[FFmpeg-soc] [soc]: r3381 - aacenc/aacpsy.c

[Robert Swain]

2008/8/17 Diego Biurrun <diego at biurrun.de>:
> On Sun, Aug 17, 2008 at 01:36:04PM +0300, Kostya wrote:
>> On Sun, Aug 17, 2008 at 11:36:20AM +0200, Diego Biurrun wrote:
>> > On Sun, Aug 17, 2008 at 10:24:28AM +0200, kostya wrote:
>> > >
>> > > Log:
>> > > Break long lines of code
>> >
>> > Ummm,...
>> >
>> > > --- aacenc/aacpsy.c       (original)
>> > > +++ aacenc/aacpsy.c       Sun Aug 17 10:24:28 2008
>> > > @@ -687,7 +696,6 @@ static void psy_3gpp_process(AACPsyConte
>> > >              IndividualChannelStream *ics = &cpe->ch[ch].ics;
>> > > -            prev_scale = -1;
>> > >              for(w = 0; w < ics->num_windows*16; w += 16){
>> > > @@ -695,9 +703,6 @@ static void psy_3gpp_process(AACPsyConte
>> > >                      cpe->ch[ch].sf_idx[w+g] = (int)(2.66667 * (log2(6.75*band->thr) - log2(band->ffac)));
>> > > -                    if(prev_scale != -1)
>> > > -                        cpe->ch[ch].sf_idx[w+g] = av_clip(cpe->ch[ch].sf_idx[w+g], prev_scale - SCALE_MAX_DIFF, prev_scale + SCALE_MAX_DIFF);
>> > > -                    prev_scale = cpe->ch[ch].sf_idx[w+g];
>> >
>> > ?
>>
>> This was a case of breaking long lines into null.
>> And this chunk is not needed anyway.
>
> Please at least mention this in the commit message then.  If you say
> "break long lines" I expect nothing but cosmetics.

Agreed. Strictly, Kostya mixed cosmetics and functional changes so he
shouldn't have committed both at the same time.

Rob