, uh, you just make yourself look really stupid."

 

You, of all people should know better. The C-C distance in diamond is 1.54 angstrom - something like 0.000154

microns, yet diamond is transparent. Opacity is a resultant property of the arrangement of atoms or molecules,

and just because a "gap" is 200 nanometers doesn't mean light of wavelength 400 nanometers won't pass through.

Its not like the photons have a 400nm radius and its not like atoms are opaque or transparent. Don't go name

calling unless you are sure you are right.

 

And finally to Bernie - try implementing a binary system (only two stable states and a threshold for switching)

and you'll know that it is not as simple as two states. Think of a mousetrap. This is a binary system, which can

have two states - set and sprung, with a threshold of weight on the lever - a fly landing on it would not trip it

but a mouse stepping on it would. This is a contrived mechanism, and occurs rather rarely in nature. If you

postulate that a grain is binary - meaning either entirely clear or entirely black, you are implying such a

mousetrap - in this case a photon trap - which is initially clear, but when some predetermined threshold of

photons fall on the grain it immediately jumps to completely opaque. (Of course, this results in absurdities like

independence from development time, because splashing developer after the photons have either tripped the

mousetrap or not is useless, more traps can't be sprung with the developer since the tripping mechanism is a

photon threshold. Sort of like you can't trip a mousetrap by blowing at it.)

 

A) Therefore the onus is on the binary crowd to show evidence of this mousetrap mechanism.

 

B) Or, if you wish to state that a grain does not jump immediately from state to state it is not binary, i.e.,

must have a continuous range of values.

 

This argument IS binary though - either A or B above have to be true. No other state is possible.

[bernardwest]

Thanks for the input Rich. I guess we need to see images from a microscope that transmits light in the same way as an enlarger. Any idea what the equivalent to an enlarger is?

 

Benny... I've been waiting for the cat to make an appearance. I was thinking about introducing it a hundred or so posts ago, but couldn't bring myself to do it. The cat's out of the box now... so to speak...

[rich_evans]

In my spare time (heh...heh) I'll endeavor to capture the appropriate images using brightfield transmitted light microscopy techniques and see what I get - that is if I can find any film..... ;-)

 

I do however have an interesting SEM micrograph of a single grain of TMAX 400 which I'll post here tomorrow when I get back to work - just for kicks.

 

--Rich

[bernardwest]

<i>In my spare time (heh...heh) </i>

 

<p>Thanks Rich. Surely you can slip this one under the work radar and claim you are "calibrating" your microscope

with it? ;)<p>

 

We've got all sorts of fancy pants stuff here at my work. I might try and see what I can do with some of it. So,

it's 'brightfield' microscope I am after?

[bernardwest]

<i>Whether this means that it is equivalent to a 23 Megapixel camera or not, I am still amazed at the detail the picture carries when you look at the original size if the slide. Back to you for discussion. Hugh</i><p>

 

Hugh, thanks for that. But I dunno. I'm not convinced. Surely if you had of scanned your film in at 78MP, you would see the same thing but with multiple pixels representing the heels (by the way, I still maintain that your image shows the heels as being multiple pixels wide at 23MP anyway). That doesn't mean your film is good for 78MP. I think your reasoning on this one is flawed.

[philip_sutton]

J Sevigny - I kind of agree with you and thanks for the comment.

 

These guys have just gone off on a tangent and are totally oblivious to anything else. I asked them a civilised

question yesterday (24 hrs ago), and they did not even notice it let alone acknowledge it. I think they should go off

on another thread and pursue this nonsense and let us plebs have a normal disucussion and try and return to the

essence of my original question. I Hope this can happen because these scientist dudes have (by virtue of their

technical jargon and hard headed attitude) excluded everybody else from this thread!

 

I am sorry to take this stand - but the route this has taken is really starting to annoy me.

 

Cheers - Phil

[jorge_a]

Phil, if you read the earlier posts you might find simple answer to your question. Film or Digital does not matter one bit if

you are happy with the end result. Don't worry about micro nano technology. Jorge

[philip_sutton]

Yes thanks Jorge - I appreciate your comment. I have tried to wade through all this technical stuff but it just rips the

heart out of what I think is the essence of photography.

 

My new 67ii arrived and just love it and the process so much. I am getting these awesome trannies (Velvia 50) that

leap off the light box. The whole process of taking my time to shoot, carrying around such a well made hunk of

engineering, feeling the film slide through the film back - all has such a hands on organic feel that I just don't get with

my digital camera (I have a small Sony 7 meg camera).

 

I have to shoot off to work now but when I come home tonight I want to ask some more questions on what you guys

think I should use for a scanner to get the best out of these huge, gorgeous trannies. I have decided to hang on to

my F5 system as well, so it will have to be able to take 35mm too.

 

I know there are other threads like this in the archive, but I want to ask about flat-bed scanners (seem to be all the go

now), as opposed to the results I may get from a good film scanner (like one of the Nikons).

 

I am sorry if I came across as unkind in my prior comments - it was just that I felt the essence of what I had asked

and the heart of this converstion has been hijacked!

