Prev: Low Capacitance Tranzorb in the 10-30V range
Next: Is this an ink flow sensor? (Ink Jet printer)
From: Jan Panteltje on 28 Jan 2010 08:26
On a sunny day (Thu, 28 Jan 2010 10:34:51 +0000) it happened Dirk Bruere at
NeoPax <dirk.bruere(a)gmail.com> wrote in <7sd7hlFhn4U1(a)mid.individual.net>:

>Any idea how long the delay is?

From 10 or so milli seconds to maybe a hundred, sounds like a real echo.
There is one issue with this, if you use 2 sound cards, each with their own crystal,
you get irregularities if for example data is read faster (say at 44100) then written (also at 44100).
If you have a sound card with multiple channels on it, running from the same crystal, it should be OK though.
At least this is my theory of what happens, may well be wrong...

For example I have a Creative 32 PCI, and a Soundblaster Live.
The Live resamples everything, and has 2 channels, I think on the same clock.
I have seen problems using the SB PCI32 as input and the SB Live as output.

IIRC I could send from one channel of the SB live to the other, but thanks to Alsa sound system
now installed and replacing OSS sound drivers this no longer seems to work.
I had /dev/dsp0 (PCI32) and /dev/dsp1 and /dev/dsp2 (SB Live), and
now Alsa no longer seems to see the second dsp device on the Live.
So to make a long story short, it may or may not work reliable on your system.
Of course it will always work reliable when playing from a wave or mp3 file.

I just tested some things, nice echo bathroom sound if I feed output of both sound cards to a speaker...
xpequ -u plughw:0,0 -t plughw:1,0 (PCI32 via equaliser to SB Live).
But this way I get an alsa buffer overflow after a while.

That seems to depend on system load too, been running OK now for 10 minutes.
So YMMV.
Of course you could make a hardware equaliser too :-)


First  |  Prev  | 
Pages: 1 2
Prev: Low Capacitance Tranzorb in the 10-30V range
Next: Is this an ink flow sensor? (Ink Jet printer)