Most tutorials for generating glitch art usually has to do with some form of manipulation of the compression of files. Should my goal instead to replicate the look of these glitches in shaders or is it somehow possible to authentically generate the compression artifacts in real time?


enter image description here

This effect which I'm particularly interested is referred to as datamoshing. It does "things" using the p-frames of a video (frames that I think store just the change in pixels). I feel like I need a better understanding of both graphics programming and data-compression.

  • \$\begingroup\$ That...is disturbing. \$\endgroup\$
    – Casey
    Commented Jun 29, 2013 at 5:47
  • 1
    \$\begingroup\$ youtube.com/watch?v=GTKmxBtaMF4 \$\endgroup\$
    – Tetrad
    Commented Jun 29, 2013 at 5:48
  • 1
    \$\begingroup\$ Also a lot more examples: reddit.com/r/brokengifs \$\endgroup\$
    – Tetrad
    Commented Jun 29, 2013 at 5:52
  • \$\begingroup\$ I think you need to provide more details about what you are after -- you're interested only in achieving the specified effect, yes? For what duration? How does this effect loop? Do you want it only applied to pre-rendered data or real-time, user-interactable data? Without further information it's really unclear. \$\endgroup\$
    – user1430
    Commented Dec 5, 2013 at 6:04


Browse other questions tagged .