bernardwest Posted February 22, 2009 Share Posted February 22, 2009 At the risk of being labeled a 'digihead', I would like to point a couple of things out. Nikons raw files are known to be not exactly 'raw'. There is known to be some level of noise reduction being performed on the real raw data before this data is presented as 'raw'. This is in reference to Les' map shot. I'm not sure if this pre-'raw' noise reduction is performed on the D2 (or whatever it was), or whether this would even account for the difference, but it is worth perhaps investigating further. On the point of the map shots Les, can you post a raw of the dslr shot? In relation to Mauro's 40D vs film shots of the containers of food, there are a few issues here. One clearly is the pixelated 40D image. This problem has been pointed out in other threads with this image, and Mauro continues to trot out this image as some sort of conclusive proof. This is the most disingenuous comparison I have ever seen. Factually, you might be correct Mauro, but these images certainly don't show it. As Daniel (and many others in the other threads) said, your method for upsampling was flawed. The other serious problem with this image is that it isn't a shot of the same thing! The can in the second image is not the same as the first image. What's going on? Now, like I said, you may be factually right, but this image doesn't show it. In fact, all it appears to show is that you have some kind of agenda in pushing film over digital. Your 'tests' that you keep trotting out as conclusive proof have had flaws pointed out many times in other threads. When this happens, the response from you or other film fanatics is usually along the lines of 'this is just a simple test as best as one man could do, and you should do your own tests to make up your mind'. While that is good advice, the problem is that these 'results' appear in other threads and get promulgated as the highest truth. And on a related point, and one that is a reflection on comments higher in the thread about poor scanning methodology being the reason for poor film results against digital results, the same goes for raw processing. Unless you are a skilled raw processor and actually understand what goes on behind the scenes, then you have about as much credibility as someone who is a poor film scanner. Be careful when you want to claim the higher ground. Link to comment Share on other sites More sharing options...
don_e Posted February 22, 2009 Share Posted February 22, 2009 <p>"I would like to point a couple of things out. Nikons raw files are known to be not exactly 'raw'."</p> <p>I recall a thread here in which the op was a forensic photographer shooting fingerprint files with complaints about Nikon raws.</p> <p>As for the "tests", describing a bit about the criteria and showing a 1/4mp jpeg as 'proof' will convince no one but the already convinced.</p> <p>And for the claims of engineer-for-decades-this-and-that, it is irrelevant if your testing criteria implies otherwise.</p> <p>Folks, this is not a difficult test to set up. I suggest the film and digital sides get together and each pop 20 bucks into the kitty, agree to critria, and run the test.<br> I also recommend not merely comparing scanned film to digital camera file, but record the digital file to the same film used in the test and compare that, too.</p> <p>I think Todd complained about the "tests" not having natural subjects. He may want to browse Ken Rockwell's site for hover-over comparisons of snaps taken with various film and digital cameras. He won't like the results, though. But, if real world tests convince you, you're in luck, because good film cameras are still inexpensive.</p> <p> </p> Link to comment Share on other sites More sharing options...
lilserenity Posted February 22, 2009 Share Posted February 22, 2009 <p>Returning to the original question, as a film shooter who is very aware of the results from digital RAW (in general) yes there is a difference.<br> I'm not going to go into this grain and resolution discussion but the difference I find varies depending on what you've shot with regard to film. Is it print or slide film? If it's print, I can't really describe the difference except that the colours have a slightly more earthy appearance and the shadows seem to have a little more detail which at first appearance makes the scan look rather dark, even underexposed! With slide, I find the colours just pop out in the area of focus, there's a very 3D look to them. Slides (K14 or E6) scan best, C41 next and I would honestly say that B&W scans are just not as good as wet prints. The scans on B+W seem to emphasise the grain compared to a wet print.<br> The best way? Well I don't know. Send it somewhere reputable, e.g. here in the UK I'd use Peak Imaging if I needed a crucial shot digitised, I often don't bother with mini labs as they save in JPEG rather than TIFF. Also I find minilabs don't always look after their scanners and I have had cases where consistency just isn't there.<br> Overall your result is going to depend a lot on the film originally used.<br> In any instance, I still use 35mm and 120 (for 6x6) film, C41, B&W and E6 and when scanned, it can look very good. I still maintain the best way to view a photo is projected or printed (be that digitally or traditionally.)<br> I'm not even going to get into the megapixels and resolution, sharpness and whateverness as really, day to day, if you're a good photographer you'll get the results you want if you have a half decent digital or film camera.</p> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p><a href="../photodb/user?user_id=862488">Michael Ferron</a> ,</p> <p>Do you want me to send you a piece of tmax with a resolution chart, do you have a microscope to see how much your scanner is leaving behind? Just let me know.</p> <p>This is a comment from Rishi Sanyal who I sent strips of different films with resolution charts:<br> "With the TMX & the Imacon, I can see past 5. Maybe even 5.5. Meaning ~34 megapixels for the entire frame. Ridiculous."</p> Link to comment Share on other sites More sharing options...
