[vorbis-dev] FLOAT_LOOKUP version of lsp_to_curve
TIMMERMANS ANDRE
Andre.Timmermans at sbs.be
Wed Jun 26 23:50:45 PDT 2002
> -----Original Message-----
> From: TIMMERMANS ANDRE
> Sent: Wednesday, June 26, 2002 8:47 AM
> To: 'vorbis-dev at xiph.org'
> Subject: RE: [vorbis-dev] FLOAT_LOOKUP version of lsp_to_curve
>
> > -----Original Message-----
> > From: TIMMERMANS ANDRE
> > Sent: Wednesday, June 26, 2002 8:40 AM
> > To: vorbis-dev at xiph.org
> > Subject: [vorbis-dev] FLOAT_LOOKUP version of lsp_to_curve
> >
> > It seems to me that in this version of the function using
> >
> > float p=.7071067812f;
> > float q=.7071067812f;
> >
> > will just leads after applying all the products and squaring
> > to p and q having a common factor 0.5 which can be extracted
> > from the formulaes.
> >
> > So I propose to replace:
> >
> > float p=.7071067812f;
> > float q=.7071067812f;
> > float w=vorbis_coslook(wdel*k);
> > float *ftmp=lsp;
> > int c=m>>1;
> >
> > by:
> >
> > float w=vorbis_coslook(wdel*k);
> > float p=lsp[0]-w;
> > float q=lsp[1]-w;
> > float *ftmp=lsp + 2;
> > int c=(m>>1) - 1;
> >
> > and:
> >
> > vorbis_invsq2explook(qexp+m)
> >
> > by:
> >
> > vorbis_invsq2explook(qexp+m-1)
> >
> I forget to add that strictly speaking the looping conditions should changed on
>
> do{
> q*=ftmp[0]-w;
> p*=ftmp[1]-w;
> ftmp+=2;
> }while(--c);
>
> but if I understand it correctly m is always greater or equal than 8
> which means the loop will always be executed at least once.
>
I have mad a confusion with windowing/mdct, from floor one I see
the 0 < m < 256. But is really m = 1 valid because in that case
not only my changes are problematic but the integer variant
of lsp_to_curve is incorrect.
André
--- >8 ----
List archives: http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body. No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.
More information about the Vorbis-dev
mailing list