Wow - read this re: Film versus Digital debate!

[bernardwest]

<I>Did I have to spell it out for you like that for you to get it, or could you have just inferred it from the tens of times in this thread I've already explained it?</i><p>

 

Neither, Rishi. I've tried both approaches, over and over and over again, and they have no influence on the man. He will ABSOLUTELY not condede his position, no matter what evidence or logical reasoning you use. There really is no point going on (except the satisfaction of watching Vijay flailing about desperately trying not to be pinned down on a point). Vijay... at the start of this thread, you had credibility. You haven't got much left to play with man.

[vijay_nebhrajani]

Dude, Rishi - during the course of this debate, you and I have agreed several times; that is all fine. But thankfully I haven't yet put forward a single argument from a position of ignorance. I haven't argued about chemistry, and you're right, I don't know much about Chemistry. I studied it in college and in engineering, but my knowledge is now lost. I know that, I freely admit that, and if you had said something on the basis of Chemistry I'd have backed out simply because my knowledge wasn't sufficient. But what you talk about - especially binary/switching systems and information theory - is my profession. I know enough about it.

 

Sorry, there is no such thing as a binary system that is not a switching system. That is the nature of reality. You can't have ONLY two states of a system and have a process that goes between those states in finite, nonzero time. It has to be zero time, or the system is not binary.

 

You just spouted off some chemistry, but you don't understand that you described an analog system. Unless you could go from 3-4 reduced metallic silver atoms to a macroscopic (billions of silver atoms) silver speck in zero time, the system is analog.

 

You're saying "state of silver atoms = f(development time)". Unless development time = 0, I can halt the reaction at any time and get a state between the defined 0 and 1; so the system is not binary. In the real world, there exist almost no binary systems at all. Even digital circuits approximate binary systems by having their switching time in the picoseconds (10^-12 seconds).

 

If the system is analog, it could result in a speck of silver that is not thick enough to be fully opaque.

 

As simple as that.

[vijay_nebhrajani]

Bernie: Vijay... at the start of this thread, you had credibility. You haven't got much left to play with man.

 

I have no problem with that. You evidently don't understand the nature of scientific debate - the credulity of the person has nothing to do with the credulity of the idea.

 

It is only in politics where discrediting the person is considered equivalent to discrediting the idea. Thanks for sinking this debate to this new low point.

[rishij]

Vijay:<i>"You just spouted off some chemistry, but you don't understand that you described an analog system.

Unless you could go from 3-4 reduced metallic silver atoms to a macroscopic (billions of silver atoms) silver

speck in zero time, the system is analog."</i>

<p>

<b>I just said that grain is not a binary system</b>, so OF COURSE I KNOW I JUST DESCRIBED AN ANALOG SYSTEM! WTF?

<p>

My point was IF the system were set up SUCH THAT with a given development time (presumably pretty long) ALL

grains that have even 1 or more 'latent' sensitivity sites (this is the 'threshold') get fully reduced, whereas

ALL grains with 0 'latent' sensitivity sites are not touched by the reducer (developer) because it's just not

strong enough even with this extended development time, THEN it'd be a binary system. Presumably this might be

accomplished by suitable ratio of developer vs. anti-foggant. But, anyway, we've deduced through

this magnificent thread that this is NOT how development happens, that silver ions are reduced from the

sensitivity sites inwards toward the center of the grain, and depending on development time can be halted at any

step during this rate-controlled process, therefore <b>it's not binary</b>. See? We're arguing the same thing

now, yet you insist on trying to 'win' and make me (us) look like idiots for not understanding binary.

<p>

You yourself say that the digital circuits approximate binary systems by setting up the artificial restraint of a

switching time. I'm saying that if we set up the artificial restraint (in the chemical development process)

described above, then, yes, grain in film

<i>could be</i> binary <b>even though in reality it's not</b>. I feel like I have to spell everything out for

you. Have some faith that I'm not an idiot, man.

