[CELT-dev] Octagon Coding

[Benjamin M. Schwartz]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Benjamin M. Schwartz wrote:
> I
> was particularly disappointed to see no measurable benefit in terms of RMS
> error for N=3, though the maximum error is improved significantly.

Oops. I meant L=3. (I think I tested N=2^24 and 2^32.)

- --Ben
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)

iEYEARECAAYFAknBaS4ACgkQUJT6e6HFtqQ6eQCgltNrN+u6HQqjATT3PzH2pdSd
54AAnRfn0dw+1RS+RXoNNr0MY1gqlDec
=HR4e
-----END PGP SIGNATURE-----