Linux-audio-dev

linux-audio-dev@lists.linuxaudio.org

4 participants
7509 discussions

Inter thread Communication: Design Approach

by Harry van Haaren

Hey All, I'm looking to improve a program's design with regards to how it communicates between the "jack" thread and its "main/gui" thread. Please note I'm *not* looking for implementation details like what ringbuffer to use, this has been discussed here before. Conditions: Gui needs to feed data trough to the jack thread (data = parameter moves etc) Jack thread needs to push data (buffers for waveforms & "playhead" info) The real question: What is a neat solution to passing various data trough a ringbuffer? My (hacky?) solution: Create a class, call it "Event". Runtime now looks like so: 1. Create a EventType enum, set the details 2. Write those "Events" into the ringbuffer 3. Switch based on EventType, and handle the event. While not terribly ugly, that Event class starts to get bigger & nastier, so I concidered sub-classing it... but I'm not sure this is going in the right direction. I'm very intrested how the "big" programs around have approached this problem... Cheers, -Harry

13 years, 10 months

IR: Lv2 reverb (without the GUI..?)

by Harry van Haaren

Hey All, I'm faced with a problem that I can't see an easy way around regarding the use of the IR reverb plugin. I'm running the GUI in a seperate *process*, and doing all lv2 communication over OSC. For the most part this is easy, the problem rears its head when one want's to use a plugin whose UI requires "instance-access". Basically, the IR plugin GUI needs access to the Lv2_Handle. But I can't provide that due to the OSC communication. So I've concidered "spoofing" a plugin on the UI side, and keeping it up to date with what the "real" one is doing in the Engine. Bit ugly, and if the UI has instance access, will it still call the normal "port" events..? Because otherwise I'm lost with trying to get at the UI data. The other problem is that loading a "sample" into the IR convolution happens in a pretty strange way: there's 3 Control Input ports, and together they make up a 64bit file hash. I understand the reasons behind this decision, and I'm not trying to criticise the implementation, I'm just not sure how I can send a certain file to these ports to make it work... -Harry

13 years, 10 months

Re: [LAD] Inter thread Communication: Design Approach

by Florian Paul Schmidt

On 09/02/2011 12:54 PM, Pedro Alves wrote: > boost::function<void(void)> serves this purpose for me in jass. To >> create and pass a functor that assigns a new auditor generator to the >> one in the engine, and then tells it to play i do for example: >> >> write_blocking_command(assign(engine_.auditor_gen, p)); >> write_blocking_command(boost::bind(&engine::play_auditor, >> boost::ref(engine_))); assign() is just a utility template to make >> creating functors that do assignments easier.. boost::bind is used to >> make all passed functors 0-ary (e.g for binding member functions to >> their instance or binding arguments to the functor.. and >> write_blocking_command is just a utility function that disables the GUI >> until the acknowledgement from the engine has come back to the GUI,, The >> command ringbuffer is just a ringbuffer holding >> boost::fucntion<void(void)> objects.. >> >> typedef ringbuffer<boost::function<void(void)> > command_ringbuffer; >> >> Examples from here: >> https://github.com/fps/jass/blob/master/main_window.h >> https://github.com/fps/jass/blob/master/assign.h >> https://github.com/fps/jass/blob/master/engine.h Regards, Flo > There are alternatives (even predecessors) to boost.function that > are much faster and avoid the heap/new. See > <http://www.codeproject.com/KB/cpp/fastdelegate2.aspx> for example. > (I used Don's original "Fastest Possible C++ Delegates" on an embedded > project years ago -- worked great). > Oh, thanks for the interesting read :D Flo

13 years, 10 months

a *simple* ring buffer, comments pls?

