由 stantang » 11日 12月 2008年, 09:54
Let me try to explain... as the writing said, sRGB is non-linear, so more data has been used for black area, therefore more steps/levels for black than for white. Since the monitor is no-linear as well, so it will display the image fine for human eyes. But we have to undo the non-linear color space before applying effects like blur, CC, etc... because the software works in linear color space (some don't though). Then you apply the conversion to sRGB IF you want your final output as sRGB.
Personally, I never use sRGB, but the same theory applied to Cineon files (dpx, cin...). Yes, your question indeed is a basic question...
dig deep, think deeper