Forums before death by AOL, social media and spammers... "We can't have nice things"
|    comp.arch    |    Apparently more than just beeps & boops    |    131,241 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 130,105 of 131,241    |
|    BGB to Robert Finch    |
|    Re: Tonights Tradeoff (3/3)    |
|    02 Nov 25 14:58:36    |
      [continued from previous message]              ends up needing palettes or encoding schemes that are slightly irregular.                            For color-cell, there are different approaches depending on how fast it       needs to be:        Faster: Simply select minimum and maximum luma;        Selector encoding is often via comparing against thresholds.        Except on x86, where multiply+bias+shift is faster.        Medium: Calculate along 4 axes in parallel;        Select axis which gives highest contrast;        Usually: Luma, Cyan, Magenta, Yellow.        Adjust endpoints to better reflect standard deviation.        Vs simply min/max.        Slower:        Calculate centroid and mass distribution and similar;        Better quality, more for offline / batch encoding.                     As noted, early on, I was mostly using real-time color-cell encoders for       Doom and Quake and similar (hence part of why they were modified to use       RGB555).              Some of this is also related to the existence of a lot of RGB555 related       helper ops. Though, early on, had also used YUV655 as well, but RGB555       mostly won out over YUV655 (even if it is easier to get a luma from       YUV655 vs RGB555).              ...              --- SoupGate-Win32 v1.05        * Origin: you cannot sedate... all the things you hate (1:229/2)    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
(c) 1994, bbs@darkrealms.ca