35mm film vs 5DII - Low light performance

[mauro_franic]

<p>Scott, there was no scientific methodology for this comparison. The purpose of the shot had nothing to do with a comparison in mind.</p>

<p>I thought the results were interesting to share, stating all the specific conditions in which they were taken. I believe they were appreciated of offering value to other people.</p>

<p>If you ignore the 5DII shots (which there are plenty more samples available in high ISO), what do you take away from observing the negatives (color and B&W) shot at EI 1600 and EI 3200?</p>

[mauro_franic]

<p>Scott, regarding the 400 ISO film at EI 1600 or EI 3200 (and/or the 5DII), please post any experience you may have (with samples if possible) that may help the discussion.</p>

<p>I perceive some frustration in your responses as if your experience/results differ greatly from these examples.</p>

[swilson]

<p>I find Mauro's comparisons to pretty much match what I see between my "low light" film shots and my low light digital shots. With film I used a flash and got the same bright foreground with a very dark background, with digital I have been at last getting a good mix between the two.<br>

I don't believe low light performance is a strong point of film.</p>

[scott_ferris]

<p>Mauro,</p>

<p>I understand when you were shooting the event there was no thought to a comparison, but that is what you have posted! And it is a very poor, heavily biased and totally unfair comparison.</p>

<p>I do have images shot at concerts, but they are not 1600iso images, I also have digital images underexposed by one stop and exposed at 1600 and push processed to simulate 3200iso images (my camera doesn't go to 3200). None of my images look anything like yours, I'd happily post them, but to what end?</p>

<p>Your post raises so many questions it is difficult to ask them all and not appear insulting. To say you didn't do it as a comparison, yet title the thread X vs X rather contradicts that, to suggest nobody is influencing their buying decisions on your "findings" also appears off the mark, as Russ even goes so far as to say it has.</p>

<p>Very disappointing post from somebody who has a true love and ability with film, yet seems to consider digital a very second class citizen. As I said, there are placating, even condescending, comments about the 5D MkII, but really this is another badly presented "film is better" thread.</p>

[mauro_franic]

<p>Scott, thanks for responding.</p>

<p>I do consider digital a second class citizen, just different. Digital is also an extremely convenient tool and provides excellent quality for most applications.<br>

<br />I make a point to stress that in my posts when possible while still mantaining an honest debate.<br>

<br />To reiterate, I think the 5DII did great for a live shot at ISO 3200, especially considering you are looking at a 21MP file at 100%.</p>

<p> </p>

[mauro_franic]

<p>Film did great too which is the reason for my post. There aren't very many examples available of film at EI 1600 and 3200 so most people are not aware how well film performs. </p>

<p>Moreover, many people are convinced, (without knowing), that film does badly and with unusable grain even at ISO 400 or 800. That is the target missconception I intended to address with this examples.</p>

[mauro_franic]

<p>Scott, no there is also no need to post your digital examples unless you think they offer good information for analysis. Digital perform great in low light and that is not in question.</p>

<p>If you have examples of film at EI 1600, 3200 or higher that provide better results, please post them. They will be a great contribution.</p>

[scott_ferris]

<p>No I don't, that is the point, film can't compete with digital in high iso tests. Correctly developed digital images are leagues better than film ones at similar, high, iso's.</p>

[mauro_franic]

<p>Film doesn't have to compete, it is a great compliment and a complete different look.</p>

<p>Could you please expand on why you concluded that your results with film at high ISO are not good?</p>

[scott_ferris]

<p>I started using film seriously in 1978, I have never taken an image above 800iso with film (135 format) that I am comfortable printing above 8x10, unless I am looking for a grainy style. I can print my digital 135 images that have been pushed to 3200iso much larger than that, or more practically, crop them much harder.</p>

<p>Of course film has to compete, unless you are just using it as a niche fancy, if that is the case then there is never any point to X vs X threads unless the two X's are different films.</p>

[mediumformat]

<p>Here's some more fuel to the fire posted by a real wedding photographer comparing a 5DMKII to Ektar 100:<br>

<a href="http://www.twinlenslife.com/2011/01/digital-vs-film-canon-5d-mark-ii-vs.html">Digital vs. Film - Canon 5D Mark II vs. Kodak Ektar 35mm </a></p>

 

[mauro_franic]

<p>"I have never taken an image above 800iso with film (135 format) that I am comfortable printing above 8x10, unless I am looking for a grainy style."<br>

Scott, go to my link in this post and retrieve from that folder any of the scans. Then print them larger than 8x10. Then comment on the grain you see on the print.</p>

<p>I would like your honest opinion.</p>

[scott_ferris]

<p>I will late next week when I am back in my office. But why? Do you believe your film images are less grainy than mine? Certainly from the original sized images on screen I see nothing to make me think your images are.</p>

[mauro_franic]

<p>I believe that you will conclude that grain is not an issue.</p>

[scott_ferris]

<p>Well it always was until I got the current generation of digital 135 format cameras.</p>

<p>If I have to shoot at 3200iso, and I very rarely do, this is what I want from 100% crops.</p><div></div>

[scott_ferris]

<p>This is the full image.</p><div></div>

[mauro_franic]

<p>Looks very good.</p>

<p> </p>

[mauro_franic]

<p>By the way, anyone in the thread is welcomed to download the full files of the scans at home to test the print quality directly. Just select a picture, click on "Original", and drag it to your desktop or Photoshop.</p>

<p>http://shutterclick.smugmug.com/Photography/Portra-400-and-TMAX-400-G/15789423_WvenE#1185545623_eRS9a</p>

[mauro_franic]