 

Cheers - Phil

[vijay_nebhrajani]

Philip,

 

I find your current attitude somewhat arrogant - you asked a legitimate

question but one that has no simple answer. You pointed to Ken Rockwell, and someone else pointed to Michael

Reichmann. What we are doing is examining the veracity of Reichmann's claim, which would sort of answer if Ken

Rockwell's original claim is credible or not.

 

Or it would answer your question "I was wondering if some of you folks who know a heap more about this subject

than me at the moment, could read this and tell me your opinion. Is this guy credible. Is it true what he asserts."

 

So here is a bunch of guys who do know a heap more about the subject (which is film, digital and resolution)

trying to come to an understanding about whether Rockwell is right, or Reichmann is, or both or neither.

 

You should be proud that your question prompted such a healthy, albeit arcane debate and that some 300 posts

later there seems to be some convergence on an answer. Everybody noticed your question - and unfortunately it was

irrelevant to the current discussion and has a simple answer - no. (Your question was: Aren't we really splitting

hairs by argueing over all the technical stuff.) We split hairs when we disagree on minutiae - here we are

disagreeing on the entire process of silver image formation, though probably not everyone realizes that.

 

I have already said that the mods can consider moving this discussion to another thread, but the original

intensity and focus is often lost when that is done. I also apologized on everyone's behalf on the arcane nature

of the debate and asked that we "please, please" be allowed to continue. Are you saying we should stop?

 

Sorry man, I don't want to sound off on you - but I absolutely hate the attitude that one should just do whatever

works without understanding the underlying processes; even everybody's hero Ansel Adams wasn't satisfied with

that approach.

 

So if someone is the greatest artist on this planet and doesn't want to understand the nitty gritty of his art,

thats okay, but for that person to deride those who want to is plain wrong.

[bernardwest]

<I>If you postulate that a grain is binary - meaning either entirely clear or entirely black, you are implying such a mousetrap - in this case a photon trap - which is initially clear, but when some predetermined threshold of photons fall on the grain it immediately jumps to completely opaque. (Of course, this results in absurdities like independence from development time, because splashing developer after the photons have either tripped the mousetrap or not is useless, more traps can't be sprung with the developer since the tripping mechanism is a photon threshold. Sort of like you can't trip a mousetrap by blowing at it.)</i><p>

 

Vijay, couple of points. I'm not experienced enough with the chemistry of the process to argue one way or the other, hence the reason why I can't directly deal with your buddy, the <i>reductio ad absurdum</i>. But that being said, I think your statement that it is a photon trap may not be right. Hopefully someone with better chemistry knowledge than me can jump in now, but I'll just keep flailing around in the dark for a little bit longer. To match it with your mouse trap analogy: sure, a fly can't set it off, a mouse can, but there are also many other things that can set it off. It's been pointed out that developer in the <b>absence</b> of any latently exposed grains can still cause the film to blacken. Clearly it is the case that photons are not the only trigger for this reaction. Think of it as a hummingbird that is hovering above the mouse trap. Let's say for arguments sake that it can't move sideways, only up and down. Now, initially, it won't set the trap off because it can hover above it. But wait enough time and it will tire, at which point it will fall, setting the mouse trap off. I think you need to modify your analogy if you want me to address your buddy r-a-a.<p>

 

And i've forgotten what my other point was, but when I think of it, I'll let you know.

[bernardwest]

<i>I am sorry if I came across as unkind in my prior comments - it was just that I felt the essence of what I had asked and the heart of this converstion has been hijacked!</i><p>

 

Phil, sorry you feel that way. But I think you received many great answers to your question. How many more answers do you need? The thread has moved off on a tangent. I'm sorry it hasn't dealt ONLY solely with your needs, but it has been a very infomative thread (particularly for me who previously had little knowledge how film worked). All threads meander their way along. If we dictated that a thread could ONLY address the intial question, then we would all be the poorer for the knowledge that didn't have chance to get out there and be shared.

[rishij]

Bernie: "So, following what i think is Hugh's logic, I should now be able to claim that my original 13mp image carries as much resolution as a 78mp dslr because I can discern edges and tones at the pixel level. If this is the wrong interpretation, let me know."

 

No, because in your blow up, there does not exist one detail that is represented by one pixel, whereas in Hugh's example, the heel is represented by 3 pixels, but one of them (in the center) is clearly darker than the two around it... meaning that the heel was sort of, almost, resolved with 1 pixel (ok maybe 2 pixels). Making it on par, approximately, with a 12-20MP digital image. I'd probably say more on the 12MP end, since it's more like an average of 2 pixels that needed to be used to resolve that heel.

[rishij]

<i>Vijay: "I've been repeatedly saying that at the visual scale, you may see black specs but there is no information at that scale that tells you that several black specs are not part of just one grain. If they are, then the grain is sort of grayscale by halftone, if they are not, then the grain is binary, i.e., a switching system."</i>

<p>

Good. I agree. But now you realize we're just getting down to an argument of: "At 400x, the black specks are simply filamentous growths from sensitivity centers, regardless of what grain they come from. So, while a grain doesn't have to be binary, as in all of it is converted to metallic silver, the final filamentous growths from sensitivity centers are 'binary' in the sense that they're predominantly opaque. What then determines tonality at lower magnification? The size of these filamentous growths, and the density of them at any given spot. Which is basically like a halftone process."

