You can build up any function by scaling and adding basis functions, but is there an analogous way to represent things as combinations of phase offsets of something? I am working with spectrograms, but encoding details favor a wrap-around value instead of just amplitude.
-
Show this thread
-
Replying to @ID_AA_Carmack
i’m aware of this domain of mathematics which was useful for image compressionpic.twitter.com/QcEyqVM6HU
1 reply 0 retweets 3 likes -
Replying to @DanielleFong @ID_AA_Carmack
not sure what you mean by “encoding details favor a wrap-around value instead of just amplitude”?
1 reply 0 retweets 2 likes -
Replying to @DanielleFong
The only encoding (that I can think of) that can have balanced statistics for every individual bit while providing a monotonic hamming distance between values involves the maximum being one step away from the minimum, sort of like a gray code.
1 reply 0 retweets 3 likes -
Replying to @ID_AA_Carmack
oh! ok, interesting! think i see what you mean
1 reply 0 retweets 0 likes
wavelet image compression usually does waaay better in terms of compression than like, fft based jpg (iirc like 10x) on real images, in think because of the highly nonrandom stats on real images, but this is only based on impressions not deep knowledge of the techniques!
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.
