Linux-audio-user May 2010

linux-audio-user@lists.linuxaudio.org

153 participants
128 discussions

Steinway_IMIS2.2 soundfont (no loops fix)

by Roberto Gordo Saez

This is Steinway_IMIS soundfont, version 2.2. ftp://musix.ourproject.org/pub/musix/sf2/Steinway_IMIS2.2 This version fixes the issue with loops. I hope this is the good one and there are no more remaining major bugs. Marcos is a little busy right now, so he asked me to make this fix. He is thinking to make other improvements, so expect more updates soon.

1 year, 8 months

[linux-audio-user] Sub-harmonic generator? (LADSPA or VST)

by Tim Howard

Hello, Does anyone know of a good plugin that will generate subharmonics? I would like to put a little more low frequency "oomph" into my bass track. Preferrable LADSPA, but VST would work, too. Thanks for any help! -TimH

8 years, 3 months

Re: [LAU] Simple, easy multithreaded circular buffer library for Linux?

by Kjetil S. Matheussen

Olivier Guilyardi: > > Ken Restivo wrote: >> It has been over 7 years since I last messed around with writing Pthreads applications. >> >> I recall it as a painful, ugly, brain-numbing task. I located an exercise I did back then to address the consumer/producer problem in Pthreads, and just the sight of it is giving me a headache. >> >> I'm being lazy, so instead of researching everything that's out there, I'll ask here: can anyone recommend a relatively simple and painless abstraction library (GPL or LGPL of course) that will give me functions to create a thread in which I can stuff things into a ring buffer, and another thread in which I can pull stuff out of it? >> >> By the way, I know that JACK has a very nice event buffer which is insanely easy to use (and I have), and makes multithreading almost transparent, but this isn't a JACK app. > > I don't know of any abstraction library, but creating/terminating a normal > thread with pthread is really an easy task IMO. It's about 10 lines in C. > > For inter-thread communication there's Portaudio's ring buffer: > http://portaudio.com/trac/browser/portaudio/trunk/src/common/pa_ringbuffer.h > > It can easily be used out of Portaudio (I'm currently doing that), and it > features memory barriers [1] which AFAIK Jack's ringbuffer doesn't. > > One problem with everything Portaudio is this heavy naming scheme. For a simpler > API, you might like my little wrapper: > http://jackbeat.samalyse.org/browser/jackbeat/trunk/src/core/ringbuffer.h > Nice. It's probably quicker to copy the jack_ringbuffer.c file out of jack though. > Portaudio actually also offers a callback mechanism (with hidden thread > creation), so if you're coding an non-JACK audio app, you might want to check it > out. > > For thread synchronization, semaphores (man semaphore.h) are really easy to use. > However, if you need a lock-free equivalent (for realtime, ...) phtread mutex > and especially pthread_mutex_trylock are your friends. > Those friends can be really cranky sometimes though. By using atomic operations instead, it's possible to avoid a lot of headache by not having to synchronize at all. Performance might be better too. Midishare has lockfree atomic functions for lifo and fifi queues: http://midishare.cvs.sourceforge.net/viewvc/midishare/midishare/src/common/…

12 years, 10 months

jackd in realtime as user: a no-go in spite of modifying limits.conf

by Der Mickster

Hey, Can't seem to start jackd in realtime as user on ubuntu studio. It does start as root, but that's hardly a consolation (unless I run all audio software as root, which would be kinda daft IMO) @audio - rtprio 99 @audio - nice -10 @audio - memlock unlimited have been added to /etc/security/limits.conf Nevertheless, starting it as user just doesn't seem to work, and I just can't figure out why. I've googled the issue, and everybody seems to agree that modifying limits.conf in the above manner should solve the problem - yet this is what I get: mick@kaizoku:~$ jackd -R -P89 -dalsa -dhw:0 -r48000 -p256 -n3 -S & [1] 3092 mick@kaizoku:~$ no message buffer overruns jackd 0.116.1 Copyright 2001-2005 Paul Davis and others. jackd comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions; see the file COPYING for details JACK compiled with System V SHM support. cannot use real-time scheduling (FIFO at priority 10) [for thread -1217255744, from thread -1217255744] (1: Operation not permitted) cannot create engine Any ideas? Mick.

13 years, 11 months

It's time to release The Salamander Grand Piano

by alexander

Well, here it is, it took a little longer than expected, mostly do to school work. As I reported earlier I ditched the giga format and went for sfz, you will need linuxsampler cvs to load the sfz file. I do not know of any other sampler for linux that can load sfz. The library also contains pedal action noise samples, these are also in the sfz file but linux sampler don't have the opcodes(on_hi/loCC#) to play them yet so I have not optimized the volume on these yet. They are located on the bottom of the sfz file. Enjoy! and feel free to mirror it where ever you like! http://freepats.zenvoid.org/Piano/SalamanderGrandPiano.tar.bz2

