Wow - read this re: Film versus Digital debate!

[pearse s]

I suppose the whole point of photography is to take photographs that work. I know that here in Dublin, Ireland,

more and more professional photographers, especially those doing weddings and 'people' photography, are going

back to film. They find that digital does not give them the satisfactory images that their old film equipment did

years ago. They've invested heavily in digital, yet they and their customers know that film is best, just by

looking at the finished image.

 

Little else matters, and that includes convenience, speed, numbers of shots per day, pixel size, etc.

 

Professional photographers have to keep the customer happy. With so much competition out there, the one who has

the edge is the one who gets the contracts. Film can give a photographer the edge, now that "everyone has gone

digital".

 

Also, now that amateurs have access to good digital cameras at a reasonable price, medium format film photography

gives professionals a real edge in terms of quality and permanence, over for example, guests at a wedding.

 

I remember when some relations of mine were professional wedding photographers in the 1960s, they'd do a wedding

with a Leica M-something, and shoot one or two rolls of film (at most) and be able to return a couple of hours

later when the guests were still eating, with perhaps two dozen excellent shots that recorded the most important

moments in great detail. We still have some of these albums and they are as perfect now, as they were then.

 

By contrast, I was at a wedding recently. The pro photographer eventually put all 750 (!!) of his shots up on his

site for people to order from. I must say that he probably had fewer than two dozen excellent images, none of

which are likely to still exist in 40 years.

 

I do own good digital cameras, but whenever I use them for important (amateur) work I always feel I could have

done better with one roll through my trusty Mamiya TLR, Mamiya Press, or Speed Graphic.

[philip_sutton]

As already suggested here - wouldn't it be great if we all lived in close proximity and we could meet up over a few

cold ones!!

 

However, I can tell by the time of your submissions that most, if not all of you are in the Northern hemisphere -

perhaps I am the only Antipodean in our midst.

 

It is sweltering heat here in the desert in Western Australia - something cold would go down well.

 

Thanks for chiming in Petrana, great to have a lady in our midst and perhaps you can sort some of these guys out on

the merits of M/F L/F film - LOL

 

Phil

[rishij]

Vijay, I don't understand what you wrote here:

 

"Rishi: "To compute the max resolution of film? No way, that would make it >10 gigapixels for a 35mm frame."

 

No, to find what you say is "an area large enough that it contains a sufficient number of grains that can give rise to a sufficient number of filamentous growths to represent thousands of tones"."

 

Please clarify.

 

Also, when a material has gaps that are considerably smaller than the wavelength of light, most of that light is scattered and/or refracted, and in some cases reflected, but more rarely transmitted. Don't know if that even matters at this point in this argument, since that part of the argument is over, for me anyway.

 

Also, I live in Seattle WA. Anyone else?

 

Cheers,

Rishi

 

P.S. Vijay, don't worry, we love you (well I won't speak for Petrana... you'll have to settle your quibbles, uh, privately). Every good argument needs a devil's advocate. And you, being the devil's advocate of like almost every argument in this thread (for me, starting with resonance/vibration of the bridge -- plausible, but unlikely given closer analysis of more images... yes I haven't followed up by posting back to this thread about it yet), are going to take lots of heat. Including irking others, like myself, from time to time. When we think that you're just being obtuse & stubborn. But no love lost. :)

[vijay_nebhrajani]

Daniel: <i>If they're too small to be observed under an optical microscope, they're too small to make any

difference in the perception of detail or tone in a final print at much lower magnifications.</i>

<p>

And the evidence of that is? Because:

<p>

Statement x: they're too small to be observed under an optical microscope

<p>

Does not necessarily imply

<p>

Statement y: they're too small to make any

difference in the perception of detail or tone in a final print at much lower magnifications

<p>

Statement y implies that the final print has tone. But tone can be created even if the particles are too small to

observe - for instance with pigments. The individual "particles" of the pigment are far too small to be observable

under an optical microscope, but they form tone just fine.

<p>

Thus statement x does not necessarily imply statement y.