by James Morris

just wondered if any more-experienced-than-i devs might comment on this. written in c for c (obviously). i realize it's not portable outside of GNU GCC (regarding the GCC atomic builtin funcs __sync_***). meant for a single reader thread and a single writer thread. comments regarding thread safety very much welcome. thanks in advance. james. #include "rng_buf.h" /* only prototypes the public functions and typedefs the struct */ #include <stdlib.h> #include <string.h> #include "debug.h" struct _RingBuffer { size_t count; void** buf; void** bufend; void** volatile w; void** volatile r; }; RngBuf* rng_buf_new(size_t count) { size_t sz = 1; RngBuf* mb = malloc(sizeof(RngBuf)); if (!mb) { return 0; } for (sz = 1; sz < count; sz <<= 1) ; mb->buf = calloc(sz, sizeof(void*)); if (!mb->buf) { free(mb); return 0; } mb->count = sz; mb->bufend = mb->buf + mb->count - 1; mb->w = mb->buf; mb->r = mb->buf; return mb; } void rng_buf_free(RngBuf* mb) { free(mb->buf); free(mb); } size_t rng_buf_write(RngBuf* mb, const void* data) { if (__sync_bool_compare_and_swap(mb->w, 0, data)) { if (!(__sync_bool_compare_and_swap(&mb->w, mb->bufend, mb->buf))) __sync_add_and_fetch(&mb->w, sizeof(void*)); return (size_t)1; } return (size_t)0; } void* rng_buf_read(RngBuf* mb) { void* data; if ((data = __sync_fetch_and_and(mb->r, 0))) { if (!__sync_bool_compare_and_swap(&mb->r, mb->bufend, mb->buf)) __sync_add_and_fetch(&mb->r, sizeof(void*)); return data; } return NULL; } void rng_buf_reset(RngBuf* mb) { /* needs work */ mb->r = mb->w = mb->buf; }

13 years, 10 months

Jack_session

by thijs van severen

Hi all My name is Thijs and i have been helping out with Hydrogen for some time now (mainly manual, website and a tiny bit of coding) Anyway, since jack_session support was recently added to Hydrogen (in V0.9.6 that is currently still in development but available from svn via hydrogen-music.org) i decided to test jack_session a bit and i must say i'm really exited about this :-) IMHO this is the most critical piece of the puzzle and it makes linux audio _so_ much more usable! so lots of KUDOS to everyone that has been working on it !! I have done several 'intro sessions' for Mac audio users and they are generally really impressed with what they see, but the lack of a good session management mechanism is a real showstopper for them... (i know there are other options, but none of those have ever allowed me to truly save my session *at any given moment*) however, the first thing that slowed me down was simply knowing what apps already support jack_session i know that jack_session is fairly new, so it's normal that some info is missing, but it would be great if the websites out there that list the available linux audio apps (like http://apps.linuxaudio.org/apps/categories/jack) would also indicate if the app supports jack_session the jack_session page on the jackaudio.org wiki<http://trac.jackaudio.org/wiki/WalkThrough/User/jack_session>also lists a number of apps, but i don't think it's up to date, is it ? any comments? grtz Thijs PS: i'd be happy to help out with documenting jack_session if needed PS: it would be nice to have a 'jack_session enabled app' banner to put on the site of the apps that support it. graphical designers ? -- follow me on my Audio & Linux blog <http://audio-and-linux.blogspot.com/> !

13 years, 10 months

interesting article on why most "music apps" fail

by Paul Davis

the discussion is about consumer-oriented apps, but its an interesting one anyway ... http://www.billboard.biz/bbbiz/industry/digital-and-mobile/why-most-music-a…

13 years, 10 months

[LAA][ANNOUNCE] New LV2 Soft Synth: The Newtonator v0.5.0

by Michael Bechard

Brand new LV2 plugin soft synth here. The Newtonator (v0.5.0) features: * LV2-compliant synth plugin * Simple, easy-to-use GTK-based GUI * 3-note polyphony (err, um, until I can do some serious optimization) * "Unique" synthesis algorithm (Tuna Pagan Fellowship) The Newtonator specializes in making crazy, harsh sounds, so if you're looking for some sounds to produce the next Yanni album, keep looking. Please refer to the user manual (http://newtonator.sourceforge.net/manual/index.html) for more gory details. Download: http://sourceforge.net/projects/newtonator/files/newtonator-0.5.0.tar.gz/do… NOTE: This plugin was written against the excellent libraries included in the lv2-c++-tools package written by Lars Luthman. However, as of v1.0.4, the GUI library only supports rev. 2 of the LV2 specification, so it may have problems on newer hosts (though I have done some (very) preliminary testing against Ardour3 and it seems to work ok). Alpha testing mostly done against zynjacku and Elven (comes with ll-plugins package) hosts. This puppy's in beta, so feel free to get in touch with me at the project forum (https://sourceforge.net/projects/newtonator/forums/forum/1820765) or submit a bug if needed. Thanks, and have fun. Michael Bechard https://sourceforge.net/projects/newtonator/