<p>

So then this entire debate boils down to an issue of semantics, essentially. Sure, then, a grain itself may not be binary because the filamentous growths that started at the sensitivity centers did not grow all the way so as to make the entire grain black, but the final resulting growths themselves are densely opaque.

<p>

Actually, I need to draw a diagram to make clear what I'm trying to say.

<p>

<i>Vijay: "So by going to smaller grains & shorter wavelengths of electromagnetic radiation, you are saying you can't increase the resolution of film? Remember I said long ago that film is not able to resolve finer than a grain, but this has nothing to do with the tonality of that grain. So you have to go to a smaller grain, but that does not make it a switching system."</i>.

<p>

Wrong, it has <i>everything</i> to do with the tonality of that grain. If that grain can represent 16,384 tones (14 bits), then, yes, you'd have film with molecular resolution. This'd give a 35mm frame of film a 10gigapixel resolution. Which is absurd.

<p>

Vijay: <i>"Molecular resolution has to be the eventual limit; what's so absurd about that? If you could create a case wherein exactly one AgBr molecule got converted to Ag, and if you had the equipment to precisely observe this, you could."</i>

<p>

No, you couldn't. One Ag molecule can't represent thousands of tones, so it'd have to be clumps of Ag molecules representing a tone, thereby making that one molecule of Ag <b>not</b> the smallest resolving element of the sensing medium (film). That is absolutely absurd.

<p>

Vijay: <i>"But if the question is if the maximum attainable resolution with film is the equivalent of 25-30 MP, then the answer is yes."</i>

<p>

No, the answer is no. Because if you're right and one grain can represent thousands of tones, then that would be the smallest resolving element on the film. And then, if you had an ideal, theoretical, lens, that had that sort of resolving power, then a 35mm film frame would hold 10 gigapixels worth of data. But, since a lens can't resolve this sort of detail, let's look at a more real-world situation. Take the same lens and use it to expose a 35mm frame of film, and then use it to expose a 20 megapixel full-frame digital CMOS sensor. Since the film, according to you, would technically have 10 gigapixels worth of resolution, whereas the CMOS sensor only has 20 megapixels worth of resolution, with the same lens here, the <i>35mm frame for film should be able to resolve at least as much as the 20 MP CMOS frame, if not more (given the lens can resolve more)</i>. This is <b>contrary to observation</b>. How do you reconcile that??

<p>

Vijay: <i>"You, of all people should know better. The C-C distance in diamond is 1.54 angstrom - something like 0.000154 microns, yet diamond is transparent. Opacity is a resultant property of the arrangement of atoms or molecules, and just because a "gap" is 200 nanometers doesn't mean light of wavelength 400 nanometers won't pass through."</i>

<p>

True, Vijay, sorry. What I meant to say is that you'll start to have diffraction effects when you have gaps, amongst opaque material, whose size are on the order of the wavelength of light. So light won't just happily transmit thru as if the gap were much wider. I forget how this is relevant to this discussion though :)

<p>

Rishi

[rishij]

Rich: <i>"Any solid material at the atomic level is not 'opaque' to light - sputter-coat a 20 angstrom film of gold onto glass and tell me what you see."</i>

<p>

VERY GOOD POINT. Which is exactly why we need to take a step back and look at the 400x magnification (feeding to the Bernie/Daniel/me camp). Which is also why, in a sense, we should probably also concede layers of grains, stacked, increase the overall tonality possible. I.e. three "black" (I use quotes b/c seriously after all this discussion it's still up for grabs what exactly a 'black' grain is :) grains stacked on top of one another may block more light than just one grain for that given Z-axis coordinate.

<p>

Perhaps we should start taking this discussion toward color film... since we can agree color dyes are somewhat transparent, and the amount of dye that forms depends on the amount of reduced silver (and concomitant oxidized developer) at any given location... which argues towards a a linear buildup of dye making a tone, which then... OMG my head hurts.

<p>

:)

<p>

Rishi

[rishij]

Vijay:<i>"If grain can't be binary, it can't be ternary, quaternary etc. By induction, it has to have a continuous range of values. The "kinetics of reduction" can't be a discrete process with thresholds either."</i>

<p>

Yes but how many of these tones matter from a macroscopic level? Also how many tones are possible depends on the total dynamic range of the grain (clear to however black/dense a grain can get).

<p>

At any rate, I don't think this is relevant because we should probably shift the discussion from talking about whether a 'grain' itself is binary to whether or not the filamentous growths growing out from a sensitivity center is binary (which I think it is in terms of opacity, but is analog in terms of size). The grain, then, just becomes an element in the film that presents sensitivity centers for filamentous Ag outgrowth. which means now we're asking 'is the sensitivity center' binary? To which I say, probably not, because the number of initial reduced silver atoms at a sensitivity center, determined by photon exposure, determines the final size of the resulting filamentous outgrowth.

<p>

I think this is exactly what Daniel is trying to argue?

<p>

Rishi