Re: [MP3 ENCODER] spreading function buggy?

[Ambrose Feinstein]

> !            if(j!=3Di) temp1 =3D (cbval[j] - cbval[i])/.33;
> !            else     temp1 =3D 0;

as im too lazy to look at the context, i dont know if this is frequently
executed code... but, do most compilers optimize /.33 into *3.030303...?
i checked with gcc on sparc, and it apparently only does this when you
use -ffast-math.
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )