hi tim again
i just noticed, that we have been talking off-list during this thread. i post that to the list again (it might be interesting for others as well).
what you describe, shouldn't happen. if you use '-rate 96000', as mentioned in the help file of [writesf~], and the files still have '44100' in their headers, then it is definitely a bug of pd, respectively [writesf~]. i'm too lazy to test it now and here with aiff-files (you are writing to aiff-files, right?).
yo, i only know how to fix the sampling rate header field in wav-files, not in aiff-files unfortunately. the rate is written in the bytes 24-27 (when counting from 0) or in hex 0x00000018 -> 0x0000001B:
you will find there:
44 AC 00 00
(0x0000AC44 = 44100)
change this to:
00 77 01 00
(0x00017700 = 96000)
maybe it works similar for aiff-files, but i don't know.
roman
On Mon, 2007-07-02 at 16:44 -0400, Tim Flood wrote:
Hi Roman-
Yes, that is it: the audio files sound pitched/slowed down. When I open them in another application (soundhack, quicktime, etc.), they are always listed at 44.1k. So maybe PD is in fact RECORDING at 96kHz but for some reason the header has the wrong sr, as you suggested. I haven't used hex editors in the past, but will look into it as a means of fixing this problem.
Thanks again,
-Tim
______________________________________________________________ From: Roman Haefeli <romanhaefeli@yahoo.de> Reply-To: romanhaefeli@yahoo.de To: Tim Flood <timflood23@hotmail.com> Subject: Re: [PD] [writesf~] sample rate problem? Date: Mon, 02 Jul 2007 02:37:14 +0200 >hm.......wierd.... i don't have a solution at hand for this. but you >could do some checks. to make sure, that pd _really_ runs at 96kHz, >check that with the object [samplerate~]. it should output '96000' on >'bang', when pd runs at this rate. if so, than i would assume, that the >resulting files _are_ recorded with 96kHz, since it is pretty sure, that >[writesf~] does not do any conversion. >if pd runs at 96kHz and the resulting files don't sound pitched down, >then i really wonder, what makes you think, they have a sampling rate of >44.1kHz. >if the resulting files sound pitched down, then it probably is a header >issue (wrong sr tag in the header). this could be easily fixed with the >help of a hexeditor. > >yo, i hope that can help you isolate your problem a bit. > >roman > > > >On Sun, 2007-07-01 at 20:13 -0400, Tim Flood wrote: > > Hello Roman- > > > > Thank you for your quick response. Yes, my soundcard (a MOTU Ultralite > > firewire interface) is set up to use a sampling rate of 96kHz. The "clock > > source" is set to internal. I have included a "-r 96000" flag in my startup > > preferences. Still, every file is written at 44.1kHz. I am running > > Hans-Christoph Steiner's latest version of Pd-extended on Mac OS 10.4. If > > you think of any other possible causes for this problem, please drop me a > > line.... > > > > Best Regards, > > > > -Tim > > > > > > > > >From: Roman Haefeli <romanhaefeli@yahoo.de> > > >Reply-To: romanhaefeli@yahoo.de > > >To: Tim Flood <timflood23@hotmail.com> > > >CC: pd-list@iem.at > > >Subject: Re: [PD] [writesf~] sample rate problem? > > >Date: Mon, 02 Jul 2007 00:49:55 +0200 > > > > > >sorry, i was inaccurate. there *is* a '-rate' flag, but it doesn't cause > > >any conversion. as the help file states, it just affects the samplerate > > >entry in the header of the soundfile. this is probably not what you > > >want. therefore it is still worth reading my previous mail. > > > > > >roman > > > > > > > > >On Mon, 2007-07-02 at 00:40 +0200, Roman Haefeli wrote: > > > > hello tim > > > > > > > > there is no '-rate' flag for [writesf~], since it doesn't do any > > > > sampling rate conversion. it just uses the sampling rate of pd. it > > > > wouldn't make much sense to upsample for recording (e.g. 44100 -> > > > > 96000Hz) anyway. > > > > > > > > in other words: you need to run pd at the desired samplerate. if you > > > > want to record with 96000Hz, run pd with '-r 96000'. BUT this still > > > > doesn't always guarantee, that pd will effectively run with the desired > > > > sampling rate. sometimes it is needed to set