On Sat, Aug 02, 2008 at 08:07:44PM +0200, Mathieu Malaterre wrote: > Hi, > > Reading draft ISO/IEC FCD 14495-1:1997(E), page 38, they are using > the value '7' for computing T3, see attached patch. Indeed, a silly typo and since nobody has tried lossy mode for bitdepths < 7, it wasn't uncovered earlier. Thank you for spotting. > Thanks, > -- > Mathieu