bernardwest Posted February 22, 2009 Share Posted February 22, 2009 Les.... did you see my post? Any comments on the nikon raw issue? More specifically, do you have the raw file that we could have a play with? After my last post I realised I forgot to mention something else strange about Les' map shots. The D2 shot is missing Lake Chad! I don't know what is going on there, but this definitely needs to be investigated further. Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Daniel, "If the 60mm macro outresolves the 40D at f/8 then photozone should not be able to detect a resolution difference on that lens between f/4, f/5.6/ and f/8, with a lesser 8 megapixel Canon camera. But they can: <a rel="nofollow" href="http://www.photozone.de/canon-eos/162-canon-ef-s-60mm-f28-usm-macro-test-report--review?start=1" target="_blank">(link)</a> "</p> <p>Daniel, once again, you have good intentions put can't pick up the camera to test yourself. The Rebel used in that test resolves only 1850 lpph, how they came up with 2166 beats me. It is probably their not understanding the true resolution of the camera and making guesses.</p> <p>It also seems like you missed my study of impact on diffraction:<br> Here is the extract pertaining to the 40 + 60mm macro (as usual shot by me):</p> <p>Link:<br> http://shutterclick.smugmug.com/gallery/6302153_PLzKe#397750363_uGbHQ-O-LB</p><div></div> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Bernie, "In relation to Mauro's 40D vs film shots of the containers of food, there are a few issues here. One clearly is the pixelated 40D image."<br> Yes, that is a problem with digital. They are pixelated ha ha.</p> <p>Scroll up and pick the raws I made public to everyone if you have a question about the quality of the shot. You can verify it extinguished the resolution of the camera by the chart placed next to it.</p> <p>About the cans being different flavors, blame that the fetuccini from the day before.</p> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p><a href="../photodb/user?user_id=862488">Michael Ferron</a> ,<br> Please clarify something to me, for the record, because I am confused. </p> <p>Two direct questions,</p> <p>"1- Do you think your 3200 dpi scanner extract all the info from TMX?"<br> "2- Do you think there is a digital camera out there that outresolves TMX 100 35mm?"</p> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Daniel, "Mauro should change it because otherwise it looks like he purposely manipulated the test to make things look worse. If he cares about the integrity of his test, he should care about the post processing differences."</p> <p>I gave you all the raws for every single test, you are welcome to post your comparisons and your conclusions about resolution. Give it your best shot at making detail appear in the digital file.</p> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Michael, how do you develop your tmax? do you mind posting an example of your resolution test also?</p> Link to comment Share on other sites More sharing options...
bernardwest Posted February 22, 2009 Share Posted February 22, 2009 <blockquote>Bernie, "In relation to Mauro's 40D vs film shots of the containers of food, there are a few issues here. One clearly is the pixelated 40D image."<p> Yes, that is a problem with digital. They are pixelated ha ha.</blockquote><P> No, not ha ha. Show me anywhere where any half competent digital shooter presents images <b>that</b> pixelated. You can't because, this isn't how digital works. <p> <blockquote>About the cans being different flavors, blame that the fetuccini from the day before.</blockquote><p> The problem is that the label you are trying to read isn't the same size on your digi shot. Now I doubt very much that that would account for the worse digital results, but it does show that you have a poor testing methodology and shouldn't use these sort of results as some sort of conclusive proof like you and your worshipers do. Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Bernie, give it the best shot at making up more detail using my raw and post it.</p> <p>The reason I upsample linerly and not bicubic is because I am comparing detail. Bicubic makes up information. With bicubic, you could end up with a good looking word saying "TARDPES" when in real life it said "RECIPIES". See example above.</p> <p> </p> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Here is the example from Daniel at working with my raw.</p> <p>I see a better defined label now (vs my linear resample) but I read "Panqetz Tardpes" not "Favorite Recipies" ha ha.</p> <p>Do you understand the point?</p><div></div> Link to comment Share on other sites More sharing options...