the clock of the audio card > > > > to the desired rate by the preferences menu of the soundcard (sometimes > > > > the application just changes the samplingrate setting of the card on the > > > > fly). > > > > for example on my system, when i want to record at 96kHz, i need to set > > > > my hdsp-card to 96kHz, so that jackd runs at 96kHz, so that pd runs at > > > > 96kHz, so that i can record in pd with 96kHz. got it? > > > > > > > > if that doesn't help you already, please tell us, on what os you work > > > > and which soundcard/driver you use. > > > > > > > > roman > > > > > > > > > > > > On Sun, 2007-07-01 at 17:42 -0400, Tim Flood wrote: > > > > > Hello Everyone- > > > > > > > > > > I have written a simple patch (see below) designed for hard disk > > >recording > > > > > at 24-bit/96kHz. The -bytes flag sets the bit depth correctly, but no > > > > > matter what value I use to set the -rate flag, the file is always > > >created > > > > > with a sampling rate of 44.1kHz. (My soundcard is set up to work at > > >96kHz). > > > > > Is this a bug, or am I just missing something? Any help would be > > >greatly > > > > > appreciated! > > > > > > > > > > Best Regards, > > > > > > > > > > -Tim > > > > > > > > > > #N canvas 224 44 449 382 10; > > > > > #X msg 129 286 start; > > > > > #X msg 198 286 stop; > > > > > #X obj 43 336 writesf~ 2; > > > > > #X obj 153 71 time; > > > > > #X obj 16 173 pack s s s s s s; > > > > > #X obj 16 96 makefilename %u; > > > > > #X obj 26 147 makefilename %u; > > > > > #X obj 153 96 makefilename %u; > > > > > #X obj 161 122 makefilename %u; > > > > > #X obj 169 147 makefilename %u; > > > > > #X obj 16 71 date; > > > > > #X obj 21 122 makefilename %u; > > > > > #X obj 43 260 inlet~; > > > > > #X obj 100 260 inlet~; > > > > > #X obj 129 336 print record; > > > > > #X text 123 173 time & date stamp; > > > > > #X text 10 4 [diskRecorder~]; > > > > > #X text 13 18 writes an audio file to disk.; > > > > > #X text 43 244 main-l; > > > > > #X text 101 244 main-r; > > > > > #X obj 16 50 bng 15 250 50 0 empty empty empty 0 -6 0 10 -262144 -1 > > > > > -1; > > > > > #X text 231 223 open 24-bit/96kHz file; > > > > > #X msg 16 224 open -rate 96000 -bytes 3 -aiff \$1; > > > > > #X obj 16 198 makesymbol %s-%s-%s_%s-%s-%s; > > > > > #X text 36 48 push me; > > > > > #X connect 0 0 2 0; > > > > > #X connect 0 0 14 0; > > > > > #X connect 1 0 2 0; > > > > > #X connect 1 0 14 0; > > > > > #X connect 3 0 7 0; > > > > > #X connect 3 1 8 0; > > > > > #X connect 3 2 9 0; > > > > > #X connect 4 0 23 0; > > > > > #X connect 5 0 4 0; > > > > > #X connect 6 0 4 2; > > > > > #X connect 7 0 4 3; > > > > > #X connect 8 0 4 4; > > > > > #X connect 9 0 4 5; > > > > > #X connect 10 0 5 0; > > > > > #X connect 10 1 11 0; > > > > > #X connect 10 2 6 0; > > > > > #X connect 11 0 4 1; > > > > > #X connect 12 0 2 0; > > > > > #X connect 13 0 2 1; > > > > > #X connect 20 0 3 0; > > > > > #X connect 20 0 10 0; > > > > > #X connect 22 0 2 0; > > > > > #X connect 23 0 22 0; > > > > > #X connect 23 0 14 0; > > > > > > > > > > _________________________________________________________________ > > > > > Get a preview of Live Earth, the hottest event this summer - only on > > >MSN > > > > > http://liveearth.msn.com?source=msntaglineliveearthhm > > > > > > > > > > > > > > > _______________________________________________ > > > > > PD-list@iem.at mailing list > > > > > UNSUBSCRIBE and account-management -> > > >http://lists.puredata.info/listinfo/pd-list > > > > > > > > > > > > > > > > > > > > >___________________________________________________________ > > >Der frhe Vogel fngt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: > > >http://mail.yahoo.de > > > > _________________________________________________________________ > > Make every IM count. Download Messenger and join the im Initiative now. > > Its free. http://im.live.com/messenger/im/home/?source=TAGHM_June07 > > > > > > > > >___________________________________________________________ >Der frhe Vogel fngt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: http://mail.yahoo.de
Need a brain boost? Recharge with a stimulating game. Play now!