<p>

I understand binary. Because I have a brain & can reason. Get over it.

<p>

Rishi

[vijay_nebhrajani]

<i>Rishi: You should read more carefully, Vijay. But I don't expect as much from you.

<p>

That's an 11 MP digital camera, and that's MEDIUM FORMAT film.

<p>

Do a calculation -- how much more information does a frame of medium format film have compared to 35mm? And then

tell me again how you think a 35mm frame of film has 35MP worth of full-tonal range information. </i>

<p>

Based on that article, it is impossible to do a valid comparison about information content - sadly you need the

much maligned resolution charts for that. No image in that article has fine information content near the

resolution limits of the media that can be used; for larger features, both will work fine; and for really small

features, scanner resolution will be the limiting factor.

<p>

Now answer me this: have you ever used medium format? Do you have, even as a rough estimate, some idea of the

information content it is capable of?

[rishij]

Bernie,

 

Though I'm frustrated with Vijay's frequent attempts to prove I don't understand binary, and his reluctance to

actually read my threads fully before lashing out with some dumb rebuttal, I have to say I'm much more in

agreeance with Vijay and the Mark/Ron camp than the 'binary' camp... though I still believe the 'halftone'

process has a large role and though I still completely disagree with Vijay's complete decoupling of 'range of

tones possible per imaging element' and resolution.

 

Rishi

[rishij]

<i>"Do you have, even as a rough estimate, some idea of the information content it is capable of?"</i>

<p>

Yeah, A LOT. Certainly more than that 11MP digital camera. Which is why I largely disagree with the conclusions of that luminous landscape article; HOWEVER, that being said, comparing 35mm film to a 39MP Hasselblad digital is just, plain, retarded. My point in pointing out that article was that if, upon inspecting large prints, a 11MP digital camera can appear as sharp (or sharper) in some areas as a medium format print, then your claim that one needs more megapixel resolution to match the equivalent film (i.e. you need more than 35MP to match 35mm film) is ridiculous.

<p>

Rishi

[vijay_nebhrajani]

Yes, if you could lower the switching time to close to zero, grain would approximate a binary system pretty closely. Of course. Sorry man, you just weren't clear enough. I apologize.

 

So, if you have silver that isn't binary (we all agree, great) why couldn't it be thin enough to pass light through? I mean just those specks that didn't grow big enough? And besides, if it were below some size, wouldn't diffraction would matter too - even though you can't see it as a sharply defined speck, it is there, dispersing light and acting "gray"?

 

Through an optical microscope these should appear as gray areas that can never be brought into focus by turning the knob, right?

[vijay_nebhrajani]

OK, Rishi - try resolving diagonal lines with digital. I mean lines that go from corner to corner in the frame. What do you expect the resolution to be? It should drop by a factor of 1.4 (diagonal of square pixels is 1.4 times the side) right?

 

So to be independent of subject orientation, we have to derate by a factor of 1.4, right? So if you said 21 MP digital (based on pixel count alone), you'd have to derate it to 15 MP anyway, because what if you had detail in the subject that was diagonal?

 

Just asking - or is my analysis wrong?

[vijay_nebhrajani]

Rishi: "comparing 35mm film to a 39MP Hasselblad digital is just, plain, retarded."

 

Yes, it is, but there are other factors at play other than the raw resolution that the medium is capable of. Correct?

 

We are talking about the same pixel count packed into the same area.

[vijay_nebhrajani]

I strongly urge you to look at this article <a href="http://old.photodo.com/art/35_m9.shtml">35 mm, medium

format, or large format?</a> from the old Photodo site. Mercifully it is film vs film in this article.

<p>

Don't underestimate 35mm film.

