Re: SV8 samplerate support

Hello,

I'll add rough translations of the parts to which I respond.

> >> if (92017 <=3D3D samplerate) return 96000;
> >> if (75132 <=3D3D samplerate) return 88200;
> >> if (55426 <=3D3D samplerate) return 64000;
> >> if (46009 <=3D3D samplerate) return 48000;
> >> if (37566 <=3D3D samplerate) return 44100;
> >> if (27713 <=3D3D samplerate) return 32000;
> >> if (23004 <=3D3D samplerate) return 24000;
> >> if (18783 <=3D3D samplerate) return 22050;
> >> if (13856 <=3D3D samplerate) return 16000;
> >> if (11502 <=3D3D samplerate) return 12000;
> >> if ( 9391 <=3D3D samplerate) return 11025;
> >> return 8000;
> >>
> >> Well, I know this is not really important  but I can't really find=3D=
20
> >> anything else to somment on yet.
> >>
> > =3D20
> >
> Interessant, identisch aussehenden Code hatte ich mal zu Lame=3D20
> hinzugef=3DFCgt, in den 2 Monaten, in denen ich Schreibzugriff hatte=3D=
20

(trans: Frank wrote almost identical code for LAME sometime)

Well, this exact code has not been part of the AAC standard for very =3D
long
(since 2001 I think), maybe some people are watching your work :)=3D20

> A typical tool to enhance huffman coding is grouping of=3D20
> samples. Often this result in better compression than=3D20
> arithmetic coding, because this also uses intersymbol correlation.

Ah yes, I guess you already defined this in the current SV8 specs.

> Hybrid-Codecs:
>=3D20
> ich habe mal dar=3DFCber nachgedacht, wann man sie wirklich=3D20
> braucht. Ich habe
> *keine*
> *einzige* sinnvolle Anwendung gefunden. das klingt so hart,=3D20
> da=3DDF ich es noch nicht gesagt habe.

(trans: can't think of any reason where a hybrid codec would be useful)

>=3D20
> Wesentlich interessanter ist es, statt "Raid-0" mit=3D20
> Audio-Daten ein Pseudo-"Raid 1" zu spielen. Fehler in=3D20
> Lossless files mit Hilfe von Lossy-Files zu kaschieren.
>=3D20
> OptimFrog --repair --lossy Axel_F.mpc --source Axel_F.ofr=3D20
> --destination Axel_F.wav

Yes, this was exactly my idea. Code the lossy coding error using =3D
lossless
coding, but what about the decoder accuracy of the lossy codec?

> PS: Erschreckend ist, da=3DDF es bei all diesen Fragen im=3D20
> wesentlich um alles m=3DF6gliche ging, nur nicht um die=3D20
> Hauptaufgabe einer Kodierung. Alles im wesentlichen Features,=3D20
> bei denen ich mir sicher bin, sie nie zu brauchen, die aber=3D20
> nach vielk=3DF6pfigen Programmierteams schreien, die sich 10=3D20
> Stunden pro Tag um das Thema k=3DFCmmern k=3DF6nnen.

(trans: all talk upto now only about features of the codec, not about =3D
the
actual codec method)

Ok, that's true, but what is there still to discuss? The bitstream =3D
format
seems about set, the coding methods seem to be determined. I don't see =3D
any
points that still require discussion. If there's still anything to =3D
discuss
about, you should name them :)

Greetings,

Menno


http://mpc.corecodec.org