@cmuratori @tom_forsyth @nothings @gunvulture For smooth regions, want larger DCTs; near edges, smaller ones. H265 has 4x4 .. 32x32 DCTs.
-
-
Replying to @rygorous
@rygorous@cmuratori@nothings @gunvulture Sounds like what you actually want is a sparse mipmap arrangement, where each node is a 4x4 DCT.2 replies 0 retweets 0 likes -
Replying to @tom_forsyth
@tom_forsyth@rygorous@nothings @gunvulture What you actually want is someone to figure out the real "visual transform" :)2 replies 0 retweets 0 likes -
Replying to @cmuratori
@cmuratori@rygorous@nothings @gunvulture I apologize for my interest.1 reply 0 retweets 0 likes -
Replying to @tom_forsyth
@tom_forsyth@rygorous@nothings @gunvulture Your apology is not accepted. GO FIGURE OUT THE TRANSFORM :P1 reply 0 retweets 0 likes -
Replying to @cmuratori
@tom_forsyth@rygorous@nothings @gunvulture (I've been bitching to@jeffatrad for years about how we need to get past the DCT...)1 reply 0 retweets 0 likes -
Replying to @cmuratori
@cmuratori@rygorous@nothings @gunvulture@jeffatrad I am kinda surprised. Seems like a metric assload of PhDs & patents waiting to be had.1 reply 0 retweets 0 likes -
Replying to @tom_forsyth
@tom_forsyth@cmuratori@nothings @gunvulture@jeffatrad It's as if there should be 50yrs worth of research that neither of you have read!1 reply 0 retweets 0 likes -
Replying to @rygorous
@rygorous@cmuratori@nothings @gunvulture@jeffatrad Damn you and your facts, Mr. Giesen.2 replies 0 retweets 0 likes -
Replying to @tom_forsyth
@tom_forsyth@cmuratori@nothings @gunvulture@jeffatrad That's right. Killing kittens and harvesting their visual cortex FOR SCIENCE.3 replies 1 retweet 0 likes
@rygorous @tom_forsyth @nothings @gunvulture @jeffatrad They did that for the central pattern generator, too (dead cat on treadmill)...
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.