[bernardwest]

<i>Now there you go saying that 0 multiplied by some number is not 0. If a single 150 nm silver speck can't attenuate light all by itself, even if it were standing alone in a freaking vacuum, then a gazillion of them placed together can't either, much less suddenly absorb all the light energy falling on them. <p>

 

How am I supposed to argue with this kind of absurdity?</i><p>

 

Is it absurd that you can't image an atom (or even the constituents of atoms), but when you clump enough of them together they are visible? Clearly it's not absurd, otherwise you wouldn't see anything in the world around you. Please address this example and tell me how it is different from the 150nm speck example you keep repeating <i>ad naseum</i><p>

 

<i>You relying on EM pictures now? Here's an EM picture of film which shows crystals with spaces less than 400 nm. X Ray film. Should be perfectly opaque then, right? Perfect Faraday cage? Nope. Still can see sun. </i><p>

 

What are you talking about man?!? Have you completely lost your marbles? How the hell would you know what is or isn't visible through this film. So, short of any real evidence, you've resorted to just blatantly <b>making stuff up</b>?<p>

 

<i>It may be complex stuff we are discussing, but it ain't rocket science. I hate to point out to you, but you if you guys need to continue arguing, you'll have to start hitting some textbooks. </i><p>

 

You are an <b>arrogant</b> twat, who has been convincingly proved wrong through both logic and evidence. Your inability to understand even the simplest of concepts is what is holding this debate up from moving forward.<p>

 

<i>and if you had said something on the basis of Chemistry I'd have backed out simply because my knowledge wasn't sufficient.<p>

 

Vijay Nebhrajani , Nov 12, 2008; 09:46 p.m.<p>

 

You could argue about the physical and chemical processes till the cows came home, but you can't escape logic</i><p>

 

<I>You can't have ONLY two states of a system and have a process that goes between those states in finite, nonzero time.</i><p>

 

Explain to us humble information theory ignoramous's how the conversion of a silver ion to a silver atom <b>doesn't</b> occur in (effectively) non-finite zero time.<p>

 

<i>I have no problem with that. You evidently don't understand the nature of scientific debate - the credulity of the person has nothing to do with the credulity of the idea.</i><p>

 

That's the point. At the start of this debate, your ideas and reasoning were just as loopy as they are now, but you had your credibility intact (briefly), that bought you a bit of wiggle room or leeway to introduce your ideas to us. But now, through your bizarre debating style which involves convoluted hypotheses which when able to be deconvolved are seen to hinge on shaky premises and/or made up stuff, you have rapidly eroded any sense of trust we may have had that you might be on to an idea that might take a little time to develop.<p>

 

<i>Bernie, <p>

 

Though I'm frustrated with Vijay's frequent attempts to prove I don't understand binary, and his reluctance to actually read my threads fully before lashing out with some dumb rebuttal, I have to say I'm much more in agreeance with Vijay and the Mark/Ron camp than the 'binary' camp... though I still believe the 'halftone' process has a large role and though I still completely disagree with Vijay's complete decoupling of 'range of tones possible per imaging element' and resolution. <p>

 

Rishi<p>

</i>

 

When finally one of two things happens: Vijay goes away; or the rest of us agree to finally ignore his confused ramblings; then we can get onto your theory. I can't say I fully understand what you are saying, but at this point I am reluctant to get into it while Vijay is still on the sidelines throwing mud at us.

[bernardwest]

By the way, I looked up Alard's 'single molecule microscopy' , and couldn't find a page that explained simply how it worked (at least in terms that I could understsand). I wonder if Alard is still hanging around he could come back and give us a laymans description of how it works? From what I could work out it uses lasers to cause electrons in molecules to flouresce, but no explaination of how this is magnified and translated to an image to be captured by the digital sensor.

[rishij]

Who the heck writes crap like this:

 

http://www.extremetech.com/article2/0,2845,1153635,00.asp

 

"By contrast, the resolution of fine grain 35mm Kodachrome film is about 2,200 lines--per millimeter! That's more than 50 times better raw resolution than digital."