___________________________________________________________ Der frühe Vogel fängt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: http://mail.yahoo.de
i hacked around a bit in aiff-files. here the flag '-rate 96000' didn't have any effect on the resulting aiff-files as well. it seems like it is hardcoded to 44100 for aiff-files. can anyone confirm this by reading the source of [writesf~]?
anyway, i found out, how to fix aiff-files. the interesting bytes are 28-31, resp. 0x0000001C -> 0x0000001F
you'll find this:
40 0E AC 44
change this to:
40 0F BB 80
aiff-files are big endian, that is why the ordering is different here. the way you set a sampling rate is a bit strange in aiff-files, i just discovered by accident. you need actually to set it to 48000 (BB 80) and then set the previous bit to one (E -> F), so that the value gets doubled -> 96000. at least mplayer -identify recognizes it now as a 96000Hz-audiofile.
roman
On Tue, 2007-07-03 at 00:13 +0200, Roman Haefeli wrote:
hi tim again
i just noticed, that we have been talking off-list during this thread. i post that to the list again (it might be interesting for others as well).
what you describe, shouldn't happen. if you use '-rate 96000', as mentioned in the help file of [writesf~], and the files still have '44100' in their headers, then it is definitely a bug of pd, respectively [writesf~]. i'm too lazy to test it now and here with aiff-files (you are writing to aiff-files, right?).
yo, i only know how to fix the sampling rate header field in wav-files, not in aiff-files unfortunately. the rate is written in the bytes 24-27 (when counting from 0) or in hex 0x00000018 -> 0x0000001B:
you will find there:
44 AC 00 00
(0x0000AC44 = 44100)
change this to:
00 77 01 00
(0x00017700 = 96000)
maybe it works similar for aiff-files, but i don't know.
roman
On Mon, 2007-07-02 at 16:44 -0400, Tim Flood wrote:
Hi Roman-
Yes, that is it: the audio files sound pitched/slowed down. When I open them in another application (soundhack, quicktime, etc.), they are always listed at 44.1k. So maybe PD is in fact RECORDING at 96kHz but for some reason the header has the wrong sr, as you suggested. I haven't used hex editors in the past, but will look into it as a means of fixing this problem.
