Linux-audio-dev

linux-audio-dev@lists.linuxaudio.org

4 participants
7441 discussions

[linux-audio-dev] Re: [Linuxsampler-devel] buffers, periods, & fragments

by Garett Shulman

Flo, Thanks for the jack suggestion. I definetly do need to spend some time working with jack on my system. Also, thanks for the great low latency documentation on tapas - ugh!. After getting 2.6 setup with | Preemptible Kernel (Low-Latency Desktop), fiddeling with my hardware to get the soundcard irq's right, and some tweeking with chrt all I can say is "wow. WOW!". I don't think there is anything you can do on a 2.4 kernel to get close to this level of performance and control. Sweet! I almost giggle as the entire system temporarily hangs while linuxsampler loads gig files at outrageous speed. Also, I wanted to let you know that after a bit more hacking I was able to get dshare working with the ice1712. It's awesome. I can send linuxsampler out of channels 1-4 and have fluidsynth playing out of channels 5 & 6, with 2 channels to spare for ecasound or something... all simultaneously, at very low latency, & hardly any processor load. I paid for the hardware... might as well take advantage of it. Besides, this way I can give my processor a break, under-clock it, & keep fan volume very low. Here is a snippet from asound.conf. -Garett |pcm_slave.66_slave { pcm "hw:1,0" channels 8 rate 44100 buffer_size 256 period_size 128 } pcm.66ch1234_dshare { type dshare ipc_key 18273645 slave 66_slave bindings.0 0 bindings.1 1 bindings.2 2 gindings.3 3 } pcm.66ch1234 { type plug slave.pcm "66ch1234_dshare" } pcm.66ch56_dshare { type dshare ipc_key 18273645 slave 66_slave bindings.0 4 bindings.1 5 } pcm.66ch56 type plug slave.pcm "66ch56_dshare" } Florian Schmidt wrote: > On Thu, 15 Sep 2005 22:40:32 -0600 > Garett Shulman <shulmang(a)colorado.edu> wrote: > >> Hello, I have been fooling around with my alsa asound.conf in an attempt >> to take advantage of the hardware mixer in my ice1712 and have multiple >> apps output to it. > > > The hw mixer on the ice1712 is not doing what one normally refers to as > hardware mixing with consumer grade cards (i.e. allowing several apps > concurrent access at the same time). It mixes and routes channels from > its single 10 in/12 out channel device. > > The way to get concurrent access with an ice1712 based card is _software > mixing_. That is, ALSA needs to do it (or whatever sounddriver you use). > > I would urge you to take a very long look at JACK though. LinuxSampler > has excellent jack support and most apps made for peolpe creating music > usually have jack support, too. You'll save yourself lots of hassles > (setting up jackd isn't so tough when you read some docs). It is very > simple to route specific LS patches to specific output channels on your > ice based card when using jack. > >> In order to acomplish this I need linuxsampler to be >> able to access devices I setup in asound.conf instead of just hw:x,x >> devices. This was easy enough to acomplish by nuking all of the '"hw:" >> +' code from AudioOutputDeviceAlsa.cpp. The next issue relates to the >> fact that the alsa dshare plugin requires that all of the "virtual >> devices" that I create from the ice1712 card share the same buffer_size, >> period_size, periods, & period_time settings. So, linuxsampler crashes >> when it tries to set buffersize and periods. I figured I would just find >> out what linuxsampler was trying to use for those values, setup the same >> values in the asound.conf and then comment the code that sets them in >> linuxsampler. This seems to work except that when linuxsampler connects >> to the device I notice click and pops from the device & the output from >> linuxsampler is rather distorted. I suspect that I'm just not setting >> the values quite right in asound.conf. However, I guess I also may be >> just mangleing linuxsampler beyond proper function. :) I was wondering >> if anywone has any suggestions. When linuxsampler tries to set the >> buffersize it is using FragmentSize=128 and Fragments=2. So I am >> interpreting this as period_size=128, periods=2, & buffer_size=256. The >> slave device in asound.conf looks like this: >> pcm_slave.66_slave { >> pcm "hw:0,0' >> channels 8 >> rate 48000 >> buffer_size 256 >> period_size 128 >> periods 2 #I also tried 1 here... counting from 0... >> period_time 0 >> } >> Then I create some dshare devices from this and some plug devices for >> each dshare device and connect linuxsampler to one of the plug devices. >> Any suggestions would be greatly appreciated. -Garett > > > dshare still won't allow you to have multiple apps use your ice1712 > card. You'll need dmix for that. Have a look at alsa.opensrc.org if you > want to go this route. dmix will probably kill latency though. > > But better take a look at jackit.sf.net. > > Regards, > Flo >