 

Even Kodak & Fuji don't advertise misinformation like that. Am I missing something here? 2,200 lines per mm? That'd be like saying that each grain, on average, is 0.5 microns, and each one of those grains could represent every tone possible.

 

What am I missing here? Or is this article just that wrong? The amount of misinformation out there is appalling. As this very thread & discussion is showing...

 

Rishi

[rishij]

OK, before I attack this all again... what's the AVERAGE grain size of Provia or Velvia... i.e. low speed/high resolution 35mm film? I've seen estimates of 0.1-2 microns... but does anyone have a definitive answer on the average size?

[rishij]

BTW, thanks Bernie, but 'my theory' is really just a combo of what Mark Smith/Ron/Vijay have been saying and the halftone (not binary!)/DLT camp (I don't even know why we're using names... we should just be saying 'analog' or 'halftone' camps).

 

And all I'm really trying to do is get to the bottom of how much of a role the 'analog' tone of a grain (what's left of it anyway) plays vs. how much halftone plays.

 

So, nothing complicated. In fact, I hope I've boiled down my arguments to things anyone can understand... I hate hiding behind technical jargon. In my chemistry explanations, for example, and correct me if I'm wrong -- anyone should be able to understand them if read thoroughly and carefully. The only 'technical jargon' I use is 'reduce' or 'reduction', which ONLY means donation of an electron.

 

I do have a comment yet to make on the whole tonality vs. resolution -- I'm trying to put together some images to prove my point.

 

Rishi

[rishij]

Vijay, ok reading... thanks for the link. I don't underestimate 35mm film -- I love it in fact. Which is why I still shoot it. I'm just trying to approach this argument very objectively. I don't favor one camp or the other. But if I *wanted* one to be true it'd be the entirely analog camp... as this'd justify why I still go about shooting film and wasting years of my life trying to scan them correctly/well/perfectly. Maybe one of these days I'll actually realize that 'perfectly' is a relative term :)

[vijay_nebhrajani]

Hey Rishi - did you see my post about derating digital resolution by a factor of 1.4 (actually sqrt(2) ) because when you try to resolve diagonal lines, the pixel width effectively increases to the diagonal of each pixel?

 

That would make a claimed 21 MP digital sensor only equivalent to about 15 MP if we wanted it to be subject (orientation) independent, correct?

 

What's your opinion about that?

[rich_evans]

Sorry I'm late to the party today - very busy...........

 

A few clarifications:

 

Single Molecule Microscopy.........actually better known as Atomic Force Microscopy, or Scanning Probe

Microscopy. Images obtained from these instruments will not help any arguments postulated here. This is a

technique which uses molecular and/or atomic level repulsive (or attractive) forces to generate a signal as a nano-

tipped probe scans across a surface. This is great if you want to image the atoms on the surface of a grain of

silver, but the grain would be considered 'macro' at this level. Effect of light in this analytical technique is zero.

 

Daniel - you are correct. Its all a matter of scale. I was merely pointing out optical theory on a point scale.

Obviously grains and crystals of the size we're seeing will only have a very minor artifact from diffraction -

depending upon substrate and emulsion thickness.

 

Vijay/Bernie - The wavelength of light is far greater than the spaces between the crystals which make up the

single grain that I imaged with the SEM. No light would be able to penetrate through those inter-crystalline gaps.

The reason an electron microscope is able to image things with the depth, magnification, and resolution that it

does is because the wavelengths of electons are much smaller then light - any light including lasers. The next

level of resolution for obtaining higher level topographical images is a newly developed microscope using Helium

protons to excite samples and generate even higher resolution images............would love to have one of those. --

Rich

[vijay_nebhrajani]

Rich - fair enough. I take back any argument I made based on "single molecule microscopy". My mistake, and apologies.

 

My other point still holds - if the silver deposit is thin enough, light will pass right through the silver, won't it? When it is thin enough, it doesn't matter if there is a filamentary mesh or not - it will let light through anyway, correct?

 

You're the PhD in Optics - could we have a definitive answer please?

