> > >…the image seems to be just fine at the end of my pixel pipeline. > (Truth be told, I do not actually know what would happen IF I dropped the > blacks into the negative).
I just did some experimentation, and I found out that some really bad stuff happens when the numbers go negative, (depending on how the image is exported), when I examine the areas outside my subjects which were clipped into negative values. As WebP, all the negative values seem to clip at zero. Any number that was not negative has a real, non-zero value. The result is that, as one approaches the darker part of the image, there is a sudden change from nearly black to pitch black. As AVIF, this phenomenon seems to disappear almost entirely in 8-bit mode, and virtually gone in 12-bit mode. As JPEG-JFIF, the issue affects more than just near-black to black, but also other hues in the shadows, making sudden jumps from, say, a dark shade of green, to perhaps a dark shade of maroon, to pitch black . This, of course, may be because the JPEG-JFIF compression algorithm favours colour accuracy in the highlights versus the shadows. Nevertheless, crushing the blacks into negative values may have undesired consequences. I may not have noticed before due to one or more of the following 1. the clipping was never on my subject, so not really noticed. 2. I mostly export to WebP 3. Our eyes see more details in the midtones/highlights, and our brains filter-out/fills-in the details of the shadows, so the phenomenon is not noticed. I must say, though, on the JPEG-JFIF, it was obvious to me. (On the others, I had to specifically study the shadows to notice). Sincerely, Karim Hosein Top Rock Photography 754.999.1652 On Wed, 26 Jan 2022 at 00:45, Top Rock Photography < ka...@toprockphotography.com> wrote: > >> > Whatever the reason, so long as I do not end up clipping (crushing) my > shadows in the [exposure] module into negative values, the image seems to > be just fine at the end of my pixel pipeline. (Truth be told, I do not > actually know what would happen IF I dropped the blacks into the negative). > >> >> ____________________________________________________________________________ darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
