Hello,
I'm in the process of learning more about PD so I've been checking some of the patches that are part of Johannes Kreidler's excellent tutorial Programming Electronic Music in Pd ( http://www.pd-tutorial.com/ ). Today I was playing a bit with the sampler-big.pd which is a sound sampler with adjustable speed and start/end points. I noticed that as I made playback speed slower I started to hear quite noticeable aliasing. The patch uses [tabread4~] for reading the sound array. So my questions are: is this aliasing a particular problem of my system or is this an expected result? Is there anyway of avoiding it? I could post the patch here if needed.
Cheers,
Hector
On Sun, 24 May 2009 00:07:45 -0400 Hector Centeno hcengar@gmail.com wrote:
So my questions are: is this aliasing a particular problem of my system or is this an expected result?
It's an expected result. It isn't 'aliasing', but an effect called quantisation, which you will hear even in perfectly sampled data.
Is there anyway of avoiding it? I could post the patch here if needed.
[tabread4~] already goes some way towards avoiding it, by interpolating (going in-between the sample values) to make the playback smoother. Since you seem interested in the low frequencies found when slowing sounds down it might help to add a low pass filter that smoothes the signal even more.
Andy
Cheers,
Hector
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Andy Farnell wrote:
On Sun, 24 May 2009 00:07:45 -0400 Hector Centeno hcengar@gmail.com wrote:
So my questions are: is this aliasing a particular problem of my system or is this an expected result?
It's an expected result. It isn't 'aliasing', but an effect called quantisation, which you will hear even in perfectly sampled data.
Is there anyway of avoiding it? I could post the patch here if needed.
[tabread4~] already goes some way towards avoiding it, by interpolating (going in-between the sample values) to make the playback smoother.
Except Pd's 4-point interpolation is at best suboptimal...
http://lists.puredata.info/pipermail/pd-list/2008-06/062878.html
Try tabread4c~ or whatever and compare..
-
Andy Farnell -
Claude Heiland-Allen -
Hector Centeno