<p>

Here is a reference: <a href="http://www.beckmancoulter.com/Literature/BioResearch/ta-401.pdf">"A Colourful Tale –

The Particle Size Analysis of Pigments"</a>

<p>

The particle size of these pigments is in the order of 100 nm, and I quote from the reference <i>"The size of the

particles (determined by SEM) is 46.9 nm in width and 130.8 nm in length, giving them an approximate aspect ratio

of 3:1."</i>

<p>

You can see the SEM pictures there as well.

<p>

When making these arguments you tacitly assume that silver particles are opaque. You can't put forward that

assumption as true because it is that very assumption that we are trying to test. <i>This</i> is an example of a

circular argument that led to this particular false assertion.

<p>

<i>So it might be an interesting question for a physicist (how does matter interact with light at the size

threshold, whatever it is for a given material, between transparent and opaque), and I would even be curious as

to the answer. </i>

<p>

But the answer is key to proving your argument. You gloss over this point as if it is too insignificant to care

about, yet this could be giving you "gray" specks of silver and you not knowing it.

<p>

<i>But it doesn't matter as far as this thread is concerned.</i>

<p>

Yes, it does. This is like saying that you can't see pigment "particles" under a microscope so pigments can't

possibly form tone on paper.

<p>

<b>Digression Alert: Ignore if you want</b>

<p>

The existence of most atomic and subatomic particles was postulated before they were observed.

In fact, theoretical physics and mathematics was and continues to be used to predict their properties and of

other particles, long before these particles are even found.

<p>

The direction that scientific enquiry takes is that you make a hypothesis first, then you attempt to refine or

reject based on observable data (a formal reductio ad absurdum) - then you test by allowing the hypothesis to

make a prediction (generally unrelated to the original observation, and something easily observable with existing

methods) and then test the hypothesis by seeing if the predicted result happens when you do that experiment. If

it does, the thing gets elevated to the level of theory which is then accepted as true until such time as it is

proved false by newer data, newer methods or when a prediction it makes fails etc. (or never proved false; whatever).

<p>

You and I are stuck at the first two levels. You, Reichmann and Adams made an observation, and came up with a

hypothesis that silver specks are entirely opaque to visible light. I am providing the refine/reject arguments

(the reductio ad absurdum) for this. If your hypothesis is to be accepted at the level of theory you have to at

least defend against my arguments. You can't say "its irrelevant" or "it does not matter" especially when you get

a potential argument that may explain your initial observation differently or even question the observation itself.

<p>

If you want me to back off from this debate, I will, but then you leave your observation at this level. "Under an

optical microscope we see opaque specks of silver. We don't know whether these are all that affect tone or there

are other phenomena that may affect tone that are not observable with an optical microscope." Don't for a moment

assume that statement of an observation amounts to even a hypothesis, much less a theory, and much less the truth.

<p>

Don't confuse all this with me making a hypothesis that grains <i>are</i> gray. They are gray for me until proven

otherwise. Why? Because there is nothing other than that one observation that says so, the hypothesis put

forward for that observation is still being tested for correctness and there is plenty of other logical evidence

that still points to the possibility of grains being not wholly transparent and not wholly opaque.

 

[vijay_nebhrajani]

"Rishi: "To compute the max resolution of film? No way, that would make it >10 gigapixels for a 35mm frame."

 

Vijay: "No, to find what you say is "an area large enough that it contains a sufficient number of grains that can

give rise to a sufficient number of filamentous growths to represent thousands of tones"."

 

Well, lets agree that to calculate resolution, we need to find a "pixel" equivalent for film. If you postulate

that this should be an area A that can at least have x tones, then we need to compute A, or find a function A = f(x).

 

Now, x tones is a function of how many silver specks y are present. So x = g(y).

 

Thus A = f(g(y))

 

Resolution in cpmm would be defined by:

 

R = 1000/(2 * sqrt(A))

 

when A were in microns x microns.

 

I already defined function g() in a previous post. Here it is again:

 

With only opaque silver specks, the number of tones you can form with n specks is n+1.

 

This is because with say: four specks you can have tones made by zero specks, one speck, two specks, three specks

or all specks, corresponding to white, 25% gray, 50% gray, 75% gray and black. This is not like binary bits -

where 4 bits form 16 values. In our case 2 out of 4 specks being opaque results in 50% black, regardless of

which specks were opaque.