[bernardwest]

<i>Vijay/Bernie - The wavelength of light is far greater than the spaces between the crystals which make up the single grain that I imaged with the SEM. No light would be able to penetrate through those inter-crystalline gaps. The reason an electron microscope is able to image things with the depth, magnification, and resolution that it does is because the wavelengths of electons are much smaller then light - any light including lasers.</i><p>

 

Thanks Rich. This is something Daniel and myself have been pointing out continuously in the this thread, but the message wasn't getting through. Since Vijay believes he has 'co-opted' you onto his 'side', perhaps he'll finally take notice of this point now.

[bernardwest]

Vijay... I notice you haven't answered my question from a few posts ago. I'll ask again:

 

Is it absurd that you can't image an atom (or even the constituents of atoms), but when you clump enough of them together they are visible? Clearly it's not absurd, otherwise you wouldn't see anything in the world around you. Please address this example and tell me how it is different from the 150nm speck example you keep repeating ad naseum

[vijay_nebhrajani]

<i>Bernie: "Vijay... I notice you haven't answered my question from a few posts ago. I'll ask again:

<p>

Is it absurd that you can't image an atom (or even the constituents of atoms), but when you clump enough of them

together they are visible? Clearly it's not absurd, otherwise you wouldn't see anything in the world around you.

Please address this example and tell me how it is different from the 150nm speck example you keep repeating ad

naseum"</i>

<p>

Bernie - you are agreeing with me now, how can I argue against you? I kept saying that that 150 nm speck would

matter, regardless of whether or not you could see it under a microscope. Here's the relevant conversation:

<p>

Daniel Lee Taylor [Frequent poster] , Nov 18, 2008; 09:36 a.m.

<p>

<i>"Vijay: That is not what I meant - I've clearly said that there would be specks at different stages of growth

(see Rich's SEM micrographs, on the left there is one tiny speck about 100-150 nm in the 102kx magnification

image). Are you implying that that speck, sitting there, is contributing nothing at all to tone in the final print?

<p>

Daniel: I'm not implying, I'm stating outright. A 150 nm particle sitting by itself in clear gelatin is for all

practical intents and purposes invisible optically and will exert 0 influence on tone, detail, anything. It might

as well not be there at all."</i>

<p>

Then I pointed out the following:

<p>

<i>

"Vijay: If a single 150 nm silver speck can't attenuate light all by itself, even if it were standing alone in a

freaking vacuum, then a gazillion of them placed together can't either, much less suddenly absorb all the light

energy falling on them.

<p>

Daniel: This is not absurd, it's the way the world works. You cannot use light to image and observe a 150 nm

particle because you cannot image something with EM radiation when that something is smaller than the wavelength

of the radiation being used. That means it has no perceptible influence on light between you and a light source.

If it had a perceptible influence then you could see it and we wouldn't need electron microscopes."</i>

<p>

Now I'm confused about what you are saying, Bernie. Are you agreeing with with me or disagreeing with me? Kindly

explain, please.

[bernardwest]

Vijay, take this quote of yours:<p>

 

<i>If a single 150 nm silver speck can't attenuate light all by itself, even if it were standing alone in a freaking vacuum, then a gazillion of them placed together can't either, much less suddenly absorb all the light energy falling on them.</i><p>

 

Take the phrase 'silver speck' and replace it with 'atom'. Apart from the 150nm size, this is exactly the situation with atoms. Individual ones are most certainly invisible to light based measurement, but when you clump a 'gazillion' of them together, you can visually measure them. See what I am saying?

[vijay_nebhrajani]

"Invisible" does not mean that they don't do their part to absorb light (when we talk about opacity) - every single atom is busy absorbing photons; down to the atomic level.

 

So that 150 nm speck, alone, or in clumps of large numbers, will do its part to absorb light, right? Nothing to do with whether you can see it absorbing light or not, right?

 

Yes or no?