14 years, 4 months

Connie, an organ template for JACK

by Martin Homuth-Rosemann

Hi, sorry, just realized that the hammond discussion mentioned below was here and not at LAD, so please allow this kind of "crosspost": this is my first post to LAD. The discussion about a hammond simulation "Fons could you make us...", Beatrix and some research for writing a (german) wikipedia article (stub) about the Vox Continental inspired me to hack a quick organ program that simulates the internal signal flow of the "Connie" with JACK MIDI input and JACK audio output. Have fun: http://cryptomys.de/horo/Connie/Connie-0.1.tar.gz Ciao Martin Homuth-Rosemann

14 years, 4 months

KDE 4.4's nice integration of Pulseaudio, ALSA, and JACK!

by Niels Mayer

http://pulseaudio.org/wiki/KDE see http://pulseaudio.org/raw-attachment/wiki/KDE/phonon-pulse-broken.pngwhich allows you to select and prioritize PA/Jack/Alsa for different audio usage contexts.... very nice way of doing things. Other desktops might take note. Context: Simon Lewis' comment about KDE working well w/o pulseaudio prompted me to switch to the KDE desktop. I think he'd already suggested I try krusader (wish they hadn't used that name && they better not start using names with multiple k's in em!), which to get fully working, I ended up installing half of KDE (i'd already gotten the first half with setting up kmplayer). His last comment finished the job and now I appear to be using KDE 4.4.2 (from f12) outright. No more issues with https://bugs.launchpad.net/ubuntu/+source/gnome-settings-daemon/+bug/404340 and https://bugs.launchpad.net/ubuntu/+source/gnome-media/+bug/472961 (which also affect fedora and gnome in general).... excerpt from #fedora-kde: (12:14:34 PM) npm: https://bugzilla.redhat.com/show_bug.cgi?id=445543#c20 (12:15:00 PM) npm: thanks... i'll try the startupnotify thing. (12:24:16 PM) rdieter: fyi folks, coling asked if we could review this for accuracy, and clarify which fedora versions apply, http://pulseaudio.org/wiki/KDE (12:27:48 PM) npm: rdieter -- i'll fwd link to http://ccrma-mail.stanford.edu/pipermail/planetccrma/ (12:28:07 PM) npm: and thanks for reply to my comment on the bug... (12:28:10 PM) rdieter: what's planetccrma have to do with it? (12:28:29 PM) npm: we have issues w/ pulseaudio and a running argument about it all the time (on LAU too) (12:28:58 PM) npm: planetccrma distributes rpms for audio/video tools and also a rt kernel (12:29:23 PM) rdieter: ah, cool. (12:29:33 PM) npm: a comment on there is actually what sent me to try KDE (12:29:43 PM) rdieter: missed the gear switch away from talking about startup notification. :) (12:30:54 PM) npm: i posted http://ccrma-mail.stanford.edu/pipermail/planetccrma/2010-May/016881.html (12:31:17 PM) npm: and http://ccrma-mail.stanford.edu/pipermail/planetccrma/2010-May/016882.htmlme… KDE (12:31:54 PM) npm: "On fc12 x86_64 with KDE 4.4.x and Planet CCRMA core packages installed - a simple remove all packages except pulseaudio-libs and pulseaudio-libs-glib2 was sufficient. On restarting KDE, KDE automatically picked-up that the pulseaudio sound server is unavailable and asks whether the pulseaudio configuration should be permanently removed - the answer is a clear yes!" Niels http://nielsmayer.com

14 years, 5 months

Re: [LAU] Apple hardware

by rosea.grammostola

On Fri, 2010-05-07 at 12:43 +0200, Florian Faber wrote: > Hi, > > > I asked also an Thinkpad user, he said that it wasn't silent and you > > better use an Desktop PC. He didn't compare it with an Macbook pro. > > There are thinkpads and there are thinkpads. > > > > So the question remains if this also is true for the Macbook pro. > > You have really bad service (unlike 3 years of 24h service with > thinkpads), you have bad linux support (unlike with thinkpads), you get > butt ugly devices (unlike with thinkpads), you have really bad keyboards > (unlike with thinkpads). > > With thinkpads you got for years now devices that run 4-8h, you have 3g > for years now, you can get really good touch screens (great for adding > notes to scores), and they can fall on a tiled floor without disintegrating. > > There really is no question. Besides Apple being evil, their hardware > really cannot keep up and is very poorly engineered. > > (I have several MacPro, MacBook Pro and MacBooks in the studio. There > really is no choice if you HAVE to use Final Cut Pro.) > > > Flo Then the question concerning the noise is still there... Maybe an independent organization should compare laptops and Macbooks for music production... BTW. Are IBM computers also popular for Desktops in the Linux world? \r

14 years, 5 months

Klick oddity

by Julien Claassen

Hi everyone! I recently downloaded the latest Klick version and notcied someting odd. When I start klick like this: klick -T 4/4 120 I can use jack_transport to start/stop/relocate and my MIDI sequencer can be synced to jack_transport control. If on the other hand I start klick in interactive mode: klick -T -i 4/4 120 I can't use jack_transport control to do anything. Klick can start/stop and this is certainly send. But jack_midi_clock doesn't get a clock-signal, somy sequencer can't be synced. Any idea, why that might be? Kindly yours 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 ======= AND MY PERSONAL PAGES AT: ======= http://www.juliencoder.de

