In Search of the Perfect Curve

Once again I am revisiting post-pro lomography. I will stop using film when they pry it from my cold dead hands. But the reality is, it’s only going to be harder to come by. So even more reason to perfect the ability to achieve similar results in the digital domain.

In tagging and archiving my work, I’ve come across some X-Pro shots of subject matter I have full-color digital images of. So, It seemed a good opportunity to see how my C-41 X-Pro curve stands up to reality. And it faired pretty well.

Long ago I did some research on this, and came up with very similar curves from many different sources. So, I created my own to taste based off the tutorials, tweaked and, and most recently tweaked it again. This could go on forever. The problem is, there is no one definitive curve for this, because every time you shoot X-Pro with a different film emulsion, you’ll get a different response, albeit probably a consistent one.

For example, if I were to shoot the same subject matter under the same lighting conditions with Kodak Ektachrome 100 and Fujichrome Velvia 100 and process them C-41, I’d get very different, unpredictable results. But as long as the film was the same batch and age, most likely I could repeat those results if I did it all over again, especially since C-41 is a normalized process.

So, the variables are really the emulsion and the exposure itself. So, If I want to have a similar digital experience, I need to build a single curve for each emulsion, and quit with the tweaking. If there’s any tweaking to be done, it’s in the exposure of the image before I run it through the curve.

And, not all images look good as X-Pro. So if the image doesn’t look good with a simple application of the curve, I feel it’s safe to say that it wouldn’t have come out well as a X-Pro film. And to force it to be kind of defeats the randomness of the whole X-Pro experience.

Thus, this becomes a project in the making. And here’s how I’m going to approach it...

First, I’m going to create a target image, which I’l print on 8-1/2 x 11 paper. All the additive and subtractive primary colors, plus shades of gray. Then, I’ll photograph this image outdoors in daylight using my film of choice, bracketing my exposures by perhaps a half stop. Then, under those same lighting conditions, I’ll photograph them digitally. I’ll then apply my curve and tweak it for the last time. But it won’t be the last curve.

I’ll then repeat this procedure for every roll of color reversal film I use. Perhaps I’ll only take a single exposure, or at most a bracketed set of three, but I’ll still be able to use them with the original digital versions. Scientific method would dictate otherwise, but let’s not go crazy here. The ballpark is all right by me.

So, my X-Pro journey should net me a set of Photoshop curves, each one labeled for a given emulsion. And given that they are curves, they should be able to be applied in Adobe Camera Raw, right?

Wrong.

Unfortunately, ACR handle curves very differently than Photoshop. If you create curves in ACR based on the input and output values for each sample point in the Photoshop curve, you’ll get very different and disappointing results. So, If I want to do this in Camera Raw as well, I’ll have to create those curves separately. And given ACR’s limited tools (like the fact that you can’t sample color in the L*a*b* color space) it’s going to take a long time. But, it’s doable, and happily undoable due to the non-destructive nature of ACR.