Kodachrome
Me too! Me too!!
So, has all the fuss about Kodachrome’s final expiration date died down? Excellent, must be time for me to stick my oar in.
I used Kodachrome a lot in The Old Days™. I’ve got stacks of mouldy decaying K25 and K64 slides, shot using either a Canon FT, a Canon FTb, or an Olympus XA3. I used in the Antarctic, because it was what people who understood photography told me to use. I barely knew an exposure from an aperture ring in those days, and the idea that a lens could have a “speed” was totally baffling. Not much has changed.
But anyway, thanks to the convergence of luck, sheer quantity, and the Shakespearean Monkey Principle, I got a couple of photos that I like. Here’s two of them, chosen because - with a lot of hindsight - I think they have a certain quality of texture and colour which I’ve been trying to recapture ever since, and which may be credited to Kodachrome.
The British Antarctic Survey ship, RRS John Biscoe, in Dorian Bay, December 1987
A couple of Adelie penguins, off Coats Land, December 1991
Both of these slides have of course gone through a scanner and Photoshop, but have been manipulated to match the originals, or at least my interpretation thereof.
It was interesting to read about another photographer’s thoughts about Kodachrome over at my current favourite blog. Although in general I’ve never been particularly interested in film emulation plug-ins (more about this later), I thought I might try the free demos of two of the most popular to see what they could do. So I downloaded Alien Skin Exposure and DxO Labs Film Pack and applied their respective K64 presets to a shot I took last week on the Plaine Morte glacier here in Switzerland.
Comparison collage
Dunno about you but I’m sorta like “yeah, whatever” (<- see I can do teen talk too!). The most striking point in my view is that the two renderings are completely different. The Alien Skin version is pretty subtle, but DxO totally changes the white balance. Now, I didn’t apply these to an out of camera file, but to a version exported from Aperture and tuned to my liking. The fact that Alien Skin doesn’t alter it much might (huge stretch here) indicate that I’m subconciously applying a Kodachrome look anyway.
Anyway, neither of them interest me very much, but I will say that the Alien Skin offering seems at a quick glance to be far better. DxO’s user interface is really not very good, the installation is clunky, and it doesnt offer anything like the range of Alien Skin. Actually this ties in with my general feeling about DxO stuff.
Generally I’m a little bit skeptical about some of this film stock talk. The differences between various films (I’m talking slide film here) are often extremely subtle, especially in an outdoors, uncontrolled lighting environment, and in any case, once inserted into a digital workflow, they’re largely irrelevant. On a lightbox I have a personal, and unconventional, liking for Fuji Velvia 100F, and I’ve used it a lot. But over the years my choice of film has generally been more influenced by ISO rating and the light levels I’m expecting to encounter than anything else. These days I’ve fixed on using Ektachrome E100G, because it is very neutral, has a wide dynamic range (for slide film) and scans very well.
So yes, I can be nostalgic about Kodachrome too. But I’m not terribly interest in artifical emulations of it.
2 comments
Project Hyakumeizan February 10, 2011 - 9:44Now, if only Kodak could produce a slide version of its Ektar 100 film - that would really be something ....
2 comments
Robert Boyer August 25, 2011 - 1:58Way way back I went through the pain of trying to analyze the Kodachrome look in various clinical and controlled ways. Unfortunately I cannot find my original materials (back in the 1990) because of a rash of incidents involving a succession of really pissed of female companions.
In any case no matter how you end up doing the clinical study it is fraught dangerous missteps and wrong-headedness. I tried photographing calibration targets, with all the Kodachome films and scanning them, then trying to match my cameras outputs, etc, etc. Anyway you slice it there are some issues where the digitization of the slide introduces it's own flavor of the day for various reasons that I am sure you can extrapolate. I am sure you can also imagine when you try to "neutralize" it to some standard that is also a bit wrong-headed.
Then there is the matter of differences in digital capture etc, etc. Lens color flavors, blah blah. You can probably guess why there is no "standard" thing you do to a file and it looks exactly like Kodachrome. The closest you can do is standardize your digital capture file to some standard and then apply some flavor of Kodachrome simulation to it. At the end of the day no matter what you do you are left with some "interpretation" of about what the actual slide looks like more or less. If you are following me at all - you can see how the fact that one of the simulation filters you tried being close to your original has no bearing on weather or not your original treatment was sort of like kodachrome. The software is not to the point where it can make a judgement that says "Hey man this already looks like Kodachrome so I will just fix it a little bit".
The best way I found of trying to match my digital results to kodachrome (which I did have a productive reason to do so at the time) was to match the contrast curve as much as possible between my commercially drum scanned Kodachrome and my digital capture, make sure the grays/whites in both were neutral based on a known pre-determined correction for the Kodachrome and then due various HSL adjustments to the neutral digital capture to emulate the color response as closely as I could.
If memory serves (questionable at this point) assuming a similar grey value for both the film and a very neutral Digital Capture (think calibrated against MacBeth) the digital file needed...
-Reds moved towards orange/yellows and more saturation.
-Yellows moved more toward orange and more saturation with a little less luminance
=Greens shifted a hair toward blue, with a little less saturation, and maybe less L
-Blues shifted a tiny bit toward magenta, a little less S, and a little more L
=Magentas the same with less S
Now... The simulations above could actually both be "correct" depending on what sort of "methodology" they went through in terms of digitizing the original Kodachrome, the characteristics of the "test case" digital capture as well as the handling of the original film and K14 processing. I have large number of occasions where K64 went magenta on me like the one simulation above - I never figured out why but my guess is the processing center that I sent it to was to blame.
After going through this like some crazed scientific nut case I found it just as effective to match the contrast curve of your standard digital capture to your visual impression of the slide and the same thing goes for the color characteristics in relation to one another. I did both and ended up more or less in the same place...
Just my 2¢.
RB