Hi,
I 've been searching the internet for two days and i couldn't find any satisfying answer, so I thought I'd ask here... I'm running PD on my Raspberry Pi, and I need to record Pd's output. I can't use the standard writesf~ because I lack disk space, and oggwrite~ doesn't work well on the RPi (it records fine but Pd's output is trashed).
I recently discovered that i could record something in raw with ALSA using arecord, and pipe this into oggenc, or flac, or whatever I need to record the file. Something like: arecord <parameters> raw | oggenc <parameter> file.ogg
The thing is that the RPi has no capture device (the single PCM doesn't seem to be capable of capture).
A lot of things seem to be possible using a .asoundrc configuration file for ALSA, but I don't know how to use it. What I would want is to copy the PCM output so that ALSA can recognize this copy as an input in a record. Note that I need to be able to listen to the output while recording.
Do you thing this is possible at all?
Cheers,
Pierre.
Or is there a way to directly pipe Pd's audio output into another program? Too bad JAC doesn't work on the Pi...
2012/11/24 Pierre Massat pimassat@gmail.com
Hi Hans,
Thanks for your reply. I can't use writesf~ alone because i want to save some disk space, wav is too big. Is there a way i can run Pd, and then pipe its output to another program to record it whenever I want? I don't want Pd to continuously write to a file because i'm running it for hours.
Cheers,
Pierre.
2012/11/25 Hans-Christoph Steiner hans@at.or.at
You could write a patch that records 16-bit mono .wav, at as low a sample rate as you'd like, to .wav with writesf~ . . . more or less copy-pasting from the help page. Recording to something besides .wav, especially into a separate application (which just because it's a separate application will need more processing power) in realtime, needs encoding (and thus processing power and RAM) as you write, which is going to be a bigger problem than disk space on a raspberry pi, I'd think. How long are you planning to record, and what is your file size limitation?
Someone correct me if I'm wrong, for my information as much as his, Greg
On Sun, Nov 25, 2012 at 2:48 PM, Pierre Massat pimassat@gmail.com wrote:
Hi,
Yes, recording to mono would be a good idea. I don't know to what extend the quality would suffer from a sample rate reduction though. I've read that it would only cut higher frequencies. Another problem is that I'm not sure i'll be able to read these file (say at 22050) in a Pd patch (at 44100).
What you're saying about CPU load and RAM makes sense, but i'd like to try anyway, just to know whether it'd work or not. Somebody on the RPi forum suggested that I use Pulseaudio. I don't even know if Pd works with Pulse. Anyways...
Thank you for your reply.
Cheers,
Pierre.
2012/11/25 Greg Schroeder gmschroeder@gmail.com
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 2012-11-27 09:38, Pierre Massat wrote:
not with [readsf~] (if you care about micky-mouse voices). but then, you cannot read mp3's with [readsf~] either.
and you could change Pd's sampelrate to 22050.
and you could convert your 22.5kHz mono wav-file to mp3 (or 192kHz 8-channel wav) after recording and before playback (the original question was about recording long soundfiles; now you are talking about playback; it's a bit hard to tell what you want to achieve)
as for the quality: if you don't want to do the maths, listen to it. if it sounds good, then it's probably good. it all depends very much on the sounds you are actually using. e.g. 8bit might sound horrible in general, but if you are doing C64-style music anyhow, there might not be so much difference.
fgasmdr IOhannes