 

Thus, to form say 65,536 tones (the equivalent of a 16-bit depth digital pixel) you would need 65,536 - 1 =

65,535 specks.

 

If these silver specks were only 200 nm x 200 nm (below which the specks become unobservable under an optical

microscope and can't contribute to tone, per Daniel) then you would need an area corresponding to 256 x 256 specks

 

(256 is the square root of 65,536, so you'd need an area of 256 x 256 specks)

 

256 x 200nm (length) x 256 x 200 nm (width)

 

or 51.2 microns x 51.2 microns

 

To make a line pair you need two such areas, or say 100 microns. This gives you the resolution of film as

1000/100 or 10 cpmm.

 

If you, instead of taking the silver speck size, decide to take the crystal size, say 2 microns, this would

compute to 1 cpmm, or 1 line pair per mm. This is why it is so important to stick to one definition of "grain".

 

So, if you use silver specks, these could be subgranular, given that the standard definition for grain is the size of

the AgBr crystal.

 

So in the volume occupied by (or cavity left behind by) an AgBr crystal (a grain per standard definition) there

could be many residual subgranular silver specks. When I mentioned halftone grain long ago - this is exactly what

I meant. This would mean, in a sense that a grain (standard definition - one crystal of AgBr) represents multiple

tones and is therefore not binary. How many tones could a it represent? I don't know.

 

There is another parallel discussion about whether a single silver speck is entirely opaque or entirely

transparent. My initial arguments were not directed at this at all. They were all about the previous issue, vis a

vis resolution of film.

 

Now I have, through some convoluted process become part of a discussion about whether a single speck can be

anything other than completely opaque - meaning it is present but it does not block 100% of the light falling on

it. While this was not part of the initial argument, now that it is, my stand is that it is gray until proven

black - or that unless it can be proven that it is entirely opaque, I'll go with the generalized case that it is

an optical attenuator.

 

Fair?

 

By the way, this calculation is based on silver specks that are completely opaque. If they were not to be so,

these computations would have to be revised, because you could now represent tones with fewer specks.

[vijay_nebhrajani]

Rishi: "Also, I live in Seattle WA. Anyone else?"

 

Silicon Valley.

 

Rishi: "P.S. Vijay, don't worry, we love you (well I won't speak for Petrana... you'll have to settle your quibbles, uh, privately). Every good argument needs a devil's advocate. And you, being the devil's advocate of like almost every argument in this thread (for me, starting with resonance/vibration of the bridge -- plausible, but unlikely given closer analysis of more images... yes I haven't followed up by posting back to this thread about it yet), are going to take lots of heat. Including irking others, like myself, from time to time. When we think that you're just being obtuse & stubborn. But no love lost. :) "

 

Good that you expressed your love. Hopefully Petrana will do the same. Can't we all just love each other? Thank you, I enjoy playing devil's advocate. But I'm not, um, opaque. I see the light when it's there. So just let me continue in this role for a while longer without telling me how err, dense I am. If there is light, make me see it.

 

P.S. Love puns. (Take a picture of 24 carrots arranged in a circle: 24 carrot ring; take a picture of a saxphone and two violins on a television: Sax and Violins on TV - both from a 1980's issue of Photographic World, an Aussie photo magazine, an article called "Punography")

[rishij]

Vijay, you've forgotten to take into your entire analysis the existence of more than one layer of grains/film. Taking this into account dramatically reduces your overall estimate of 'area required to resolve 2^16 tones'. Which then increases your max resolution of film from 10 cpmm to something much more reasonable, like 60 cpmm.

 

Certainly not 200 cpmm like you were suggesting in some earlier post, but we all knew, even back then, that that was a ridiculously high resolution for 35mm film.

 

Rishi

[rishij]

Haha Vijay! Good puns! Loved 'em :)

 

Now, the more important question -- if you're in CA, you're in the same time zone as me. Which places you at 4:43 a.m. WTF are you awake?? I'm awake b/c, as per your previous post, I was already drunk at 10 p.m. But you? You better have a good excuse! And it better not have anything to do with film vs. digital :)

 

Rishi

[vijay_nebhrajani]

Rishi: "Vijay, you've forgotten to take into your entire analysis the existence of more than one layer of

grains/film. Taking this into account dramatically reduces your overall estimate of 'area required to resolve

2^16 tones'. Which then increases your max resolution of film from 10 cpmm to something much more reasonable,

like 60 cpmm. "

 

Not with opaque specks. An opaque speck occludes all other specks in above it or below it, resulting in nothing

but a black speck. (Remember the binary AND operation of long ago?)