19 years

[linux-audio-dev] [OT] OGG on Tiger

by Hans Fugal

On Sat, 17 Sep 2005 at 11:20 +0200, Adrian Prantl wrote: > i'm afraid this is slightly offtopic, but does anyone know of an ogg/ > vorbis plugin for the new iTunes running on 10.4? XMMS through darwin ports works, and that is what I currently use. I've heard good things about VLC, too. While we're on the subject, I've done a little bit of research on the problem with quicktime. The qtcomponents project that worked before qt7 did things as a component, when they arguably should have made a codec from the start. Apple hadn't (and hasn't) solidified the API, and so things stopped working in Tiger and also with QT7 on Panther. I'm not sure, but if it had been done as a codec to begin with it might still be working. In any case it looks like doing it as a codec is the way to go at this stage. I think this will require basically taking Apple's AudioCodec example and wiring it up to libvorbisfile. I and at least one other person intend to do this "when I get time." If anyone out there is good with Apple codecs or good with libvorbisfile, help would be appreciated, and speed up the process. See this discussion for more information: http://tinyurl.com/8z6rb Also this bug reporter got some good info from Apple: http://tinyurl.com/cy35h -- Hans Fugal | If more of us valued food and cheer and http://hans.fugal.net/ | song above hoarded gold, it would be a http://gdmxml.fugal.net/ | merrier world. | -- J.R.R. Tolkien --------------------------------------------------------------------- GnuPG Fingerprint: 6940 87C5 6610 567F 1E95 CB5E FC98 E8CD E0AA D460

19 years

Re: [linux-audio-dev] convenient ogg vorbis wrapper

by Conrad Parker

On Sat, Sep 17, 2005 at 08:49:27AM +1000, Erik de Castro Lopo wrote: > Richard Spindler wrote: > > > Hi, > > > > I'm looking for a convenient wrapper library for ogg vorbis, because > > what I've found on the xiph.org pages looks a little overengineered to > > me. > > > > I've used libsndfile most of the time, so this is the API Style that > > I'd prefer, basically I need 4 functions I believe: > > Conrad Parker has been working on adding Ogg Vorbis and Speex support > to libsndfile. Its been very close to working for some time now :-). aye, if you'd like to play with it, it's in an arch repository at: http://www.metadecks.org/arch/ The library dependencies etc. were described in this email (which references Erik's libsndfile--hack branch): http://music.columbia.edu/pipermail/linux-audio-dev/2005-July/013288.html cheers, Conrad.

19 years

[linux-audio-dev] convenient ogg vorbis wrapper

by Richard Spindler

Hi, I'm looking for a convenient wrapper library for ogg vorbis, because what I've found on the xiph.org pages looks a little overengineered to me. I've used libsndfile most of the time, so this is the API Style that I'd prefer, basically I need 4 functions I believe: what_samplerate_are_you( vorbis_file ); how_long_are_you( vorbis_file ); seek_to_position( vorbis_file, 23456 ); read_data( vorbis_file, buffer ); libvorbis and libogg however have to interact in some obscure way and I honestly don't want to waste my time reinventing the wheel, while I could be hacking on a new asskicking application ;) -Richard

19 years

[linux-audio-dev] research on lau / lad -statistics

by Michael Iber

Hello administrators, hello list members, for my research for a lecture I will be holding on opensource and audio in October in Stuttgart the following informations would be very interesting: How many people are subscribed to the lau / lad -lists. How many participate actively ? Is it possible to roughly estimate, where they are from (Amerika, US, Asia, part of Europe, Germany). Is anybody capable of providing those informations? Thank you very much, Michael

