Re: FLL change

On Sat, 2006-07-29 at 21:25 +0200, Robert Jonsson wrote:
> Hi,
> 
> > >
> > >     What was the problem that he was having when he slammed the limiter?
> >
> > The original issue was that high transients caused the limiter to pump
> > with horrible results.
> >
> > First message:
> > http://article.gmane.org/gmane.comp.audio.jamin.devel/1405
> >
> > Second message:
> > http://article.gmane.org/gmane.comp.audio.jamin.devel/1409
> 
> Interesting, I didn't see the patch in the links, is it available somehwere, 
> I'd like to try it?

Sorry, I've been handling this off-list in irc so the communication to
jamin-devel has been a bit... lacking. sorry.

The patch is really quite simple. This is done to
fast_lookahead_limiter_1913.xml in swh-plugins (0.4.14)

line 89-90 (old code):
    const float this_delta = (max / chunks[p] - atten) /
                             ((float)i * fs * 0.0005f + 1.0f);


line 89-90 (new code):
    const float this_delta = (max / chunks[p] - atten) /
                             ((float)(i+1) * fs * 0.0005f + 1.0f);
                                     ^^^^^

Disclaimer: line numbers may not match. They are from 0.4.14 originally,
but they might have changed when i did testing & debugging.

This may cause slight clipping due to very high and fast peak transients
(fast an is only for a few samples), but this clipping should be
virtually inaudible because the peaks which get past the limiter are so
very tiny in length.

-- 
Sampo Savolainen <[email protected]>


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV