Photoshop CC 2015: Info window - confused about maximum value at 16-bit color depth

  • 1
  • Problem
  • Updated 3 years ago
  • Not a Problem
  • (Edited)
When tracking actual color values of 16 bit image, the Info window shows color values which are about half of the correct ones. For example when I create gray-scale 16bit image the maximum color value (pure white) should be 65535 (ie, 2^16 - 1; because the colors are counted from 0, pure black). But the Info window will display the value of 32768 (BTW, 2 * 32768 = 65536 !== 65535) for pure white. You may also notice that when you click on eyedropper icon in Info window, there are 3 options titled "8-bit (0-255)", "16-bit (0-32768)", and "32-bit (0.0-1.0)" (see picture below). It is obvious that the range of colors for 16-bit option is incorrect. It should be (0-65535).


OS: Win 7 64-bit, Photoshop CC: 2015.1.1
Photo of MilP

MilP

  • 58 Posts
  • 6 Reply Likes

Posted 3 years ago

  • 1
Photo of Chris Cox

Chris Cox

  • 20280 Posts
  • 823 Reply Likes
No, the maximum value is 32768 in Photoshop.
That allows for a center value (important in some calculations), and much faster math with 16 bit/channel data.
Photo of Chris Cox

Chris Cox

  • 20280 Posts
  • 823 Reply Likes
Because it is using 16 bits, not 15.
Photo of Alain Bouchard

Alain Bouchard

  • 2 Posts
  • 0 Reply Likes
They are just using signed 16bit. Any scientific image is therefore destroyed. If you follow their logic, 8 bit should be 0-127.  Seriously flawed.
(Edited)
Photo of Chris Cox

Chris Cox

  • 20280 Posts
  • 823 Reply Likes
No, the value is 0..32768, not 0..32767.
Just because you don't understand something does not make it flawed.
Photo of Alain Bouchard

Alain Bouchard

  • 2 Posts
  • 0 Reply Likes
If you create a 16bit TIFF-LZW grayscale image with Matlab, IDL, FITS liberator, or Image Pro Plus and open that file in Photoshop (using CS6 still) and then resave it over itself, it will cause the file to shrink half the original size. Upon inspection of the resulting 16bit LZW image, all the odd value or rounded up so a ramp of 0,1,2,3,4,5 becomes 0,2,2,4,4,6,6 ... 1023 is now 1024, etc.  Your conversion is not reversible and damages the files. So if your plan is to destroy image data, then yes it is working flawlessly. I have files to demonstrate the problem if you'd like.
PS You don't know me so it is inappropriate for you to decide what I understand.
Photo of Chris Cox

Chris Cox

  • 20280 Posts
  • 823 Reply Likes
If you wanted bit preserving edits - then yes, Photoshop's 16 bit/channel mode is probably not what you want.  Photoshop's 16 bit/channel mode is designed for editing images, not bits.
And using the 0..32768 range still offers a significant performance and quality advantage over a 0..65535 range.