This content has been marked as final. Show 10 replies
>Is better keep them in linear gamma or is better convert them to a gamma corrected color space before import them in ACR?
Just keep them in linear. As long as there is a correct profile attached to your TIFF, ACR should read them fine. The fewer conversions the better.
And if you're working linearly (gamma 1.0) you really need more than 8 bits/channel.
Thank you Jao, thank you Chris.
One last question: my first edit in PS on linear image is tweak the exposure with a linear curve. Make the exposure slider in ACR the same thing?
(I suspect this is a stupid question because ACR works linearly so the exp slider should work exactly like my linear curve in PS)
Exposure in ACR/LR is designed to mimic the effect of real exposure when taking a photo. Just like in photography, it modifies the luminosity by multiplying it with 2^EV. It only does this after substracting the black value (which is less photographic of course but more practical). I am not entirely sure what you are doing with your linear curve, but if the only thing what you're doing with that is changing the whitepoint, then yes indeed, it is the same thing. The slope of the line in the curves panel will be basically the exposure setting in ACR/LR. Steeper than one slopes correspond to exposure increase and smaller than one slopes correspond to exposure decrease.
> my first edit in PS on linear image is tweak the exposure with a linear curve. Make the exposure slider in ACR the same thing?
These are vastly different.
1. The "exposure" adjustment in ACR increases the raw pixel intensities, like increasing the exposure would have done. This is a linear adjustment
b before the non-linear mapping.
Consequently, the linear adjustment becomes non-linear in the converted data.
Try it out: increase the exposure in ACR by one stop, then two stops, etc. and watch the displayed RGB values at the picker: very dark spots almost double, while bright spots increase only little.
2. The "linear adjustment" in PS works on the mapped data.
Try it out on several GREY spots: move the top right corner of the straight line in "curves" to the middle of the top edge (this corresponds to +1 EV). Make a selection on a dark spot and watch for the average under the histogram; turn the adjustment layer on and off. The RGB values got doubled. Now do this on a middle dark area: the same happens. Then pick an area close to but less than 128: you get close to 255. Try it with an area over 128, no matter how intense: you get 255.
2) ...not if he imported linear data into Photoshop. If the data is linear in Photoshop, it'll stay that way. It sounds like your test imported gamma encoded data (possibly after saving from ACR).
> not if he imported linear data into Photoshop
The OP was mentioning importing linear data
i into ACR,
not into PS.
>The OP was mentioning importing linear data into ACR, not into PS.
No he didn't. He wrote:
>my first edit in PS on linear image
Describing his normal, non ACR workflow. So yes, they are indeed the same thing. Of course if you take a linear image through ACR it will get gamma mapped.
I was referring to
> I convert with a third party tool some images in tiff (linear gamma) and I want edit them in ACR
Anyway, we are on the same page.
I try to explain better the two workflows that I use only sometimes:
1. I import my linear gamma images in PS, assign them my ICC profile made by ProfileMaker 5 with a linear target image converted in the exactly way like the images, I tweak the WB and exposure with linear curves (moving only the white point of RGB, R, G and B channels) and then convert them in a gamma encoded color space before using the PS tools.
2. the new sperimental workflow should be: convert the raw in linear images and assign the ICC profile, then import them in ACR.
WB and Exposure compensation (highlight recovery?)should be made directly in ACR.
Thank you to all.