How come I can't choose 48-bit color palette and color picker when editing a 48-bit RGB image?
Akyhne said: How come I can't choose 48-bit color palette and color picker when editing a 48-bit RGB image?
48 bit is not a color palette, it is a bit depth. I.E. all RGB color models in 48 bit images are 48 bit.
Well, that's exactly my point. You can't choose more than 24 bit color. because it is only in 24 bit.
If you create a custom image in 48-bit 2550px width * 200px height and make a fill from right to left with color black to white, you will get 10px in width with each of the 255 grayscale colors. It's not true 48-bit!
Akyhne said:make a fill from right to left with color black to white, you will get 10px in width with each of the 255 grayscale colors. It's not true 48-bit!
Actually, you do get a true 48 bit fill when you do the above. But the Photopaint eyedropper misrepresents the results, by only showing a 24 bit value.
I've attached a 1000x10 pixel 48 bit tiff which was saved without compression to make it easier to identify the pixels in the saved file. But lets concentrate on just the first three pixels, which are almost white ...
And sure enough, photopaint's eydropper tells me those first three pixels are all the same colour -- R = G = B = 255 or hex FE FE FE.
But opening the TIFF in a hex editor (I'm using PsPad Hex) tells a different story ...
Ignore the first 48 bytes of header -- the pixel values start at 0x0030 and I've circled the RGB values for each of those first three pixels.
To get actual hex values, we have to reverse the order of the bytes. So, those first three pixel values are:
And so, we confirm that the pixels generated by photopaint are not all identical, as the eyedropper implies, and if we look further down the file we will see 1000 different values not just 256.