Wow - read this re: Film versus Digital debate!

[vijay_nebhrajani]

<blockquote>

<p>Daniel: Individual silver grains or deposits on film are opaque, black.</p>

 

</blockquote>

<p>There were a couple of scans from Baines' book, one showing partially transmissive silver specks that appear gray under a microscope. Comments on this?</p>

<p>There is the argument that underdevelopment produces finer grain, lower contrast and increased tonality. How is this possible with a strictly halftone (silver specs black or absent) process?</p>

 

<blockquote>

<p>They are not like individual pixels, which can record any level of tone.</p>

</blockquote>

<p>That is true and we made this distinction. But if the process weren't purely halftone, it would require far fewer grains to make a tone, thereby giving you realistic resolution numbers with 1:1.6 target contrast, which is really two very close gray tones.</p>

<p>We already agreed that once a crystal becomes a latent image site, there is no predictability or controllability about how the silver forms. It could be a metal film thinner than 100 nm or it could be a sphere. That is the whole point: partially transmissive silver specs are not precluded, and under the right conditions (underdevelopment) they are easily observed under a microscope (Baines).</p>

[rishij]

<p>Yes I have Baines in front of me right now and under certain development conditions/developers, the resulting silver deposits can consist of filaments that are not extremely dense... when I hold the book farther away from me, these areas appear 'gray' in contrast to a darker speck nearby (which contains very dense filaments).</p>

<p>Perhaps underdevelopment results in lower density of filamentous growths. Therefore you'd need more of such filamentous growths stacked on top of one another for a given spot to appear entirely opaque. That would give you more tones representable by a small area of film.</p>

[vijay_nebhrajani]

<blockquote>

<p>If a silver grain or deposit is large enough to see under an optical microscope, then you can bet its thickness is comparable to its height and width, and that all three dimensions measure much greater than 100 nm.</p>

 

</blockquote>

<p>Film manufacturers try to control this by controlling the shape of the original halide crystal; such as T-grain which thin and "spread out".</p>

<p>Ordinarily the process is symmetric in the three dimensions, but when the crystal is surrounded by gelatin, it doesn't necessarily have to be so.</p>

 

<blockquote>

<p>It does not specify behavior just below this boundary, whether or not the metal shifts through a range of grays as it becomes thinner or simply goes transparent (except for reflection at the surface, i.e. the one way mirror example).</p>

</blockquote>

<p>The metal shifts through a range of grays; the absorption of a uniform density of photons per unit solid angle by electrons distributed uniformly in three dimensional space can't be a binary process (going from opacity to transparency at a sudden point) precisely because it is a statistical, probability based process.</p>

[danielleetaylor]

<p><em>Daniel,"DSLRs consistently delivered real world images with just as much detail, resolution, and sharpness as Acros. Sometimes they would even exhibit a bit more detail in areas of low contrast and in shadows"</em></p>

<p><em>Not true.</em></p>

<p>I'm sorry Mauro, but are you actually telling me what I saw in the side-by-side prints I made of scenes I photographed both digitally and on Acros?</p>

<p>Were you there with me while I shot, processed, and made those images? I don't remember you being there.</p>

<p><em>In a different thread I asked you to post the examples you said you had.</em></p>

<p>No, you asked me to post examples from the web and I did. And you simply ignored them as "marketing" when they were not from companies and had nothing to do with the marketing of any product. Your response caused me to doubt that you even bothered to click one link.</p>

<p>A red flag goes up in my mind when I look at your link: why does the 40D shot look so pixelated? 30 inch prints from a 40D aren't pixelated like that. A multitude of other comparisons, all available via the web, don't show such pixelation when comparing DSLRs and film even at great magnification. What exactly are the parameters of this test which yield such horribly pixelated images?</p>

[rishij]

<p>DLT, the 40D images also raised a red flag to me, but they're that pixelated because he used nearest neighbor upscaling. Which is unfair. Someone else posted a result with bicubic upscaling which fared better, but still not as good as the film (especially in the small writing).</p>

<p>DLT if you have any side-by-side comparisons (obviously digital vs. scanned film) of real-world examples, I would love to see them!</p>

<p>Cheers,<br>

Rishi</p>

[rishij]

<blockquote>

<p>...the absorption of a uniform density of photons per unit solid angle by electrons distributed uniformly in three dimensional space can't be a binary process (going from opacity to transparency at a sudden point) precisely because it is a statistical, probability based process.</p>