13 years, 10 months

RT-Safe UI/Engine Decoupling using Functional Programming and Reference Counted SmartPointers

by Florian Paul Schmidt

Hi, during the process of writing a new small jack sampler which fits my workflow I came up with this little scheme to solve the UI/engine decoupling problem. For the purpose of spreading the idea or alternatively getting answers about how it's broken and sucks I decided to write a little article describing it.. http://178.63.2.231/~tapas/wordpress/?page_id=45 The (largely unfinished and unusable) sampler project is here: https://github.com/fps/jass Let me have it.. Regards, Flo

13 years, 10 months

[HELP WANTED] ALSA Driver developer

by Gabriel Beddingfield

All: A recruiter approached me looking for an ALSA driver developer to enhance an existing driver. This is a short-term, off-site gig. Some details below. Please contact Mr. DeAmelio directly. Thanks, Gabriel ---------- Forwarded message ---------- From: David DeAmelio <david_deamelio(a)oxfordcorp.com> Date: Tue, Aug 16, 2011 at 2:29 PM Subject: Job Description - Oxford International To: gabrbedd(a)gmail.com Hey Gabriel, Here’s the description….Thanks. Location: OFFSITE Start Date: ASAP Duration: 1-2 Weeks Manager needs consultant to come onsite for 1st few days. Client will pay for flight there and back. Must: Linux ALSA Device Drivers—Alsa- advanced linux sound architecture Blackfin Plus: D2D: Consultant will be enhancing an audio driver to support a voice quality recorder __________________________________ David De Amelio Team Lead Software/Hardware Consulting Services Oxford International a division of Oxford Global Resources 100 Cummings Center Suite 206N Beverly, MA 01915 877.258.9982 x 7608 Office 978.922.7547 FAX david_deamelio(a)oxfordcorp.com www.oxfordcorp.com Oxford Europe The Right Talent. Right Now.® This e-mail, and any attachments thereto, is intended only for the addressee(s) named herein and may contain privileged and/or confidential information. If you are not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this e-mail, and any attachments thereto, is strictly prohibited. If you have received this e-mail in error, immediately notify the sender by return e-mail and permanently delete the original, including any attachments thereto, and any copy and printout thereof.

13 years, 10 months

Yoshimi Midi Learn: 1st testing release

by louis cherel

Hello everyone, Every one knows Yoshimi, the fork of ZynAddSubFx. One thing was lacking to yoshimi to be perfect: to be nearly fully controlled by midi controls ( no OSC, sorry ). ZynAddSubFx had possibilities to control a few parameters with complicated NRPN, Yoshimi recently had ( in the test versions ) some features too. But now I'm proud to announce you the work of licnep ( not me, I'm just a bug reporter ) who made the "midiLearn" function for yoshimi. It's not stable for now because it's recent, and not full, but here are the present features: * Control System effects, Part Insert Effects * Master/Part Volume, Pan, System Effect Sends * Most of ADsynth parameters * Add/Remove controller * detect the channel and the number * reset the knob ( its position ) I think it's a feature that's very useful and could help many yoshimi/zyn users. To use it, that's simple: connect your controller to yoshimi, right-click on a blue knob ( yellow are ones which are not supported for now ) and click "midi Learn" move your controller, it detects automatically the controller. To see and modify controllers, go to the Yoshimi> MIDI controllers menu. To erase midi control of a knob, simply right click on it and click on "remove midi control" Here is the gitHub repository: https://github.com/licnep/yoshimi To download and install it, follow the explications link ( gitHub ): https://github.com/licnep/yoshimi/wiki/How-to A light page to understand how to control others not implemented controllers: https://github.com/licnep/yoshimi/wiki/Source-code Pages to follow the news of the project: Facebook: https://www.facebook.com/pages/Yoshimi-midi-learn/224823617534934 Twitter: http://twitter.com/#!/YoshimiMIDI So if you're interrested, all bug requests are deeply recommended. Cheers, Louis CHEREL.

13 years, 10 months

Jump to page:

2025

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