On 2012-01-02, Paulo da Silva <p_s_d_a_s_i_l_...@netcabo.pt> wrote: > Em 30-12-2011 11:23, mblume escreveu: >> Am Fri, 30 Dec 2011 07:17:13 +0000 schrieb Paulo da Silva: >> Alternatively you might just generate (t,signal) samples, write >> them to a file and convert them using "sox" (under Linux, might >> also be available under Windows) to another format.
That's how I'd do it. Sox can cope with raw samples, provided you tell it stuff like the sample-rate, sample-size, channels etc. E.g. from "man sox" sox -r 16k -e signed -b 8 -c 1 voice-memo.raw voice-memo.wav (Once it's in wav form the wav header contains that information.) > As much as I could understand at a 1st look you are > writing to a wav file and then play the file. > It would be nice if I could play directly the samples. "-d" as an output-file means the "default" output, which means it plays to your sound card. E.g.: sox -r 48k -e float -b 32 -c 2 input.raw -d The input file can be "-" so you can pipe directly to it: sox -r 44100 -e float -b 32 -c 2 - -d It should also run under windows and macos, and can generate lots of formats, see "man soxformat". It also has a "synth" effect which can generate simple waveforms easily. Sox is useful. Regards, Peter -- Peter Billam www.pjb.com.au www.pjb.com.au/comp/contact.html -- http://mail.python.org/mailman/listinfo/python-list