14 years, 5 months

Re: [LAU] [LAD] like "qjackctl", but trimmed of all fat

by Aaron Krister Johnson

On 5/29/10, akjmicro(a)gmail.com <akjmicro(a)gmail.com> wrote: > Hey all, > > Yes, grepping for the port type which appears underneath with a 'jack_lsp > -t' will be more consistent and dependable. Or, using a python-jack lib > function and not depending on any system shell calls. The problem then > becomes, is the jack lib for python well documented? If so, I think that's > the real future of jackctl.py > > PS Qjackctl may be lightweight, but installing the entire QT toolkit just to > use it is not! > > AKJ > > Sent from my Verizon Wireless BlackBerry > > -----Original Message----- > From: Robin Gareus <robin(a)gareus.org> > Date: Sat, 29 May 2010 14:00:52 > To: Julien Claassen<julien(a)c-lab.de> > Cc: Aaron Krister Johnson<aaron(a)akjmusic.com>; > <linux-audio-user(a)lists.linuxaudio.org>; > <linux-audio-dev(a)lists.linuxaudio.org> > Subject: Re: [LAD] [LAU] like "qjackctl", but trimmed of all fat > > Hi Julien, Hey Aaron, > > read 'jack_lsp --help'. > > '-t' does not take any arguments; it just makes jack_lsp print the type. > the filter-string only acts on the port-name (BTW, not only the > beginning of the port-name; but it's case-sensitive: strstr() ) > > Anyway I can reproduce the problem, some jack-midi ports show up in the > audio-tab of jackctl20100528b.py. > > jackctl20100528b checks for lowercase 'midi' in the port-name instead of > looking up the port-type. So a2jmidi for example with an upper-case M > "Midi.." ends up in the audio-panel. > > Your suggestion to parse the output of 'jack_lsp -t -c' is spot on. > the (currently 2) possible return values are (indented by tab): > > #define JACK_DEFAULT_AUDIO_TYPE "32 bit float mono audio" > #define JACK_DEFAULT_MIDI_TYPE "8 bit raw midi" > > ..or as you suggest using the python-module for JACK may also simplify > things and make jackctl easier to maintain. > > Cheers! > robin > > PS. Oh, and which of qjackctl's features makes it 'fat'? it's not > bloated in any way. I'd rather put it the other way 'round and say that > jackctl is 'slim'. Sorry could not resist. > > > On 05/29/2010 12:23 PM, Julien Claassen wrote: >> Hello Aaron and Jack-Team! >> There seems to be a bug in my jack_lsp. I just started a2jmidid and >> j2amidi_bridge. when I do a jack_lsp I get all the ports. >> When I do: jack_lsp -t midi I only get one port from jack_midi_clock, >> but none of the other ones. >> When I type: jack_lsp -t, I can't see a difference between the >> jack_midi_clock port and the others: >> jack_lsp -t >> [...] >> a2j:Virtual Raw MIDI 0-0 [16] (capture): VirMIDI 0-0 >> 8 bit raw midi >> a2j:Virtual Raw MIDI 0-0 [16] (playback): VirMIDI 0-0 >> 8 bit raw midi >> a2j:Virtual Raw MIDI 0-1 [17] (capture): VirMIDI 0-1 >> 8 bit raw midi >> a2j:Virtual Raw MIDI 0-1 [17] (playback): VirMIDI 0-1 >> 8 bit raw midi >> a2j:Virtual Raw MIDI 0-2 [18] (capture): VirMIDI 0-2 >> 8 bit raw midi >> a2j:Virtual Raw MIDI 0-2 [18] (playback): VirMIDI 0-2 >> 8 bit raw midi >> a2j:Virtual Raw MIDI 0-3 [19] (capture): VirMIDI 0-3 >> 8 bit raw midi >> a2j:Virtual Raw MIDI 0-3 [19] (playback): VirMIDI 0-3 >> 8 bit raw midi >> a2j:M Audio Delta 1010LT [20] (capture): M Audio Delta 1010LT MIDI >> 8 bit raw midi >> a2j:M Audio Delta 1010LT [20] (playback): M Audio Delta 1010LT MIDI >> 8 bit raw midi >> j2a_bridge:playback >> 8 bit raw midi >> a2j:j2a_bridge [129] (capture): capture >> 8 bit raw midi >> Jack MIDI Clock:midi_out >> 8 bit raw midi >> >> Or is the argument "midi" only seen as the start of a port_name? >> If so, Aaron, you must rewrite this part of jackctl (I guess you do >> what I described, because I get exactly your output). You should rewrite >> it using: >> jack_lsp -t >> And then parse the type info underneath each name. I think a simple >> grabbing for "audio" or "midi" will do. But I guess, that in the long >> run, using the python module for jack, will be more efficient and easy >> to use. >> Kindly yours >> Julien >> > -- Aaron Krister Johnson http://www.akjmusic.com http://www.untwelve.org

14 years, 5 months

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-user May 2010