 

A clear area therefore implies not a single opaque speck of silver anywhere in the entire thickness of the

emulsion. Otherwise that area wouldn't be clear. That this is an impossibility (or at least very low probability)

given that the light itself doesn't have bright and dark areas - that not even one speck forms and leave the

entire area clear - but we will leave that alone for now.

 

So those specks must be in the X-Y dimensions. My analysis is correct for opaque specks. You would be right if

the specks could be transmissive, i.e., had tone.

[vijay_nebhrajani]

Rishi: "I'm awake b/c, as per your previous post, I was already drunk at 10 p.m. But you? You better have a good excuse!"

 

I have the perfect excuse.

 

Kids.

[rishij]

Vijay: <i>'Kids'</i>

<p>

Petrana, y'hear that?? Strings attached!

<p>

:)

<p>

Vijay, as for your comment: <i>"Not with opaque specks. An opaque speck occludes all other specks in above it or below it, resulting in nothing but a black speck. (Remember the binary AND operation of long ago?)"</i>

<p>

That's it. You and all your opaque specks can go to... I don't believe in opaque specks as per the rigorous definition of 'opaque' because I don't believe <b>black holes</b> exist on the film I have stored in my gosh darn fridge. I can still eat my eggs tomorrow morning, making it highly unlikely that black holes exist in my fridge. Yet my developed film resides in my fridge. Both statements being true is a logical impossibility.

<p>

That's your kinda language, ain't it Vijay?

<p>

:)

<p>

Rishi

[vijay_nebhrajani]

Rishi: "Certainly not 200 cpmm like you were suggesting in some earlier post, but we all knew, even back then, that that was a ridiculously high resolution for 35mm film."

 

Look back. I suggested putting 100 cpmm as the upper bound on film - under the best of best cases. This implies 35 MP as an upper bound; i.e., you could never achieve more than 35 MP with film for a 35mm area. Realistically I'd expect something in the 60 cpmm as well; but with the right conditions - like high contrast resolution targets, stable tripods etc, the upper bound (for the sake of coming up with a limit) would be 100 cpmm.

 

I also said repeatedly that I'm glad that the debate of film and digital is largely over with 21 MP cameras and we can focus now on increasing sensor area etc.

[vijay_nebhrajani]

Rishi: "That's it. You and all your opaque specks can go to... I don't believe in opaque specks as per the

rigorous definition of 'opaque' because I don't believe black holes exist on the film I have stored in my gosh

darn fridge. I can still eat my eggs tomorrow morning, making it highly unlikely that black holes exist in my

fridge. Yet my developed film resides in my fridge. Both statements being true is a logical impossibility. "

 

What are you talking about man?

 

I was just referring to opacity - the ability of materials to absorb all light falling on them - not about light

not being able to escape the gravitation of a black hole because it distorts the space-time continuum enough.

 

Shine the (visible range) light of ten thousand suns on the outside of your fridge - see if your film gets

exposed. That's opacity.

 

Everyone seems to agree that a speck of silver (subgranular or whatever) is solid silver, and will be opaque,

meaning absorb light. So a clear area of processed film can't have any speck of silver anywhere in the Z

dimension now, can it?

 

Also you can make light gray + light gray = dark gray, but not black + black = any shade of gray. If the silver

specks are opaque, a clear section of film has no silver specks, and an opaque area of film could have any

number, stacked on top of each other even, but always resulting in full opacity.

 

If you want stacked silver specks to be able to make multiple tones, then they have to be transmissive - dead

against the observations of Daniel, Reichmann and Adams. I didn't say opaque specks at all, you guys did.

 

Get it?

 

Rishi: "That's your kinda language, ain't it Vijay?

 

:) "

 

Yeah, but you can't use reductio ad absurdum to prove an absurdity. It is used to disprove it.