Thanks again,
-Tim
______________________________________________________________ From: Roman Haefeli <romanhaefeli@yahoo.de> Reply-To: romanhaefeli@yahoo.de To: Tim Flood <timflood23@hotmail.com> Subject: Re: [PD] [writesf~] sample rate problem? Date: Mon, 02 Jul 2007 02:37:14 +0200 >hm.......wierd.... i don't have a solution at hand for this. but you >could do some checks. to make sure, that pd _really_ runs at 96kHz, >check that with the object [samplerate~]. it should output '96000' on >'bang', when pd runs at this rate. if so, than i would assume, that the >resulting files _are_ recorded with 96kHz, since it is pretty sure, that >[writesf~] does not do any conversion. >if pd runs at 96kHz and the resulting files don't sound pitched down, >then i really wonder, what makes you think, they have a sampling rate of >44.1kHz. >if the resulting files sound pitched down, then it probably is a header >issue (wrong sr tag in the header). this could be easily fixed with the >help of a hexeditor. > >yo, i hope that can help you isolate your problem a bit. > >roman > > > >On Sun, 2007-07-01 at 20:13 -0400, Tim Flood wrote: > > Hello Roman- > > > > Thank you for your quick response. Yes, my soundcard (a MOTU Ultralite > > firewire interface) is set up to use a sampling rate of 96kHz. The "clock > > source" is set to internal. I have included a "-r 96000" flag in my startup > > preferences. Still, every file is written at 44.1kHz. I am running > > Hans-Christoph Steiner's latest version of Pd-extended on Mac OS 10.4. If > > you think of any other possible causes for this problem, please drop me a > > line.... > > > > Best Regards, > > > > -Tim > > > > > > > > >From: Roman Haefeli <romanhaefeli@yahoo.de> > > >Reply-To: romanhaefeli@yahoo.de > > >To: Tim Flood <timflood23@hotmail.com> > > >CC: pd-list@iem.at > > >Subject: Re: [PD] [writesf~] sample rate problem? > > >Date: Mon, 02 Jul 2007 00:49:55 +0200 > > > > > >sorry, i was inaccurate. there *is* a '-rate' flag, but it doesn't cause > > >any conversion. as the help file states, it just affects the samplerate > > >entry in the header of the soundfile. this is probably not what you > > >want. therefore it is still worth reading my previous mail. > > > > > >roman > > > > > > > > >On Mon, 2007-07-02 at 00:40 +0200, Roman Haefeli wrote: > > > > hello tim > > > > > > > > there is no '-rate' flag for [writesf~], since it doesn't do any > > > > sampling rate conversion. it just uses the sampling rate of pd. it > > > > wouldn't make much sense to upsample for recording (e.g. 44100 -> > > > > 96000Hz) anyway. > > > > > > > > in other words: you need to run pd at the desired samplerate. if you > > > > want to record with 96000Hz, run pd with '-r 96000'. BUT this still > > > > doesn't always guarantee, that pd will effectively run with the desired > > > > sampling rate. sometimes it is needed to set the clock of the audio card > > > > to the desired rate by the preferences menu of the soundcard (sometimes > > > > the application just changes the samplingrate setting of the card on the > > > > fly). > > > > for example on my system, when i want to record at 96kHz, i need to set > > > > my hdsp-card to 96kHz, so that jackd runs at 96kHz, so that pd runs at > > > > 96kHz, so that i can record in pd with 96kHz. got it? > > > > > > > > if that doesn't help you already, please tell us, on what os you work > > > > and which soundcard/driver you use. > > > > > > > > roman > > > > > > > > > > > > On Sun, 2007-07-01 at 17:42 -0400, Tim Flood wrote: > > > > > Hello Everyone- > > > > > > > > > > I have written a simple patch (see below) designed for hard disk > > >recording > > > > > at 24-bit/96kHz. The -bytes flag sets the bit depth correctly, but no > > > > > matter what value I use to set the -rate flag, the file is always > > >created > > > > > with a sampling rate of 44.1kHz. (My soundcard is set up to work at > > >96kHz). > > > > > Is this a bug, or am I just missing something? Any help would be > > >greatly > > > > > appreciated! > > > > > > > > > > Best Regards, > > > > > > > > > > -Tim > > > > > > > > > > #N canvas 224 44 449 382 