19 years

[linux-audio-dev] JACK error 4294967295

by Hans Fugal

2.6.13.1-rt6, rlimits-patched pam and configured thus: # in /etc/security/limits.conf * - rt_priority 0 * - nice 0 @audio - rt_priority 50 @audio - nice -10 jack starts happily with realtime enabled, and the xruns are very few. But no jack application will start, they all give this error: jack_create_thread: error -1 switching current thread to rt for inheritance: Unknown error 4294967295 It's like the ol' jack is root apps can't connect days, but I thought rtlimits was supposed to make it possible for a user to get rt_priority. Am I missing something? Ah! It came to me just now - I set the priority field in qjackctl to 0 instead of 1 (where it was) and now jack apps can start. Heads up there. Maybe applications need a way to recognize what rt_priority level to ask for based on what jack is running at? On a tangent, how exactly does that work? Is rt_priority=0 sufficiently prioritized? (because it is the only thing running realtime) -- Hans Fugal | If more of us valued food and cheer and http://hans.fugal.net/ | song above hoarded gold, it would be a http://gdmxml.fugal.net/ | merrier world. | -- J.R.R. Tolkien --------------------------------------------------------------------- GnuPG Fingerprint: 6940 87C5 6610 567F 1E95 CB5E FC98 E8CD E0AA D460

19 years

[linux-audio-dev] Re: Impulse response visualize tools

by Juhana Sadeharju

>From: Anders Torger <torger(a)ludd.luth.se> > >Does anyone know of a tool especially useful for visualizing room >impulse response measurements? Magnitude responses, waterfall plots and >such. Did you find anything? I just started browsing http://www.opendx.org which should be open source. It is a modular visualization system. Don't know if it helps but the gallery looks advanced. I have a problem in checking out the CVS, though. Juhana -- http://music.columbia.edu/mailman/listinfo/linux-graphics-dev for developers of open source graphics software

19 years

[linux-audio-dev] [OT] problem with parallel gcc

by Julien Claassen

Hi! I wonder if someone could provide me with the right hint, for I feel it is not much to get me where I want. I have an old suse system (gcc3.2 glibc 2.2.5), now I wanted to install a newer gcc (3.4.3). It compiled fine and I can compile other programs with it. Yet there is - apparently - some configuration problem. For if I start linuxsampler or ecasound compiled with new gcc it says, it has a problem with libgcc_s.so.1. It can't find some symbol VERSION_3_3 or something to that effect. I tried ldd on those programs and it told me that they contain both libgcc_s.so.1 from /lib and /usr/local/gcc343/lib. What am I doing wrong. For I know this can work fine. I recompiled my libstdc++ (from the 3.4.3 package) with my already installed compiler to prevent it from including the old libgcc_s, but no effect. What can I do? Should I put the new library-paths at a specific point in my /etc/ld.so.conf? Set some special environment variable? Any help is GREATLY APPRECIATED! Kindest regards Julien -------- Music was my first love and it will be my last (John Miles) ======== FIND MY WEB-PROJECT AT: ======== http://ltsb.sourceforge.net - the Linux TextBased Studio guide

19 years

[linux-audio-dev] swh-plugin errors with gcc4 - RedHat FC4

by Michael Stein

I am having issues getting my FC4 - x86-64 - GCC4.0 box to compile swh-plugins cleanly. I have applyed the patches that I have seen on this list but I still get errors ranging from not determining the proper -march value to simply failing with the following error: /bin/sh ./libtool --tag=CC --mode=link gcc -march=k8 -O3 -module -avoid-version -Wc,-nostartfiles -o hermes_filter_1200.la -rpath /usr/local/lib/ladspa hermes_filter_1200.lo util/libblo.a -lrt -lm -lm -lm *** Warning: Linking the shared library hermes_filter_1200.la against the *** static library util/libblo.a is not portable! gcc -shared .libs/hermes_filter_1200.o util/libblo.a -lrt -lm -march=k8 -nostartfiles -Wl,-soname -Wl,hermes_filter_1200.so -o .libs/hermes_filter_1200.so /usr/bin/ld: util/libblo.a(libblo_a-blo.o): relocation R_X86_64_32 against `a local symbol' can not be used when making a shared object; recompile with -fPIC util/libblo.a: could not read symbols: Bad value collect2: ld returned 1 exit status make[2]: *** [hermes_filter_1200.la] Error 1 make[2]: Leaving directory `/usr/local/src/audio_apps/swh-plugins-0.4.13' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/usr/local/src/audio_apps/swh-plugins-0.4.13' make: *** [all] Error 2 Any help would be appreciated.

19 years

[linux-audio-dev] Impulse response visualize tools

by Anders Torger

Does anyone know of a tool especially useful for visualizing room impulse response measurements? Magnitude responses, waterfall plots and such. I'd prefer something more efficient and quicker to work with than octave and gnuplot. /Anders Torger

19 years

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Linux-audio-dev