Talk:Tutorial - Recording Computer Playback on Linux

From Audacity Development Manual
Revision as of 20:51, 15 December 2011 by Stevethefiddle (talk | contribs) (→‎PulseAudio: added comments from Bruno)
Jump to: navigation, search

Steve 16Sept11: Most (by popularity) Linux desktop systems use PulseAudio as the default sound system. In the past PulseAudio has been buggy on many systems, but is very much improved these days, so I propose that the first method in the Recording Computer Playback on Linux tutorial should be to use PulseAudio.

In cases where PulseAudio is the default sound system, I don't think it is necessary to go into a description of what PulseAudio is (you don't need to explain the engine management system within instructions for how to drive a car). So here's a proposal of some revised text (taken from this post on the forum):


PulseAudio

If the Device toolbar lists pulse then you can record computer playback using PulseAudio.

You will need the Pulse Audio Volume Control application (pavucontrol), so if not already installed, install it now. If available, it may also be useful to install the PulseAudio Device Chooser (padevchooser) for easy access to the Pulse Audio Volume Control.

  • Open Audacity and select pulse as the recording device.
  • Click on the Recording meter to activate it, then open Pulse Audio Volume Control. Audacity should be listed in the Recording tab of pavucontrol.
  • Set Audacity (in pavucontrol) to record from the "Monitor" of your sound card. The available options are typically Internal Audio Analog stereo and Monitor Internal Internal Audio Analog Stereo, in which case you would select the latter.

Audacity will now record whatever is playing through the sound card.

For a more detailed description, see Stereo Mix with Ubuntu 10 on the Audacity wiki.


I had a look at it and here are my thoughts:

1. Using a loopback cable is shown as the first option on top of the page. I hope this doesn't give the idea that this is the preferrable solution. I think this should be a last resource solution when everything else fails and I'd put it at the end.

2. The Alsamixer section looks a bit too long and specially the first 3 or 4 paragraphs sound a bit confusing and contain some unnecessary information.

3. ALSA PCM file: to make it more newbie proof, it should be more clear that setting the variable FLASH_ALSA_DEVICE=teeraw on the command line won't affect already running instances of the web browser, neither will affect freshly started ones unless they're started from the command line. The example given "FLASH_ALSA_DEVICE=teeraw firefox" will start firefox from the command line if and only if there's a binary or link/script named firefox in the path. In some distros this might not be true and if the user uses a different browser this doesn't apply either.

4. ALSA PCM file: either I missed or it's not said on the text that the "recorded" sound will go to a file in the /tmp folder and that (in the first example) it is saved as "raw PCM".

5. PulseAudio: it's the default on Ubuntu, but I'm so sure about Debian, though I haven't made a fresh install of the latest release so it might have changed... I also don't like the second phrase where it says PulseAudio works like a plugin for ALSA.

6. PulseAudio: pulse loopback module should be mentioned here.

7. OSS: not sure if anyone is going to write this section... Mostly everyone is using ALSA on Linux nowadays and OSS support on Audacity (AFAIK) is very limited. Is it worth writing such a section?


Cheers, Bruno

Retrieved from "http://alphamanual.audacityteam.org/m/index.php?title=Talk:Tutorial_-_Recording_Computer_Playback_on_Linux&oldid=13647"