[vijay_nebhrajani]

Oh and Rishi, don't forget to do a similar analysis for chromogenic (C41) B&W. This is presumably made of dye

clouds a bit larger than grains in conventional silver halide film. Say 5 microns x 5 microns at most for a dye

cloud. But there is an important difference - a single dye cloud can represent all 64K tones or maybe even more,

so is effectively a complete pixel. So chromogenic film must be able to resolve like 100 lpmm, or like ten times

the resolution of conventional silver halide!

 

Right?

 

(This is my kind of language. Good buddy RAA.)

[vijay_nebhrajani]

Bernie: "Vijay.. why are you continually putting forth hypothetical grain arrangements and sizes and claiming this is somehow related to the reality of film grains? If film was constructed of such a hypothetical arrangement, it wouldn't function as film as we have come to know it. Whatever your line of reasoning is, it is vague. Please state it more clearly so we can have a chance at refuting it."

 

Daniel put forth the argument of a Faraday cage - the sheet of metal with holes in your microwave oven that lets you see your coffee heating up but prevents the microwaves from coming out. He invoked this to "prove" that a mesh formed by silver filaments can't possibly pass light. His argument was that the holes were roughly the same size as the wavelength of light and so light wouldn't get through. I was pointing out that such holes could occur in film as gaps between different silver deposits from different grains, not necessarily from filaments within the same crystal, and you'd have the same effect - the film would go opaque - as a matter of course.

 

"Now, I see some form of reasoning in your examples of thin strips of gold and diamonds. Clearly if you clump enough gold molecules together they will become opaque (or at least effectively opaque). So why doesn't silver act the same way. Well according to some stuff I googled earlier (but don't have the links anymore) silver can also be used in the same way. Thin films of silver are used in devices for viewing solar eclipses. ie. It isn't fully opaque. It functions more or less the same way as gold, but is of course cheaper (although performance wise it isn't up to the scratch of gold, and therefore the use of gold in space). So it is a fair point that you raise, and i'm not sure what the answer is. But I suspect, as Rishi, DLT and myself have been arguing all along, it doesn't matter. All that matters is what is visible at human perceivable magnifications."

 

My point being that as the silver deposit is being formed by the developer - it is increasing in size from a few atoms to possibly the size of the whole crystal, so as it was growing, it would be nearly transparent in the beginning, and as more silver atoms continued to get added, it would get thicker and more opaque, until finally it had gotten completely opaque. Since development could be halted at any time during this process, you could wind up with a speck of silver that was not completely opaque.

 

Now when you want to print that image, that very same speck of silver that was not completely opaque would absorb some of the enlargers light, contributing to tone. It wouldn't matter whether this speck is visible even at "human perceivable magnifications" (not sure about that, but I'll assume through an optical microscope) - but as long as it absorbed some of the enlargers light it contributed to tone.

[vijay_nebhrajani]

Oh and if only particulate matter that is visible at "human perceivable magnifications" could contribute to tone, then pigments wouldn't work at all - see the reference I quoted earlier (http://www.beckmancoulter.com/Literature/BioResearch/ta-401.pdf) - pigment particles are in the order of 100 nanometers, perhaps invisible at "human perceivable magnifications" but they form tone just fine.

[rich_evans]

Rishi - Thanks! ;-)

 

Daniel - screw the beer, make mine a Belvidere, double, on the rocks, with lemon.........

 

Good morning guys (and ladies........) - here in NJ its raining and windy, leaves finally falling off all the trees (has

not been a colorful fall this year, ergo only one photo........oh yeah - its digital.

 

--Rich<div></div>

[galan]

I think Fredrik may have been a little harsh using 'crackpot' to describe Ken 'not worth comment' Rockwell. The word that comes to mind for me after reading his babblings is 'idiot'

Although I use both digital and film (Nikon), I can see differences between the two which is precisely why I have opted to do both.

I have purchased (my bad) one of Ken's books on photography in my earlier days. About 1/2 of the way through it I realized I was wasting my time with this effort to associate reality with his banterings about how to's and seeing the images he provided as 'examples' was a wake-up call. He's an idiot who is immersed in his own importance.

