|On Tue, 27 Nov 2001 00:02:36 +0100 (CET), _-¯-_ wrote: | |> |let me know how it works... |> |>hm, can't seem to get a clean signal out of it so far |>have been trying on a proper machine now, at 48000, 128kbit stereo but |>still those crackels (on frame boundaries?) | |i think i have the same problem. please listen to |http://radiostudio.org:8000/puredata | |i'm using the linux version.
i also checked the date's of the files i was using yesterday and it seems they were the latest versions (nov 26, 14.20) right from the start ..
-- > < d V . O r G
Well, I don't have Linux to test it but the Windows version works without crackles. So it can't (?) be a problem concerning the audio buffers of shoutcast~, I think, since they use the same code. Probably a problem with sending data via TCP on Linux? shoutcast~ should display a message in case data is not sent correctly (lines 214 to 229 in shoutcast~.c). Does 'send' on Linux give the same return value (i.e. number of bytes sent)?
To test if crackles are related to buffer size you could try to compile with say double buffer size (MY_MP3_MALLOC_IN_SIZE in line 64). Then crackles should occur half that often... If they still occur in same frequency it might be a problem of mp3 encoding which is done differently on linux. Or try to increase x->x_lamechunk = 1024 in line 884 to 2304 (which is the amount of data we send LAME on win when encoding in 48kHz, 224kbit).
Olaf
-¯- schrieb:
|On Tue, 27 Nov 2001 00:02:36 +0100 (CET), _-¯-_ wrote: | |> |let me know how it works... |> |>hm, can't seem to get a clean signal out of it so far |>have been trying on a proper machine now, at 48000, 128kbit stereo but |>still those crackels (on frame boundaries?) | |i think i have the same problem. please listen to |http://radiostudio.org:8000/puredata | |i'm using the linux version.
i also checked the date's of the files i was using yesterday and it seems they were the latest versions (nov 26, 14.20) right from the start ..
-- > < d V . O r G
I'll take a look at this problem quite soon... but I can't today, one of my friend crashed his Window$ system.
later,
Yves/
----- Original Message ----- From: "Olaf Matthes" olaf.matthes@gmx.de To: "-¯-" jdl@xdv.org Cc: "smoerk" smoerk@gmx.de; "pd-list" pd-list@iem.kug.ac.at; "Yves Degoyon" ydegoyon@free.fr Sent: Tuesday, November 27, 2001 3:43 PM Subject: Re: shoutcast~ lin
Well, I don't have Linux to test it but the Windows version works without crackles. So it can't (?) be a problem concerning the audio buffers of shoutcast~, I think, since they use the same code. Probably a problem with sending data via TCP on Linux? shoutcast~ should display a message in case data is not sent correctly (lines 214 to 229 in shoutcast~.c). Does 'send' on Linux give the same return value (i.e.
number
of bytes sent)?
To test if crackles are related to buffer size you could try to compile
with
say double buffer size (MY_MP3_MALLOC_IN_SIZE in line 64). Then crackles should occur half that often... If they still occur in same frequency it might be a problem of mp3 encoding which is done differently on linux. Or try to increase x->x_lamechunk = 1024 in line 884 to 2304 (which is the amount of data we send LAME on win when encoding in 48kHz, 224kbit).
Olaf
-¯- schrieb:
|On Tue, 27 Nov 2001 00:02:36 +0100 (CET), _-¯-_ wrote: | |> |let me know how it works... |> |>hm, can't seem to get a clean signal out of it so far |>have been trying on a proper machine now, at 48000, 128kbit stereo
but
|>still those crackels (on frame boundaries?) | |i think i have the same problem. please listen to |http://radiostudio.org:8000/puredata | |i'm using the linux version.
i also checked the date's of the files i was using yesterday and it seems they were the latest versions (nov 26, 14.20) right from the start ..
-- > < d V . O r G
-
_-¯-_ -
Olaf Matthes -
Yves Degoyon