bernardwest Posted February 22, 2009 Share Posted February 22, 2009 "Bernie, give it the best shot at making up more detail using my raw and post it." This is the problem Mauro. This has been done by myself and others in other threads, and it gets ignored by you and yours anyway. I'm not saying that you aren't right, I'm just saying that you have no credibility claiming one way or the other, because your methodology is so weak. "The reason I upsample linerly and not bicubic is because I am comparing detail. Bicubic makes up information." Sure, but upsampling linearly you get NO detail because it gets mashed up with pixels. Do the comparison properly by upsampling how everyone upsamples in the real world, and let 'recipies' (sic) become 'tardpes'. It won't hurt your comparison, and it will give you far more credibility. Link to comment Share on other sites More sharing options...
mjferron Posted February 22, 2009 Share Posted February 22, 2009 <p>" Do you think your 3200 dpi scanner extract all the info from TMX"<br> No but I doubt 4000dpi is an earth shattering jump either.<br />"<br> Do you think there is a digital camera out there that outresolves TMX 100 35mm"<br> I believe any of the top end dslr's will out resolve 35mm tmax. And no where have I ever seen anyone prove different. If you have a link proving otherwise please let me know so I can stand corrected. Personally I shoot B&W film because I like the texture, look and tones.</p> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Bernie, can you please state what your understanding is regarding the resolution of TMX, Velvia and Ektar, Canon 40D, Nikon D2X, Sony A900 and Canon 5DII?</p> <p>If you don't know on a lines per pph basis, what is your relative understanding?</p> <p>Please do this for me if you would. It seems like many people argue but no one disagrees and I want to see clearly what you think.</p> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Bernie, I think I understand. You are saying that your results agree with mine, but linear upsampling make digital looks worse than it really is.</p> <p> </p> Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Michael, "I believe any of the top end dslr's will out resolve 35mm tmax. And no where have I ever seen anyone prove different. If you have a link proving otherwise please let me know so I can stand corrected."</p> <p>Of course I have a link (from my own test as usual.</p> <p>This is a scan of 35mm TMX with a Coolscan, the conversion from the chart to LPPH is 890. It resolves 4.4-4.5 which corresponds to 3900+ lpph (because this is the limiting resolution of the Coolscan). Under the microscope I can see 5500-6000 lpph.</p> <p>For comparison, the Canon 5DII and the Sony A900 only resolve about 3000 lpph.</p> <p>Michael, email me your address and I'll send you the film if you want to. I mailed a strip of the film to Rishi Sanyal and he quoted "With the TMX & the Imacon, I can see past 5. Maybe even 5.5. Meaning ~34 megapixels for the entire frame. Ridiculous.".</p><div></div> Link to comment Share on other sites More sharing options...
bernardwest Posted February 22, 2009 Share Posted February 22, 2009 <blockquote>Bernie, can you please state what your understanding is regarding the resolution of TMX, Velvia and Ektar, Canon 40D, Nikon D2X, Sony A900 and Canon 5DII?<P> If you don't know on a lines per pph basis, what is your relative understanding?<P> Please do this for me if you would. It seems like many people argue but no one disagrees and I want to see clearly what you think.</blockquote><p> Mauro, the funny thing is I actually don't care. Resolution at ridiculous scales as Daniel has highlighted, is so unrelated to the real world that only a zealot could possibly care. You and Les seem to be on some sort of religious crusade to defeat the evil digital doers. For all I know, and care, you may be right that film out resolves digital. But with your flawed methodologies and unanswered inconsistancies, you have no credibility. I haven't looked into the claims of Koren and Clark et al (remember, I don't actually care), but I know of their reputations and I would rather trust them with their internationally peer reviewed and acknowledged excellence than a couple of religious zealots with shakey methodologies. You asked me if I understood what the point is. I understand well, and you two would do well to understand that trying to deceive to get your point across can only back fire. If the case against digital is so overwhelmingly strong, as you indicate, then you should be able to prove it simply without resorting to shonky testing. Link to comment Share on other sites More sharing options...
bernardwest Posted February 22, 2009 Share Posted February 22, 2009 Les, I don't know what your problem is, but I only tried to ask you some simple questions relating to your methodology. You obviously didn't have a clue about Nikon's raw data, and you haven't answered why Lake Chad is in one image and not in the other. The only thing I can conclude from this is 1. You have no idea about raw processing, and 2. You possibly used two different maps for your comparison. By the way, I went to your links, but see no details about your testing methodology. All I can assume is that you didn't use raw, but in-camera jpgs. If that is the case, what were the settings? Link to comment Share on other sites More sharing options...