<p>On the prints I made here at native 360 dpi (this is 11x16 instead of just 8x10) on Epson Ultrasmooth and Ilford Fibre Gold grain is not an issue.</p>

<p>At 16x20 you can see it but it is pleasing. It also can be removed with one click in PS if someone prefers to.</p>

<p>Look forward to your feedback from the prints Scott.</p>

[rishij]

<p>Mauro, what I really wonder is this-- what's better when shooting film at a concert scenario with flash:</p>

<ol>

<li>ISO 400 film exposed at EI 400 w/ auto flash, then pushed 2 stops during processing</li>

<li>ISO 400 film exposed at EI 1600 w/ auto flash, then pushed 2 stops during processing</li>

</ol>

<p>In a concert setting, one desires a good balance between flash-lit foreground & ambient-lit background. Therefore, I will typically set the ISO on my 5D to 800 or 1600, manually select a shutter speed that won't result in offensive blurring given the level of ambient light, select the widest aperture on my lens, then use the white card on my 580EX II aimed at the performer (flash pointed up, perhaps adding some ambient light to the room). Basically, in this scenario, upping the ISO ensures that the flash exposure is low enough that when analog gain (magnitude of which is determined by the ISO setting) is applied to the raw data off the sensor prior to being fed to the A/D converter, you haven't blown out the performer (foreground) exposed by the flash. Furthermore, the higher ISO helps b/c, holding shutter speed/aperture constant, the analog gain increases the raw pixel values of the shadows (well, everything for that matter) which then allows more bits for their representation after A/D conversion (theoretically, if you had infinite bits, there'd be no need for an ISO setting on a digital camera).</p>

<p>But since negative film has such exposure latitude, one may not have to worry about the push processing blowing highlights on the flash-exposed foreground, no? In which case, one could expose as ISO 400, let the flash fire at higher power, maybe even helping to expose the background some more, which'd help with the film shadows.</p>

<p>So basically I'd love to see what'd happen at a concert with ISO 400 film shot as EI 400, with E-TTL flash, then pushed 2 stops.</p>

<p>Which also brings to mind another question I had: does shooting ISO 400 film at EI 400 & then pushing 2 stops lead to more shadow detail than developing it with no push, & then brightening shadows in post? The question really is whether or not push processing helps silver growth for very unexposed grains or only really extends the growth of mid to well-exposed grains. After all, there's a critical threshold below which a grain is considered 'unexposed' & just won't get reduced, right? (Well, I guess it's not the whole grain that gets reduced but sensitivity specks that have large enough clumps of reduced silver from the exposure... ACK. Flashbacks to that epic thread... :)</p>

<p>Rishi</p>

[dogbert]

<p>Hilarious post. I wonder how anyone can take this seriously. Titled 35 mm film versus 5D mark II low light comparison, but then Mauro claims when the criticism gets too much that he was not intending it to be a film versus digital comparison.<br>

Mauro also makes it clear the purpose was not a scientific or controlled test, but then runs to draw conclusions everywhere that would not be supported by a controlled test.<br>

The conclusions that I draw from this thread is that in low light conditions, carefully developed and porcessed film shot with a flash at 5500k at f4, will look different to digital shot at ISO 3200 and f1.8 with ambient coloured concert light.</p>

[Karim Ghantous]

<p>R.J. Fox, thank you for that link. I didn't realize that Ektar 100 was that good! Based on that simple comparison, it seems that a 6x7 negative might match an 80Mpx MF back for resolution. Maybe. :-)</p>

<p>I wonder if anyone should ever bother pushing negative film. I'll have to do more research on that.</p>

[tony_leinster]

<p>Oh come on Geoff, how can you draw that conclusion! You obviously looked at the pictures and methodology! If you're not careful you will be accused of "lies and deceipt" and even "corruption" and agreeing with the evil companies whose only plan is to keep us all in digital and away from film. Then you would have to be endlessly told that it's not a film vs digital debate (you were fooled by the title there weren't you) and that actually digital isn't that bad really but let's get back to the superior world of film. You might claim that you want an unbiased comparison but we all know you're probably a digital user who just wants to claim that you have the best system. Shame on you!</p>

[rishij]

<p>While that 5DII vs. Ektar 100 comparison was interesting, I think the Ektar coulda performed much better if he'd used a decent scanner. When I scan Ektar 100 on my Minolta 5400 or Nikon LS-9000/4000, I can definitely see grain. His scan is probably throwing away half the information on that Ektar. Now, whether or not that information is useful is another matter ;)</p>

<p>Also, the fact that the 1st 5DII shot at 100% looked like it had motion blur but in the nighttime shot comparison the Ektar 100 shot looked like it had motion blur kinda invalidates his test. He shoulda had a more controlled environment: tripod + triggered release.</p>

<p>Karim -- let us know if you ever figure out the answer to whether or not it's worth pushing negative film. Certainly probably makes a difference for slide, since there you're re-expanding the tonal range of the medium (which is initially small after the exposure) to make a final product for presentation.</p>

<p>While we're on the subject -- does anyone know if Velvia (or any positive film for that matter) actually captures the same # of stops of dynamic range as negative film in the latent image? But then loses it in the chemistry of making a high-contrast pleasing image for presentation? Might not even be a valid question since it's likely a combination of latent image + development that determines the # of stops recorded... though... I don't know.</p>

[15sunrises]

<p>There seriously seems to be something wrong with the 5DII shots, seems like they were underexposed and then pushed in post or something. I'd expect nearly the kind of ISO3200 performance in the example from my 40D on a properly exposed shot...</p>