We have a brand new look! Take a tour with us and explore the latest updates on Adobe Support Community.
Is there any reason why 32bit currently does not support all photoshop functionality after all these years of it being around?
I think Brecht was referring to 32 bit/channel bit depth rather than OS or Application support.
Brecht - It probably stems from many of the core features in Photoshop being originally written for 8 bit. 16 bit in Photoshop is actually 15 bits+1 so does not use the full 16 bit/channel capability. For image purposes it makes no difference but for things like height maps it does. But even in 16 bit some filters go AWOL.
As for 32 bit floating, I would suggest you add a feature request stating what functionality you would want to see. All functionality would not be required as 32 bit is really still a specialist mode for HDR and some 3D rendering. There is not a monitor made that can display it (the better monitors currently work in 10 bits/channel) and so working with 32 bit is always a compromise with only parts of the dynamic range visible.
Well since I am working in sizes of about 20.000 px x 6500 px would zooming in and out not not compensate for the 10 bit screens since I would make more of the bit depth visible by zoomin in on it? I am mainly lacking the layer functionality almost none of them work in 32 Bit but the dynamic range makes everything look fenomenal.
Bit depth has absolutely nothing to do with pixel size of the image. Zooming in and out does not change the bit depth of the image*
Bit depth describes how many levels are used for each colour within each pixel. So 8 bit there are 256 levels of Red, 256 of Green and 256 of Blue. For 16 bit there are 65536 of Red, 65536 of Green and 65536 of Blue. (Actually in Photoshop there are 32769 levels of each). Both 8 bit and 16 bit use the same black and white points (defined by the colour profile) so 16 bits splits the same span into smaller steps.
For 32 bit floating point there are 3.4 x 10 to the power of 38 levels each. In addition the 32 bit linear gamma image has a black and white point that is beyond that displayed on the screen, so you are only ever looking at part of the range in the on screen preview. The dynamic range in a 32 bit image far exceeds anything that can be seen on your screen.
Most screens use 8 bits (some 6+dither) giving 256 levels per colour channel. Higher end screens use 10 bit giving 1024 levels per colour.
* Actually when zoomed out to less than 66.7%, Photoshop previews all images in 8 bits/channel. That is for drawing speed not a limitation of the image or function of the screen.
Well I do notice a deffinite difference in quality, I would have guessed that zooming in om 32 bit images would reveal the levels that would have been squished down in a zoomed out image. What makes it that it looks so much better then?
You're chasing a red herring. As Dave explained, those two are completely unrelated. Two different things that have no interaction and nothing to do with each other.
Zooming does a lot of things in terms of resampling the original data. Depending on zoom ratio, an image pixel may be represented by any fraction number of screen pixels.
The only way to take this out of the equation is to view at 100%. This maps exactly one image pixel to exactly one screen pixel. This is the only zoom ratio where you see the actual pixel structure of the image.
Zooming in on any image reveals more pixel detail. When you zoom out several image pixels are combined into a single screen pixel. The only true preview is View -100% which maps 1 image pixel onto 1 screen pixel.
Zooming does not increase the bit depth or the capability of the display system to show that bit depth (the exception being that at zoom levels less than 66.7% 8 bit previews are used).
Edit : Sorry Dag, I hadn't seen your reply when I typed mine. We are both saying the same thing though 🙂