color shift in Aperture ?
the owls are not what they seem
Hmm…
For reasons that will become clear in a forthcoming post, I’ve been using Iridient RAW Developer quite a lot recently, which is always a pleasure.
Just now out of curiosity I decided to compare IRD’s interpretation of a RAW file with Aperture’s. I was actually more interested in sharpening (and IRD won, easily), but I ended up getting sidetracked by what looked like a pretty large colour shift introduced by levels in Aperture.
This is a crop of the photo I was playing around with (Olympus E-3 RAW, unadjusted):
It looks pretty much identical in both programs at default settings. Well actually if you look closely, IRD has slightly more detail, but that’s not the issue here.
Here’s the image with Level adjustments in the Luminance channel in Aperture:
And here’s the same thing in Raw Developer, or at least an approximation using a Lightness curve:
It looks to me that Raw Developer has nicely retained the color balance, but Aperture has introduced a pretty drastic magenta cast. It may be that my assumption that Lightness=Luminance is wrong, but I tried the same thing in RGB and got pretty much the same result. And of course my curve is not a perfectly accurate match of Aperture’s levels settings, but still, we’re not talking minor details here. The colour shift is huge.
I guess it’s all down to taste, but even so, it is easier to adjust to taste when you’re starting from an unadulterated baseline…
Addendum: after a bit more investigation, I noticed two things: 1, Aperture doesn’t actually appear to offer level adjustment in the Luminance channel. The Luminance / RGB options just change the underlying histogram display. The fact that the edit points stay the same seems to confirm this, although I could well be talking rubbish. 2, I can replicate the RAW Developer result by sticking to the Exposure and Contrast slider in Aperture. The lesson seems to be to treat Levels with caution.