sjmurray Posted February 22, 2009 Share Posted February 22, 2009 <p>"Is there a difference between scanned film and digital?" Simple question, but actually too simple. "Film" can mean anything from Minox to 8x10 format. Digital can mean anything from 3mp to 4x5 digital back. Sure, you can compare drum scanned 4x5 with a 10 mp DSLR and there is no comparison. IMO it all boils down to what you need: what you like to work with, what you can afford, and what gives you the results you are happy with when the other conditions are met. Furthermore, resolution charts and MFT ratings are science, but we are all quite subjective when it comes to what we like when looking at a photo hanging on the wall. <br> Unfortunately this is what happens in this type of thread: we all have different subjective opinions, and then start to equate these with the tools and methods used to create the images we favor, and then assume that these are the best tools to use to create images for everybody. Well, sorry; we are all different in our tastes, and there is no universally perfect technology that will make "better" images than the other technologies, since "better" is totally subjective when it comes to art, as long as we are talking about art, and not about photographing circuit boards or arial photography of the surface of mars or something like that.</p> Link to comment Share on other sites More sharing options...
Dave Luttmann Posted February 22, 2009 Share Posted February 22, 2009 <blockquote> <p ><a href="../photodb/user?user_id=1501968">Bernie West</a> <a href="../member-status-icons"><img title="Frequent poster" src="http://static.photo.net/v3graphics/member-status-icons/2rolls.gif" alt="" title="Frequent poster" /> </a> , Feb 22, 2009; 07:03 p.m.</p> <p>"Bernie, give it the best shot at making up more detail using my raw and post it."<br> This is the problem Mauro. This has been done by myself and others in other threads, and it gets ignored by you and yours anyway. I'm not saying that you aren't right, I'm just saying that you have no credibility claiming one way or the other, because your methodology is so weak.<br> "The reason I upsample linerly and not bicubic is because I am comparing detail. Bicubic makes up information."</p> <p>Sure, but upsampling linearly you get NO detail because it gets mashed up with pixels. Do the comparison properly by upsampling how everyone upsamples in the real world, and let 'recipies' (sic) become 'tardpes'. It won't hurt your comparison, and it will give you far more credibility.</p> </blockquote> <p>Actually Bernie, I've never seen your comparisons. Actually, I've never seen anything from you. The rest of us have been posting tests and comparisons for years.....from you I only see opinion pieces. Feel free to post some samples proving Mauro incorrect. If you can't, I'd appreciate your moving on as your contribution to this thread has been nothing (just like Michael) but vacuous opinions.</p> Link to comment Share on other sites More sharing options...
bernardwest Posted February 22, 2009 Share Posted February 22, 2009 Dave, actually you've missed the point. I'm not contenting that Mauro is wrong. Would you like to post a quote where I said this? I'm contending that he and Les have shonky testing methodologies, and both appear to have purposely manipulated their results (I can only assume this, as particularly in the case of Les, he hasn't commented on why Lake Chad is mysteriously missing in one of his images). So the point is you can't make a meaningful decision on the subject one way or the other going by these two flawed tests. If you've got some evidence, I'd be happy to look at it and give you my thoughts. If not, like I said, I don't care either way. Link to comment Share on other sites More sharing options...
mauro_franic Posted February 22, 2009 Share Posted February 22, 2009 <p>Bernie, if you don't care, you will never find answers.</p> <p>The reason why I ask you what your understanding was about the resolving powers of fim is because it is clear you don't know. </p> <p>Your idea that me or someone else would try to cheat you with information is foolish. First why would I want to do that. Also you have my raw files as well as open access to all my film that is at your disposal. </p> <p>Lastly, and the main point, is that I always find out that people that argue, in general, don't know. And the last of the arguments always is "Ok, film outresolves digital, but I don't care" or "OK, Dynamic Range of negative film is better but I don't care". So what is the point in arguing?</p> Link to comment Share on other sites More sharing options...
mjferron Posted February 22, 2009 Share Posted February 22, 2009 <p>Steve Murry for the record some folks are saying 35mm tmax 100 is worth 40 mp in actual resolution. 10x that of a 10mp digital slr. No one has proved that or provided an independent link proving such a claim but I'll continue to shoot my 35mm B&W film just because I like it.</p> Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now