Subject: Re: [ecasound] ecasound resample problem
From: Aaron Heller (heller_AT_AI.SRI.COM)
Date: Sun Jul 25 2004 - 11:28:32 EEST
Kamil Wencel wrote:
> Yeah, even with the latest secret rabbit code I get these clicks , too.
> But, and here comes the itch :
> Not always. I am going the other direction ( 44.1 -> 96 )
> And I hear the same clicks mostly only when the track
> is near clipping or parts of the music are very very dynamic.
> I am still rather clueless what could be behind all that,
> so any ideas would be gladly appreciated.
It sounds to me as if a sample (or two) is being dropped periodically.
This would make the amplitude of the click be proportional to the
instantaneous slew rate (dv/dt) of the audio, which is consistent with
your observations. I'd guess that the clicks are there all the time,
but you can only hear them during loud or dynamic parts of the music.
In fact, during silence, you would not hear it at all.
This is why I picked sine waves for the tests I ran. It is consistent
amplitude and they don't mask the clicks. The clicks are louder with
the higher frequency (higher slew rate), which is consistent with the
dropped sample conjecture.
This archive was generated by hypermail 2b28 : Sun Jul 25 2004 - 11:28:48 EEST