Film and Digital--Unscientific Survey

[sarah_fox]

<p>Hi Alan, I understand, and thanks for clarifying! My point in the first paragraph was clumsily made. Perhaps here's a better attempt: The pre-canned contrast curves used for creating jpeg images in digital cameras do actually have film-like shoulders and toes (sigmoidal rolloffs in the highlights and shadows, respectively) to ease highlight and shadow transitions. However, these rolloffs are not very substantial, for the most part. Even so, they prevent hard clips in the highlights, making them somewhat abrupt, instead. ;-) I think these rolloffs are not made more generous because they would otherwise diminish the latitude or dynamic range of the image and therefore make it more difficult to achieve an acceptable exposure. Again, I'm talking about pre-canned jpeg conversions. In Canonese, the term would be "picture styles." (No hable Nikonish.)</p>

<p>Anyway, these rolloff regions (shoulders and toes) are natural properties of film and are already imposed on digital representations, as they have benefit. The fashion most recently in film is to achieve very broad linearity with minimal shoulders and toes and expansive latitude (e.g. Tmax films). So it seems digital is already somewhat like film, and film seeks to be more like digital. Of course this has not always been the case. Fashions change.</p>

<p>What I do that is somewhat different is to shoot in RAW and create my own contrast curves (not push exposure sliders) that roll off highlights and shadows where I want them, while creating a higher contrast in the midtones that I like. I'm referring mostly to my B&W work. Color is not nearly so permissive, and if it becomes necessary to create generous shoulders and toes, it is also generally necessary to knock saturation down to accommodate the higher midrange contrast. But that's not a comfortable solution.</p>

<p>But to be clear, the objective in all of this is not to emulate film. Rather, it is to create the contrast curve that best fits the image I've recorded in RAW. This would be similar to Kodak refining an emulsion for a certain look -- and also of a technician or photographer adjusting chemistry/times to modify that look. The rolloffs in the contrast curves are a good thing in either film or digital, and I like my rolloffs a bit broader than most people. So that's what I create.</p>

<p>Anyway, your points are well taken, especially that there is a digital component to a scanned negative. But I would say in today's world that component exists more often than most people recognize -- whenever they take their film into a lab for processing. People still rave about film, but it's not the "same" as film from decades ago (unless they do their own processing or do business with an old-school lab). They just don't know it. In today's world, the emperor of film might be wearing clothes, but maybe not much more than his boxers and socks! ;-)</p>

[AlanKlein]

<p>Sarah: When I scan a film (flat-no pre scan adjustments), I usually get only a small part of the full histogram. Maybe half of the full range. The result is a flat, rather dark image with drab colors and little contrast. The first thing I do is adjust levels which usually takes care of 90% of the work to bring the image to something that looks decent. It also requires a lot of sharpening.</p>

<p>Please look at my film shots that I scanning and let me know if you can tell if the results look like digital or film. I really am not sure. There are both 35mm and MF. Then look at the BW set that I converted from digital to see if they look like digital or film. I'd really be interested in what you think.<br>

<a href="http://www.flickr.com/photos/alanklein2000/sets/">http://www.flickr.com/photos/alanklein2000/sets/</a></p>

 

[dan_south]

Since my name has been mentioned a few times since I last had time to check in on this discussion, I would like to clarify

that my initial comment was not sarcastic. It was economic.

 

Film processing incurs a frame by frame expense that digital capture does not. Shooting film is therefore a luxury that few

wedding photographers in this day and age can afford. - I would thank Mr. Bayer for reinforcing this luxury point with his

They're-Billionaires-So-They-Can-Afford-Film story, but it's just too far outside of the scope of what the average photo

business deals with to be illustrative in this regard.

 

My "too young to know better" comment was also directed at the young business owner who does not yet fully appreciate

the economics of cost containment, possibly because they never had to survive a downturn in the economy. I watched a

show about Ted Turner recently. Someone mentioned that he once went out of his way to pick up a quarter that someone

had left behind. This, in my experience, is the mindset of the seasoned business owner: waste nothing.

[wogears]

<blockquote>

<p>Film processing incurs a frame by frame expense that digital capture does not. Shooting film is therefore a luxury that few wedding photographers in this day and age can afford.</p>

</blockquote>

<p>Depends. If the clients need lots of images on a DVD, you're dead right. If they want a beautiful album and a CD/DVD with a hundred or so images, not so much. As I pointed out in the OP, many shooters use a combination of digital (low-light/candid/reception) and MF film (HQ portraits/formals).</p>

<p>My kid just got married. I didn't shoot the wedding, although in retrospect I should have. Had I done so, I would have used my TLR for the outdoor shots, and my Fuji X-E1 for the indoors. I probably would have shot five rolls of Portra 400 and maybe two-hundred shots with the Fuji. The film would add about $100 to the total cost of what would be about a $1500 package. Not negligible by any means, but not a profit-destroyer either.</p>

