[daala] Why not upgrading Daala range coding with much faster rANS or tANS?

Timothy B. Terriberry tterribe at xiph.org
Fri Aug 14 07:58:47 PDT 2015


Jarek Duda wrote:
> rANS: similar to range coding, but many times faster: using single
> multiplication/symbol, more convenient frequency handling, better for
> SIMD, e.g. https://github.com/rygorous/ryg_rans

We're aware of the work. What we have is even more flexible 
(conveniently handles non-power-of-two frequency counts), has no 
multiplies at all, and can also take advantage of SIMD.

We'd be happy to see someone experimenting with this, but the core team 
is focused on improving quality per bit, and rANS is unlikely to be much 
help there.


More information about the daala mailing list