Hi there,

if I do a simple

[adc~] [phasor~ 44100]
\ /
\ /
\ /
\ /
\ /
[ samphold~]
|
|
[dac~]

, the original sound gets 'stained' with noise (white, pink, whatever ?).
To me this should theorically lead to a 'bypass', but doesn't.

Does someone know how I could get rid of this problem ? I noticed this learning how to bitcrush and downsample audio, as the downsampling operation seemed way too noisy (and this can't be filtered out because widespread over freq. domain).

Thanx,

Nau

To me, this should lead to silence, and that's what I get. [samphold~] samples every time the [phasor~] decreases (i.e. at the end of each cycle). But 44100 Hz can't be reproduced at a sample rate of 44100. This frequency gets aliased to DC. And since every harmonic of [phasor~] is a multiple of the fundamental, they all get aliased to DC as well, leaving you with a flat signal.

Hi,

Sorry Maelstorm, I was wrong. The noise was from another source in my patch (bad debug scheme).

But that's strange you get zero. I get something distorted, but can clearly recognize the initial sample. Moreover, when messages converted to signals control the [samphold~] objects, this sound artifact appears and disappear randomly when toggling from 44099 to 44100 (see attached patch).

Of course you remembered me that I am playing with the limit, so a 44099 value will be my choice for control knob upper limit.

Thanx,

Nau

http://www.pdpatchrepo.info/hurleur/downsampling_noise.pd

I tried your patch and it still works as it should for me--no sound at 44100. I'm not really sure what could be causing the problem, other than maybe your patch's sample rate is not actually 44100, but I'm sure you checked that...right?

I was expecting such an answer. My Pd samplerate is 44100 as shown in the menu. Feels like we'll never know the end of the story ...
I work on TangoStudio, rt distro based on Ubuntu Lucid.

Thanx anyway.