10; > > > > > #X msg 129 286 start; > > > > > #X msg 198 286 stop; > > > > > #X obj 43 336 writesf~ 2; > > > > > #X obj 153 71 time; > > > > > #X obj 16 173 pack s s s s s s; > > > > > #X obj 16 96 makefilename %u; > > > > > #X obj 26 147 makefilename %u; > > > > > #X obj 153 96 makefilename %u; > > > > > #X obj 161 122 makefilename %u; > > > > > #X obj 169 147 makefilename %u; > > > > > #X obj 16 71 date; > > > > > #X obj 21 122 makefilename %u; > > > > > #X obj 43 260 inlet~; > > > > > #X obj 100 260 inlet~; > > > > > #X obj 129 336 print record; > > > > > #X text 123 173 time & date stamp; > > > > > #X text 10 4 [diskRecorder~]; > > > > > #X text 13 18 writes an audio file to disk.; > > > > > #X text 43 244 main-l; > > > > > #X text 101 244 main-r; > > > > > #X obj 16 50 bng 15 250 50 0 empty empty empty 0 -6 0 10 -262144 -1 > > > > > -1; > > > > > #X text 231 223 open 24-bit/96kHz file; > > > > > #X msg 16 224 open -rate 96000 -bytes 3 -aiff \$1; > > > > > #X obj 16 198 makesymbol %s-%s-%s_%s-%s-%s; > > > > > #X text 36 48 push me; > > > > > #X connect 0 0 2 0; > > > > > #X connect 0 0 14 0; > > > > > #X connect 1 0 2 0; > > > > > #X connect 1 0 14 0; > > > > > #X connect 3 0 7 0; > > > > > #X connect 3 1 8 0; > > > > > #X connect 3 2 9 0; > > > > > #X connect 4 0 23 0; > > > > > #X connect 5 0 4 0; > > > > > #X connect 6 0 4 2; > > > > > #X connect 7 0 4 3; > > > > > #X connect 8 0 4 4; > > > > > #X connect 9 0 4 5; > > > > > #X connect 10 0 5 0; > > > > > #X connect 10 1 11 0; > > > > > #X connect 10 2 6 0; > > > > > #X connect 11 0 4 1; > > > > > #X connect 12 0 2 0; > > > > > #X connect 13 0 2 1; > > > > > #X connect 20 0 3 0; > > > > > #X connect 20 0 10 0; > > > > > #X connect 22 0 2 0; > > > > > #X connect 23 0 22 0; > > > > > #X connect 23 0 14 0; > > > > > > > > > > _________________________________________________________________ > > > > > Get a preview of Live Earth, the hottest event this summer - only on > > >MSN > > > > > http://liveearth.msn.com?source=msntaglineliveearthhm > > > > > > > > > > > > > > > _______________________________________________ > > > > > PD-list@iem.at mailing list > > > > > UNSUBSCRIBE and account-management -> > > >http://lists.puredata.info/listinfo/pd-list > > > > > > > > > > > > > > > > > > > > >___________________________________________________________ > > >Der frhe Vogel fngt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: > > >http://mail.yahoo.de > > > > _________________________________________________________________ > > Make every IM count. Download Messenger and join the im Initiative now. > > Its free. http://im.live.com/messenger/im/home/?source=TAGHM_June07 > > > > > > > > >___________________________________________________________ >Der frhe Vogel fngt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: http://mail.yahoo.de
Need a brain boost? Recharge with a stimulating game. Play now!
___________________________________________________________ Der frhe Vogel fngt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: http://mail.yahoo.de
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Telefonate ohne weitere Kosten vom PC zum PC: http://messenger.yahoo.de
On Tue, 2007-07-03 at 00:49 +0200, Roman Haefeli wrote:
i hacked around a bit in aiff-files. here the flag '-rate 96000' didn't have any effect on the resulting aiff-files as well. it seems like it is hardcoded to 44100 for aiff-files. can anyone confirm this by reading the source of [writesf~]?
anyway, i found out, how to fix aiff-files. the interesting bytes are 28-31, resp. 0x0000001C -> 0x0000001F
you'll find this:
40 0E AC 44
change this to:
40 0F BB 80
two additions are needed:
- the posistion in the header is not fixed. bytes no. 28-31 is only
valid for files made with pd/[writesf~].
- 40 UV WX YZ seems to be kind of a floating point format, where UV is
the exponent part (where 0x0E means '*1', 0x0F means '*2', 0x0D means '*0.5', etc.) and WX YZ the fraction part.
roman
___________________________________________________________ Der frühe Vogel fängt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: http://mail.yahoo.de
-
Roman Haefeli