[dan_south]

<p>Les, I agree completely that the cost of film would not add much to the total price in the scenario that you described.</p>

<p>However, it's possible (and highly probable) that the couple will ask for digital copies of the images to share online. And not just for the select images that you chose for their album. They might want all of them. That would result in a lot of scanning that you would have to either do yourself (lots of time) or farm out (lots of money). Those expenses would have to be built into the film/digital hybrid business model.</p>

<p>I shot a wedding with B&W film some years ago (T-Max 100 using flash). The look was amazing, but the expense is not negligible. If the clients are willing to pay for the extra overhead, and if it doesn't impede the rest of the workflow, film can add an elegant dimension to the package. </p>

[sarah_fox]

<blockquote>

<p>Alan: I'd really be interested in what you think.</p>

</blockquote>

<p>Well, first of all, I enjoyed looking through your work, particularly the MF landscape! I think you and I have stood in some of the very same locations. ;-)</p>

<p>Your chromes and some of your color neg images do have a certain "look" to them, but I don't think this "look" is anything a digital image can't have. I suppose you can always take an image with lots of latitude and hike up the contrast, resulting in higher saturation and less dynamic range. You obviously can't go the other way. On the other hand, you can flatten out film's response curve digitally and get something that probably doesn't look particularly like either.</p>

<p>Having said all this, I really think any of these images can look just about however you want it to look, provided the information is there. There are so many ways to edit an image digitally that the possibilities are virtually endless. I think the only substantive differences in media have to do with limitations in information content (e.g. with the limited latitude of the chromes).</p>

[AlanKlein]

<p>Sarah: Thanks so much for looking at my photos. My question is not whether a digital can be made to look like film. My question is whether my Velvias look like film? Of course I'm working with a digital once I scan it. But ithe scan is pretty flat and I have to adjust it to make it look "normal". I don't use any recipe. I just adjust the levels first than continue with adjustment so the picture looks acceptable to my eyes in color and contrast. I don't try to make the result look like the original chrome. I don't even look at the film after the scan.</p>

<p>So that's the question. Do my Velvia photos look like film?</p>

[sarah_fox]

<blockquote>

<p>So that's the question. Do my Velvia photos look like film?</p>

</blockquote>

<p>This is a difficult question for me to answer, Alan, because I'm reluctant to say there's actually a "look" that belongs to film and not digital. But in the very narrow context of what one would get back from the lab after dropping off film, vs. the jpeg file one would get straight out of a digital camera using a pre-canned "picture style," I would say your scanned Velvia images do look much more like "film" and much less like pre-canned digital.</p>

<p>Here's the thing, though: You say you don't try to make the scan look like the original chrome. You want it to look a certain way, and you make it that way. That's exactly what I do with digital, and our end points might not be so different. So it really becomes pointless to identify what "looks" like film or digital. Either can look like just about anything, depending on how it's processed.</p>

[sarah_fox]

<p>For instance, here's an garish, "HDR'ish," eye-grating, cartoon version of the lovely water scene I posted in the other survey thread, which of course was captured on film:</p><div></div>

[alan_ginman]

<p>" I find it fascinating that with analog film and audio, there is literally no limit to the resolution of the recording"</p>

<p>Strictly speaking, 'analogue' film is actually just as digital as digital shooting. Each grain of silver in an exposed film is either exposed (1) or not exposed (0). So you end up with a series of 1s and 0s in both media.</p>

<p>Just a thought :)</p>

<p>Alan</p>

[AlanKlein]

<blockquote>

<p>Here's the thing, though: You say you don't try to make the scan look like the original chrome. You want it to look a certain way, and you make it that way. That's exactly what I do with digital, and our end points might not be so different. So it really becomes pointless to identify what "looks" like film or digital. Either can look like just about anything, depending on how it's processed.</p>

</blockquote>

<p>Agreed. Thanks for the review again. In the end I don't really try to make my Velvias look like film or Velvia for that matter. Only that they look well. Since I don't look at the chromes after the scan to compare, my interpretations can vary. Here are two shots taken from the same roll. I don't have the original film handy. So I don't know whether the differences in contrast and saturation are on the film due to different exposures. Or because I arbitrarily adjusted each after the scan independent of each others result. So which is the real picture? Which looks more like the original Velvia? Who cares? Interesting, most people have told me that the wide angle shot is "better" even though there's less saturation and contrast and probably looks less like Velvia. So in the end, other things can count for more.</p>

<p><img src="http://farm6.staticflickr.com/5281/5270429762_bb58f11c37_z.jpg" alt="" width="640" height="521" /></p>

<p><img src="http://farm6.staticflickr.com/5283/5270637805_5e15b955ab_z.jpg" alt="" width="640" height="531" /></p>

[sarah_fox]

<blockquote>

<p> So which is the real picture? Which looks more like the original Velvia? Who cares?</p>

</blockquote>

<p>Exactly!</p>

<p>BTW, I, too, prefer the wider shot because of the inclusion of the near shoreline. Nice photo! :-)</p>

[marc_bergman1]

<blockquote>

<p>So in the end, other things can count for more.</p>

 

</blockquote>

<p>Alan,</p>

<p>To me the position of the clouds in the wide shot draws one's eyes into the picture where it can linger. I can recall a photography teacher talk about this very concept. The fact that the wide shot has less contrast and less punchy colors doesn't matter to one's appreciation of the image.</p>

<p> </p>

[dan_south]

Beautiful shots, Alan!

