Copy link to clipboard
Copied
Hi!
In my work I use Photoshop mainly for edits of 3D renderings, so 16-bit images. In workflow I am in between either using Camera RAW for main colour/exposure/contrast adjustments + adjustment layers, but sometimes I find it way more useful to have the edits possible in the layer stack from the very beginning. However, when switching to the adjustment-layer only workflow I notice a bit of limitations from Photoshop, for example:
a image, that is a bit under exposed (but without black areas) with correctly exposed light, now to brighten it up in general:
1) In Camera RAW I just put up the exposure, then take down highlights and in those two simple moves I can solve the problem - shadows are brighter and lights are still not burnt.
2)Trying to do the same in adjustment layer. First of all: exposure makes everything way too burnt. Then, when I try to clip the bright spots through Levels, it simply turns these burnt areas gray...
In general: Camera RAW somehow remembers the information that got brightened even to pure white and still lets us retrieve it, but when using adjusment layers somehow the previous adjustment becoems a tabula rasa for the next one. So my question is: am I doing something wrong? Is there another way or method I am missing? Or is it just the way it is?
Thanks!
2)Trying to do the same in adjustment layer. First of all: exposure makes everything way too burnt. Then, when I try to clip the bright spots through Levels, it simply turns these burnt areas gray...
By @damianw75349309
One thing you are running into is that many adjustment options, even with the same name, do not do the same thing in Camera Raw and Photoshop.
For example, in Camera Raw, Exposure is a general purpose brightness adjustment, with an emphasis on the midtones. In Photoshop, Exposure appears to be intended for correcting HDR images in gamma 1.0 (linear)
...Copy link to clipboard
Copied
It's not about ACR vs Photoshop. It's about linear raw data vs gamma encoded RGB data. Those are two very different classes of data.
A raw file can contain up to 14 stops of dynamic range, recorded by the camera sensor. This is possible because of the linear tone curve, which compresses the tonal range and gives room for brighter highlights and darker shadows. This isn't a usable image that resembles any real scene. It looks very drab, very dark and very flat.
This extra dynamic range can be recovered and remapped in ACR, still working with linear data.
To process this into an image you recognize, it has to be gamma encoded. A gamma encoded image can contain about 7 or 8 stops of dynamic range from black to white, at most. This produces an image with a contrast curve resembling a real scene.
In this process, a lot of data are discarded. The result is that you don't have the same headroom for highlight and shadow recovery when working with an RGB file in Photoshop.
The conclusion is very significant and important, and the golden rule is: Do as much work as possible in ACR, on linear data, before moving to RGB data in Photoshop.
Copy link to clipboard
Copied
I agree with D Fosse, the RAW data is far more open to manipulation than the processed output (which is what youre working on in Photoshop).
If you can do those edits in RAW.
I hope this helps
neil barstow, colourmanagement net - adobe forum volunteer - co-author: 'getting colour right'
google me "neil barstow colourmanagement" for lots of free articles on colour management
Copy link to clipboard
Copied
Hi
D Fosse and NB Colourmanagement are correct in terms of the difference between RAW and RGB gamma encoded data. However, I read your first post as referring to rendered images from a 3D application which are not raw images. It looks like you have chosen to render in a 16 bit RGB gamma encoded space.
If rendered in 16 bit, I often use the Camera Raw filter on a smart object which keeps adjustments non destructive and gives the tools with which you are familiar. If you prefer adjustment layers, then the best adjustment tool in those circumstances is Curves. As you have found in 8 bit and 16 bit, Exposure just brightens the image and sends the extreme highlights into clipped white. With curves you can control each value range and avoid such clipping.
As an aside, Exposure in 32 bit linear does not clip as the values in the image can be way beyond the range displayed on screen (blacker than displayed black and whiter than displayed white). That is why tone mapping to a 16 bit or 8 bit image is required. This is an alternative workflow from a 3D render, but at some point you still need to tone map to 16 or 8 bit for use as a conventional image.
Dave
Copy link to clipboard
Copied
@davescm "rendered images from a 3D application" ah, yes, thanks, your experience with 3D should be rather useful to the OP
neil barstow, colourmanagement net - adobe forum volunteer - co-author: 'getting colour right'
Copy link to clipboard
Copied
2)Trying to do the same in adjustment layer. First of all: exposure makes everything way too burnt. Then, when I try to clip the bright spots through Levels, it simply turns these burnt areas gray...
By @damianw75349309
One thing you are running into is that many adjustment options, even with the same name, do not do the same thing in Camera Raw and Photoshop.
For example, in Camera Raw, Exposure is a general purpose brightness adjustment, with an emphasis on the midtones. In Photoshop, Exposure appears to be intended for correcting HDR images in gamma 1.0 (linear), not the more common conventional gamma-corrected SDR images. In Photoshop, a conventional overall brightness adjustment is supposed to be done by changing the Brightness option in Brightness/Contrast, the middle Input Level in Levels, or a midpoint in Curves.
Same with Highlights. In Camera Raw, the Highlights option employs a sophisticated algorithm that incorporates highlight rolloff and highlight recovery (using multiple channels if needed). If you adjust the white clipping value in Levels, this is not the same thing at all, because all that does is clip hard, without any intelligence. If you tried to be more subtle about it by shaping the highlight part of the curve in Curves, that would at least be doing some kind of rolloff, but it would not do any kind of highlight reconstruction/recovery, so that still would not match the visual results of the Camera Raw Highlights option.
The closest thing Photoshop has to the Highlights and Shadows options in Camera Raw are the options in the Image > Adjustments > Shadows/Highlights command. But again, it’s not the same code, so you won’t get the same results. The version in Photoshop came first, but I’m not sure if it’s ever been updated, so it can seem like a blunt instrument compared to the more recently and more frequently updated Highlights and Shadows options in Camera Raw.
Also, Adjustment Layers are quite simple; they just take a layer pixel and do math on it involving the same pixel in underlying layers. In contrast, the options in the Basic panel of Camera Raw are image-adaptive, so the visual results of applying a certain value (such as Highlights -20) can vary depending on the content (neighboring pixels) in each image. For example, the Camera Raw Highlights and Shadows options don’t just take each pixel and change the value, they analyze surrounding pixels and can alter groups of them, employing masks and other code to improve local contrast and the visual result. A Photoshop Curves adjustment layer that increases contrast for shadows necessarily reduces it in midtones and highlights, while the Camera Raw Shadows option increases contrast for shadows but automatically masks off other tonal ranges to protect them…fewer compromises.
In short, if you want those adjustments in a Photoshop layer stack with the results that Camera Raw would produce, it would be better to choose Filter > Camera Raw Filter. And if you want that to be editable and nondestructive like an an adjustment layer, convert the image layer to a Smart Object before you apply the Camera Raw Filter. The cost of the Smart Object approach is a larger file size and more complex workflow than with adjustment layers.
Copy link to clipboard
Copied
Right. I really missed the target on this one.
But in the best Socratic tradition, we're getting there, if by a convoluted path 😉
Copy link to clipboard
Copied
Hi! Thank you all for these meaningful and fantastic inputs, although I have never professionally learned PS, I now feel a gap. However, I reckon everything was explained in depth, so it has been a good read and everything has been understood and noted down, indeed experimenting with curves only gave me some more satisfying results, but maybe there's no more profound reason to look for this type of workflow when ACR seems more potent as a tool. At least I have learned something new. Thanks a lot!