</blockquote>

<p>Spot on, Vijay.</p>

<p>P.S. Bernie -- looks like there's hope this thread will pass the 1000 posts mark!</p>

[bernardwest]

 

 

<blockquote>

<p>Vijay>>...the absorption of a uniform density of photons per unit solid angle by electrons distributed uniformly in three dimensional space can't be a binary process (going from opacity to transparency at a sudden point) precisely because it is a statistical, probability based process.<br>

P.S. Bernie -- looks like there's hope this thread will pass the 1000 posts mark!<br>

Spot on, Vijay.</p>

 

</blockquote>

<p>Not necessarily spot on in my book. If the transition from clear to opaque occurs from say 0.01 to 0.1 microns, and most silver deposits are microns thick, then it is irrelevant to the discussion. Infact, it is fairly non-sensical because if Vijay insists that there are grey silver deposits in the negative, then this means they must occur at the scale of 0.1 microns. This would give film phenomenal resolution, which it doesn't.</p>

 

<blockquote>

<p>There is the argument that underdevelopment produces finer grain, lower contrast and increased tonality. How is this possible with a strictly halftone (silver specs black or absent) process?</p>

 

</blockquote>

<p>I thought Rishi explained this. I read his explaination to read that it is entirely possible that it could be from a halftone process. The only thing not clear is at what resolution this halftone process occurs.</p>

 

 

[danielleetaylor]

<p><em>There were a couple of scans from Baines' book, one showing partially transmissive silver specks that appear gray under a microscope. Comments on this?</em></p>

<p>Where are these scans? I can't find them in the thread (which is hopelessly long).</p>

<p><em>There is the argument that underdevelopment produces finer grain, lower contrast and increased tonality. How is this possible with a strictly halftone (silver specs black or absent) process?</em></p>

<p>Silver specs are black or absent, but their size is continuously variable.</p>

<p><em>We already agreed that once a crystal becomes a latent image site, there is no predictability or controllability about how the silver forms. It could be a metal film thinner than 100 nm or it could be a sphere. That is the whole point: partially transmissive silver specs are not precluded, and under the right conditions (underdevelopment) they are easily observed under a microscope (Baines).</em></p>

<p>* Is a metal ever partially transmissive? The link I included states that at >100 nm metal is opaque. It doesn't say that it has a range of transparency below that. What is the boundry behavior?</p>

<p>* If you can see a grain, meaning it's much wider and taller than 100 nm, then you can bet it's thicker as well. This is not a process that's going to yield grains which are tall/wide enough to observe visually yet just happen to be only 100 nm thick in the orientation that matters under the microscope or enlarger, i.e. facing them head on. That's just not how it works.</p>

<p><em>Film manufacturers try to control this by controlling the shape of the original halide crystal; such as T-grain which thin and "spread out".</em></p>

<p>And still larger than 100 nm.</p>

<p><em>The metal shifts through a range of grays; the absorption of a uniform density of photons per unit solid angle by electrons distributed uniformly in three dimensional space can't be a binary process (going from opacity to transparency at a sudden point) precisely because it is a statistical, probability based process.</em></p>

<p>And light is also both a wave and a particle, which means we can't predict its behavior based entirely on an understanding of either model. I would like to hear from someone with experience in very thin metal coatings before reaching any conclusion here. The range of thickness and tones between opaque and transparent may be so small as to be effectively binary for this discussion.</p>

<p>100 nm thick grains, eh? Even if they can be shown to sometimes form, they would be as much the result of random chance (crystal size, shape, and orientation) as of light exposure, and would therefore not carry any meaningful data from the original scene. Again I go back to my old line: practically this debate was over long ago. The basis of tone in B&W film is essentially a halftone pattern of present/missing silver. If it were otherwise, we would observe "gray" grains of silver in great quantity in any frame of film with gray tones. To form the basis of gray tones in print these gray grains would have to be found in quantity. If they are not found in quantity (ever? I have to see the Baines image) then by definition they can't be the source of gray tones in prints.</p>

<p>Rishi - <em>the resulting silver deposits can consist of filaments that are not extremely dense... when I hold the book farther away from me, these areas appear 'gray' in contrast to a darker speck nearby</em></p>

<p>Can you scan/post this image? (Fair use.) If filaments can be clearly seen, then isn't this an EM image? And if you have to hold the book at a distance to see gray, then aren't we still talking about a halftone process?</p>