[floyd_davidson]

<blockquote><b>Alan Ginman:</b><p>Strictly speaking, 'analogue' film is actually just as digital as digital shooting. Each grain of silver in an exposed film is either exposed (1) or not exposed (0). So you end up with a series of 1s and 0s in both media.</blockquote>

<p>

To be digital means that there is a finite set of discrete symbol values. Analogue means an infinite set of continuously variable values. Film is indeed analogue. It is a mistake to assume that each grain of silver is either a 1 or a 0. Two grains can have any value from 1 to 2.

<p>

There might well be a finite number of grains of silver, but their value is not discrete. If it were, an area with 2 grains would necessarily be twice as opaque as it would be with only 1 grain, but in reality the number of grains is only one of multiple determinants for the opacity. The size, the shape, and the overlap, all of which are continuously variable, also affect the opacity. Hence there is never a finite set of discrete symbol values. Any gradient can be made half as much, no matter how small.

[dan_south]

<p>Film grains can arrange themselves randomly. Digital pixels are bound to a grid. Digital processing engines interpolate pixel data and attempt to smooth transitions, but this is one of the reasons why the two media look somewhat different. Other visibly significant differences include contrast, dynamic range, and the look of noise versus grain. </p>

<p>Film doesn't have infinite resolution. The grains can arrange themselves in an infinite number of patterns and positions, but they're still grains. I have yet to see a film where grain was not visible at some degree of magnification. At least with 4x5 or 8x10 you don't need to enlarge the grain as much as with small or medium formats. Unfortunately, lenses for large format don't resolve as many line pairs per unit distance as small format lenses. For every gain in frame size, there's some degree of offsetting loss in resolution.</p>

<p>Having shot MF and LF film extensively, I'm completely satisfied with the resolution of 20+ MP cameras. 21 MP matches the detail of any MF frame that I've ever shot, and I can use a wider variety of lenses and take advantage of advancements such as image stabilization and live view for critical focusing.</p>

<p>I have seen outstanding four-foot-wide exhibition prints from 12 MP cameras (by professional photojournalists). How they accomplished that level of detail with so few pixels, I don't know, but they did it. And they didn't seem to need the 360 pixels per inch that some folks believe to be required for acceptable print quality. - 45 inches from 21 MP is the best that I've been able to manage.</p>

<p>Film is lovely, but it has no practical advantage over digital imaging. I have enjoyed shooting film over the years and look forward to shooting a bit more from time to time, but there's no denying the power, resolution, and utility of today's best digital sensors when coupled with high-quality lenses. </p>

[sarah_fox]

<blockquote>

<p>Film grains can arrange themselves randomly. Digital pixels are bound to a grid. Digital processing engines interpolate pixel data and attempt to smooth transitions, but this is one of the reasons why the two media look somewhat different. Other visibly significant differences include contrast, dynamic range, and the look of noise versus grain.</p>

</blockquote>

<p>At some level, I agree. However, when you get to a certain level of noise, the distribution becomes much coarser than the Bayer array, and it does appear rather random. I posted this in a parallel thread, but here is a 100% crop of such an image:</p>

<div></div>

[sarah_fox]

<p>I admit if you peep even closer, the non-random distribution of pixels becomes evident, and if you squint hard enough, you can even see (or perhaps just imagine) the individual pixels in rows and columns in this 100% crop. However, this grid is greatly overpowered by noise relating to the random nature of photon capture.</p>

[AlanKlein]

<p>Scanning film often give grains that aren't really grains but come from the scan process. Not sure what causes these.</p>

[dan_south]

<p>Digital noise can be suppressed rather effectively. I haven't tried the same with film grain. Any thoughts?</p>

[AlanKlein]

<p>Masking in LR3 in the sharpening section will reduce the grain in open sky areas without reducing the sharpening elsewhere. Well to a certain degree. I use that in combination with luminance noise reduction but of course you need a light hand as that also reduces sharpness throughout the photo.</p>

<p>Caveat: This is only for what I see on the screen. Since I haven't been printing, I don't know what results you'd get with a print with these adjustments. If you try printing, let us know how it worked out and any solutions you come up with.</p>