Audacity & JACK: quest for the ultimate configuration
Posted: Wed Mar 04, 2009 11:05 pm
Since upgrading to 1.3.5, I've had to start using JACK to avoid terrible, terrible skipping while recording. There is a large visual lag upon zooming in, or following the cursor, which is a nuisance ... anyone know the cause of this? More troubling, obviously, is the record quality, so I've set out to remedy this by configuring JACK.
I tried many different settings, and am now at a 232ms latency rating. Running Ubuntu, I recorded audio into the conputer as a stereo mix from a tape sent into a USB mixer. I recorded for a total of 90 minutes, and had the system playing back the audio as it was recording it. Also, I was running Baobab, a disk usage analyzer, which draws a visual representation of disk usage, a system intensive scanning process. I had no xruns during this test, a test which I designed to attempt a simulation of the system's most stressed state, while striving for clean recording.
While the test worked, I had a jam the other night, and encountered 2 xruns over the span of 1 1/2 hours. I was shocked! Why, why, WHY? Audacity and JACK were the only things running. Maybe my battery was charging (it's a laptop), affecting something? That shouldn't be. And something else senseless, the xruns are reported as some ridiculous number, like "xrun occured of at least 192383402834.2439ms." That's gotta be a few days, or something ...
My question is: what scenario could possibly cause xruns to occur that wouldn't have occured during the resource-intensive test?
I tried many different settings, and am now at a 232ms latency rating. Running Ubuntu, I recorded audio into the conputer as a stereo mix from a tape sent into a USB mixer. I recorded for a total of 90 minutes, and had the system playing back the audio as it was recording it. Also, I was running Baobab, a disk usage analyzer, which draws a visual representation of disk usage, a system intensive scanning process. I had no xruns during this test, a test which I designed to attempt a simulation of the system's most stressed state, while striving for clean recording.
While the test worked, I had a jam the other night, and encountered 2 xruns over the span of 1 1/2 hours. I was shocked! Why, why, WHY? Audacity and JACK were the only things running. Maybe my battery was charging (it's a laptop), affecting something? That shouldn't be. And something else senseless, the xruns are reported as some ridiculous number, like "xrun occured of at least 192383402834.2439ms." That's gotta be a few days, or something ...
My question is: what scenario could possibly cause xruns to occur that wouldn't have occured during the resource-intensive test?