> We use Sfront with DirectSound and are now having problems with =
> processorpower (PIII 500Mhz), resulting in ticks in our audio.
First and foremost, I rendered it, and it really sound like cool
music!
Some comments:
-- I rendered it into a WAV file, then played it back, and there
were clicks on my soundcard, which is not surprising, since 48000
sampling rate with my soundcard (Creative PCI 128) using Linux
introduces clicks during the resampling. I set the sampling rate
in the coolsound.saol file to 44100 kHz, rerendered, and the
clicks went away. You might try this experiment on your system
(both rendering into a file at 48000 and 44100 and listening,
and rendering using Directsound) to make sure the problem is
what you think it is, i.e make this chart:
Clicks? WAV Directsound
--------------------------------------
44100 | | |
--------------------------------------
48000 | | |
--------------------------------------
And see which boxes have a yes and which have a no.
-- I used the -aout linux option on my 500Mhz PIII under linux
to play the file out as it was computing (the equivalent of your
-aout dsound) and it played out with no clicks. So, both the
raw amount of processor and the variance in computation needs are
sufficient to stream out audio using sfront 0.53 and its (very)
primitive real-time model ...
-- I rendered the 20-second piece to a WAV file and timed it
using the UNIX time command:
% make timing
/usr/bin/time -p ./sa
real 7.72
user 7.66
sys 0.06
Verifying that its indeed not computationally not overbearing. All
these indicate to me that the 48000 -> 41000 experiment is worth
trying first to see if it solves it. In addition, though, I'll spend
a few minutes looking over the code, and send along some efficiency
suggestions in a follow-on note,
--john lazzaro
This archive was generated by hypermail 2b29 : Mon Jan 28 2002 - 11:46:36 EST