<p>I'm dying to see this image.</p>

<p> </p>

[rishij]

<p>Yes DLT I agree it's a halftone process. I've never disagreed with that. Looser filaments means gaps in between silver filaments through which light can pass.</p>

<p>Yes I'll scan the image.</p>

<p>Also, if absorption is mediated by mobile electrons, and you have X electrons within a given area of film, but you shine 2X photons at this same given area that arrive <em>at the same instant</em> , then, would your surface be able to absorb all those photons or would some go through? What if you shone 10X, or 100X, or 10,000X photons at this same area containing X mobile electrons? How would all those photons still be absorbed at the same instant?</p>

<p>I'll admit I'm at the limits of my knowledge here, so sorry if I sound dumb/naive.</p>

<p>-Rishi</p>

[danielleetaylor]

<p>Rishi,</p>

<p><em>DLT, the 40D images also raised a red flag to me, but they're that pixelated because he used nearest neighbor upscaling. Which is unfair. Someone else posted a result with bicubic upscaling which fared better, but still not as good as the film (especially in the small writing).</em></p>

<p>I see. What is the scale of those images? Put another way, if I made prints from the full frames, at the resolution seen on screen, how big would those prints be?</p>

<p>I wouldn't expect a 40D to do as well as the tested films on high contrast details near/past its Nyquist limit. It has both a smaller frame (APS vs. 35mm) and a lower high contrast resolution. That doesn't mean it can't produce prints at common print sizes with equal or better overall IQ. The point I was trying to get at above is that shooting high contrast targets and then analyzing the smallest details stresses one aspect of an imaging system, and not necessarily the most important aspect for most applications. You can prove that Acros has more resolution than a 10D-40D (not sure about the 50D) with the right target and lighting and magnified studies of the smallest details. But that doesn't mean that prints from Acros of things we normally photograph will be visibly better.</p>

<p>Though it would be if Fuji would just start using gray grains! ;-) Seriously though, I have to get back to work so no more posts from me tonight.</p>

[mauro_franic]

<p>Daniel, I believe Rishi explained the upsampling.</p>

<p>Regarding whether you need that much detail, or can benefit from it using film, it depends;</p>

<p>If you are shooting landscapes and printing larger than 11x14, yes. If printing smaller less so, but you still have more color resolution and dynamic range.</p>

<p>If you are shooting portraits in the studio, with contral lighting and no interest on capturing the skin imperfections, I'd pick digital.</p>

<p>If you want to shoot landscapes and print larger than 16x20, I shoot 6x7 film.</p>

<p>etc...</p>

<p>The difference is so large than the 50D won't make a difference. I am looking for someone to test a 5DII next to 35mm film (and I will also shoot 6x7 film just for kicks for people that back their points with links to the luminous landscape).</p>

<p> </p>

[mauro_franic]

<p>And to quote the fabulous ludicrus landscape; Michael R. himself likes to post other people's opinion to defend his own arguments.<br /> <br /> He quoted this as supporting claims at the bottom of his article comparing the Canon D30 (true 1.8 megapixel resolution) to provia:</p>

<p align="left">A year after the D30's announcement and some 8 months after this article first appeared the debate has died down somewhat, and the supporters are outnumbering the naysayers by a considerable margin. Here's a recent letter from a photographer in the Canary Island.</p>

<p align="left"><em>"Hi Michael, I've come across your article on the D30 versus film, I agree absolutely with all your conclusions, and I can add something. I own a drum scanner capable of 10,000 ppi and 4.6 D and also made tests with Provia 100F, scanning at 5000 ppi (the theoretical limit of film resolution). At that res the grain shows clearly so the D30 prints is vastly superior. The most approximated look to a D30 file is a 4x5 transparency drum scanned to match the size of a print from the D30+GF Pro (30x40 cm). So I too am impressed by the performance of the D30. It seems that all we used to know about photography no longer applies !!" </em></p>

[mauro_franic]

<p>"<em>The most approximated look to a D30 file is a 4x5 transparency drum scanned" </em></p>

<p><em>Sweet momma!! makes me laugh so bad.... <br /> </em><br>

<em><br /> </em></p>

[mauro_franic]

<p>Hence my point, he is either marketing ill intended or worse.</p>

[vijay_nebhrajani]