 

Nuff said.

[don_e]

Ok. I'm exhausted and can't read this anymore, no matter how much I'd like to. When it is all over and done with, maybe the moderator can post a summary for us weaklings. I want to thank you all for a most entertaining and interesting discussion, but life is too too short...

[s_u]

It's unfortunate that this thread quickly descended into technobabble and other such minutiae; Mr. Rockwell's broader points seem to have become lost in a chemical thicket. I applaud, and quite agree with, his principal points regarding the general superiority of film. I particularly appreciate his offhand comment to the effect that the fanatically digital-minded are victims of corporate greed and manipulation, whether they realize it or no. I have offered the challenge here on previous occasions to someone of perhaps investigative bent to explore the full story of how digital was foisted on the mass consumer by corporate market manipulation. I'm considering taking on that challenge myself. I suspect anyone seriously undertaking this task would uncover a typically sordid fable of corporate greed, cynicism, clever manipulation of the market, the generation of "artificial demand," and so on. I further conjecture that the whole story would buttress Herr Marx's observations regarding the inherent tendency of capitalist economies to generate useless and frivolous products.

 

As for efforts to attack Mr. Rockwell's credibility by mindlessly hurling epithets such as "idiot" at the man, I have found that tactic appallingly common among "digiheads." I have visited Mr. Rockwell's site a number of times and have found the information he provides quite useful, most notably in the area of lens testing and evaluation. The spewing of such terms as "idiot" and "moron" in the context of this increasingly tiresome debate seems to flow mostly from the self-deluded "digiheads," who often are, I imagine, cybergeeks of the Gen X-through-Zero demographic for whom each day is an entire new world.

[02Pete]

Some of this thread really goes against the grain.

[benny_spinoza]

When reading these posts, take them with a grain of salt.

[James G. Dainis]

<I>"...maybe the moderator can post a summary..."</I><P>

 

Certainly. In short, after careful and diligent study and analysis of all the posts, I can state quite assuredly

and without fear of contradiction that film grain may or may not be opaque. <P>

 

I'm hanging in there in the hope that they can actually prove that film is binary. That will lay to rest forever

the film vs. digital debates since it will be proved to be all the same anyway.

[vijay_nebhrajani]

Shane - I tend to agree with most of what you (and Rockwell) pointed out; corporate greed and such.

 

I work in the semiconductor industry - I design chips for a living; and if I were to cheer for a side, it would

be digital. After all, film doesn't pay my bills, and if the whole world were to buy new digital cameras every

year, why that would keep me employed for a long long time, perhaps even make me rich.

 

I suppose the digital folks had their shills testing out 3 Megapixel cameras and declaring them superior to

medium format just a few years ago but I never bought into the hype. Besides, after sitting in front of a

computer at work all day, I didn't want to do that again just to post process images.

 

But today, with 35mm sized sensors approaching 21 MP, the gap has all but disappeared. Even Ken Rockwell's

controversial photo of trees with film and his D3 shows relatively minor differences that would likely be

minimized with his Canon 5DmkII when he repeats the same experiment.

 

Film probably won't go away soon - there are many applications for film yet - especially in the larger formats,

but for smaller formats, digital will be an equal option - shoot, upload the pictures to Costco, go pick up the

prints at your convenience. (With film, shoot, drop off the pictures at Costco, pick up the prints.)

 

Digital has other problems that will never go away - such as storage. You need to have reliable digital storage,

and for anything other than modest personal use, this means RAID servers, automatic backups, remote backups, data

security, high speed networking, and a whole lot of other solutions (read expenses). You have to have accurate

color, so think profiling, expensive monitors etc (read expenses).

 

But like I said, for the purposes of 35mm film the quality from digital has finally arrived, and the debate can

switch to other things - like the optics - a la Nikon vs Canon - or sensor size - a la Leica vs Hasselblad.

 

Mercifully.

[vijay_nebhrajani]

James: "In short, after careful and diligent study and analysis of all the posts, I can state quite assuredly and

without fear of contradiction that film grain may or may not be opaque."

 

Blasphemy! You doubted Reichmann and maybe, just maybe you could be forgiven for that, but more importantly you

doubted Ansel Adams! Gasp!