Re: [AD] [SCM] Allegro branch 5.1 updated. 5.1.3-21-g1a7e633

[Peter W. <nov...@gm...>]

On Sun, 2 Sep 2012 11:29:34 +0200, Evert Glebbeek <eg...@dd...> wrote:
> On 2 Sep 2012, at 3:27 , Peter Wang wrote:
> > commit cd66cd6ca44bd2177babb1b27589ec7dc01fbe3d
> > Author: Peter Wang <tj...@us...>
> > Date:   Sat Sep 1 23:35:11 2012 +1000
> > 
> >    audio: Add cubic interpolation (for float mixers only).
> > 
> > commit 809a75c10d21c9de1c82487db1d8d2ac37a6d959
> > Author: Peter Wang <tj...@us...>
> > Date:   Sat Sep 1 18:51:30 2012 +1000
> > 
> >    audio: Support linear interpolation for 16-bit mixers.
> 
> In a completely different context (not related to audio) I recently got rid of cubic and linear interpolation and switched to monotonic cubic interpolation. I wanted to get rid of the linear interpolation because of lack of smoothness and the cubic interpolation because of spurious oscillations in the output.
> 
> The method is extremely simple, described in http://adsabs.harvard.edu/abs/1990A%26A...239..443S. Is this something that is worth looking into for Allegro?

Yes, worth a shot.  I can see the overshoots clearly with a sawtooth wave.
I'll try it tomorrow or so.

Peter