<blockquote>

<p>Daniel: And light is also both a wave and a particle, which means we can't predict its behavior based entirely on an understanding of either model. I would like to hear from someone with experience in very thin metal coatings before reaching any conclusion here. The range of thickness and tones between opaque and transparent may be so small as to be effectively binary for this discussion.</p>

</blockquote>

<p>Rich Evans already stated that a 20 angstrom film of gold is transparent. A less than 100 nm (1000 angstrom) film is darker, but still not fully opaque. So if there is a threshold for "opacity" it must be equal or greater than 100 nm. The range of thicknesses between opaque and transparent is at least 20 angstroms to (less than) 1000 angstroms; not "small enough to be effectively binary" at all. Not binary means grayscale. Grayscale for a stochastic process translates to analog, and usually linear.</p>

<p>But I'd like to hear Rich's opinion anyway. He's the Ph.D. expert in optics. Rich, we need you.<br>

<img src="file:///tmp/moz-screenshot.jpg" alt="" /></p>

[rishij]

<p>Anyone want to try and answer my question? It seems a very basic one, and if you all claim yourselves to be deep thinkers, answer it. Tackle it. WTF.</p>

<p>Also, Luminous Landscape said this in the D30 vs film article: "It is inescapable that the D30 produces sharper, better looking images than the scanned film combination at sizes up to about 10 X 13". Larger than this 35mm wins, but it isn't till above 11X15" or so that this starts to become obvious. Most lay observers can't see the difference."</p>

<p>WTF?! If he just admitted that at larger sizes 35mm wins, then HOW ON EARTH can he say that at smaller sizes the digital wins?! That's just a down-sizing algorithm problem then! OMG, this is INFURIATING. Now I seriously doubt ANY digital vs. film comparison. It all just depends on the bias of the observer!</p>

<p>This is ridiculous. Hideous. Disgusting. Revolting. Makes me wanna puke. WTF. Mauro, stop me. At least he's honest when he says <em>"I'm a photographer, not a scientist."</em> Well then you know what? <strong>Get lost MOFO.</strong></p>

<p><strong>F'in UNBELIEVABLE, </strong> Luminous Landscape should be <strong>SHUT DOWN</strong> in the interest of science and knowledge<strong>,<br /> </strong><br>

Rishi</p>

[rishij]

<p>Or a sharpening issue! Who knows how much sharpening goes on in-camera? WTF?! OMG. LOL. ROFLMAO. This is ridiculous. What a f'in MORON.</p>

[rishij]

<p>Mauro, I'm so sick of people promulgating misinformation. I am so infuriated right now. You don't understand. We (or, YOU) need to do something about this... it's so wrong that this isn't a SOLVED debate published somewhere in a SCIENTIFIC JOURNAL. $hit I publish scientific findings in my research... why can't we do the same with regard to film resolution?</p>

[vijay_nebhrajani]

<p>Rishi,</p>

<p>Waiting for someone to tackle your question before I chime in.</p>

<p>As for your LL comments, I'll say this: spot on!</p>

<p>Actually, as I browse through websites that claim to be "references" on "imaging", I come across so many factual errors that my own reaction mirrors yours. And everyone seems so smug and secure in their erroneous knowledge that it actually shocks me. Something in me wants to scream out like you do, so I cheer you when you do so. Bravo, Rishi.</p>

<p>That LL statement is actually self contradictory as you pointed out; and at the defense is pathetically weak: "I'm a photographer"? Then please don't comment on something that relates to science and engineering, or as you so aptly put it, "Get lost, mofo."</p>

<p>I've taken the approach that the days of film may be numbered, so there may come a time I may be forced to go digital - so I'll just enjoy film until then. Nothing in my experience or knowledge puts film any inferior than any 20 MP digital camera - and Mauro's tests quite amply demonstrated that (the 26 MP camera test). I've been shooting LF for a while now but am beginning to miss 35mm - especially Kodachrome 25, so I just went ahead and bought some. Speed aside, it is very hard to beat Kodachrome for sharpness and subtlety of colors. Eh, who cares: when digital reaches 35 MP with 20 stops of DR, it will truly have beaten film, so I'll just get that. It appears that something like that could happen inside of five years, and I'm hoping film will last at least until then so that the transition will be seamless.</p>

<p>Once again for your LL comments, Rishi - Bravo!</p>

[vijay_nebhrajani]

<p>I found this on the LL website: <a title="Merklinger" href="http://www.luminous-landscape.com/reviews/merklinger.shtml" title="Merklinger">Harold Merklinger's comments</a> .<br>

<br /> Interestingly he writes: <em>"I generally agree that high quality photographic image is equivalent to perhaps 40 million pixels: about 5000 pixels per inch for a full-frame 35mm image, or 200 pixels per mm."</em><br>

<br /> He actually computes the upper limit for film resolution:<em> "It also turns out that this limit is nearly independent of film format. So 40 megapixels is about it for most pictorial photography."</em><br>

<br /> I hadn't read Merklinger's views until this moment, and I am rather glad that this 40 MP (actually 35 MP or so) number so accurately captures an upper limit. A little validation for me from someone rather well regarded in optics, although I disagree that 40 megapixels is a hard limit for all formats. 8x10 with good lenses easily can get 82.5 MP worth, even if you take 20 cpmm for the lenses which is rather on the low side.<br>

<br /> Then he proceeds to compute the resolution of color film and comes up with one third the linear resolution, or a pixel count that is one ninth: <em>"So, for 35 mm images, effective pixel count is typically on the order of 4.5 megapixels (40 divided by 9) for colour images. Medium and large formats can do better, of course, approximately equalling B&W for 4 by 5 inch colour materials."</em><br>

<br /> Now here I disagree - Royal Gold 25 was around in 2000 when Merklinger's article was written, and so was Kodachrome 25. Their resolution is nowhere near 33 cpmm (maybe some consumer 200 ASA print film only resolves 33 cpmm, who knows?). Today, Ektar 100 resolves far better than 33 cpmm, so does Velvia and many other films.<br>

<br /> Anyway, just thought you folks should read this article. Merklinger is well regarded, and many of his other articles are excellent.</p>

[rishij]

<p>Just to throw a rock into the picture... Vijay, see if my last post on Mauro's thread has any merit:<br>

<a href="00RV7N"><br /> </a><br>

<a href="00RV7N">http://www.photo.net/film-and-processing-forum/00RV7N</a></p>

<p>-Rishi</p>

[mauro_franic]

<p>Rishi, I loved your remarks on the luminuos landscape.</p>

<p>Imaging how upset a person who, after reading Michal R. articles, chose to take a 10MP DSLR instead of his 6x7 Velvia for his once in lifetime trip to Alaska or the Amazon. No that 'd get me really pissed.</p>

[mauro_franic]

<p>And Michael added on the same article, ha ha (laugh or shoot him):<br>

Resolution<br /> <br /> <br /> "The first thing I looked at was, of course, resolution. As can be seen by the detail blow-ups immediately above, the D30 image shows finer detail."<br>

<br /> NOT SO BY A FACTOR OF AT LEAST 10-20 TIMES.<br>

Dynamic Range<br /> <br /> <br /> "In both prints the white tennis dome in the lower right hand corner displays identical clean detailed white. (<em>The prints show this much better than the low-res JPG images shown here; trust me</em> ). Looking at the shadow details it appears to me that there is about a half stop more detail in the D30 image."<br>

<br /> WELL OBVIOUSLY ON THE 14TH SIXPACK WHEN HE THOUGHT OF WRITING THIS.<br>

Colour Saturation<br /> <br /> <br /> "The D30 blows it away in colour accuracy."</p>

<p>NOT BY A 100 TIMES STRETCH.</p>

[mauro_franic]

<p>A note about genuine fractals;<br>

<br /> GFs are quite wonders of nature and very good friends with Michael R. GFs are little creatures that place hundreds of pixels in between any two pixels and the fly away to the scene of the shot;<br /> they ask the scene:<br>

<br /> GF: "Scene, can you tell me what was in between the 2 pixels I have"<br>

<br /> Scene: and the scene responds "Yes GF, there were 2 trees, a boat, dozens of yellow, green and red leaves,<br>

and bird. Oh oh and a crack on the second tree."<br>

<br /> GFs then fly back to your computer instantly, -since they are powerful magical creatures- place the information on the image, and repeat until the upsampling is completed.<br>

<br /> Then you can print your 16x20, 20x30, 30x40 and 44x60 using your DSLR.</p>

[rich_evans]

<p>Sorry guys - I've been away. Give me a day or two to cach up on the latest since my last post and I'll get back to you ;-) --Rich</p>