Linux-audio-user

linux-audio-user@lists.linuxaudio.org

3 participants
20662 discussions

by mehrdad ghassempoory

I want to write a small audio signal processing application. What I want to do is to successive blocks of N samples, process them and display the results in real time. The processing takes a lot more time than sampling interval, however it can be comfortably done in the time it takes to collect N samples. That is why I like to have an arrangement where I can queue two or three buffers for audio input and process the first full buffer in plenty of time while the audio software is filling in the next buffer. As soon as I finish processing the buffer, I could queue it to be filled again. I would also like use callbacks as much as possible. I have tried PORTAUDIO and it does not seem to do multi-buffering I have tried JACK briefly but it suffers from lack of documentation and as far as I could gather it had no facilities for this kind of thing. I know I could write a higher level interface for PORTAUDIO to simulate multi-buffering but it means I have to copy input buffer into intermediate buffers to simulate multi-buffering and apart from synchronisation problems that I may encounter, it is a solution that appears very inelegant to me. Is there anything out there which can do multi-buffer audio I/O ? Any pointers to web pages, advice or admonition is very welcome!

10 years, 4 months

Re: [LAU] Behringer and Linux

by Len Ovens

On Fri, 19 Jun 2015, Felix Homann wrote: > have 16. The motu 8M has only 8 but is a class compliant USB2 IF and > can > be controled by OSC (I downloaded the OSC spec out of interest) or by > browser. > > > Oh, cool, I didn't know that MOTU has become more open to standards! I think it is the AVB series that is more standards compliant. Even still the AVB part does not allow everything AVB can provide... but that is bound to be true with almost any AVB box. The Mac AVB implementation is a bit strange too. It will talk to multiple AVB boxes but then keeps those boxes from talking to each other. It would be through the motu's NIC port that OSC (or browser) control would be done, not through usb. The audio and maybe MIDI (if the box has it) go through USB (or AVB if your computer supports it) But control is network only. -- Len Ovens www.ovenwerks.net

10 years, 5 months

Debian packages for Advanced Gtk+ Sequencer v0.4.2 and many more ...

by Joël Krähemann

Hi all Advanced Gtk+ Sequencer v0.4.2 was improved in many ways and offers for now debian packages. Please visit: http://gsequencer.org/download.html You may for now import MIDI files using provided `midi2xml` and `xsltproc`. Here's a brief howto: $ midi2xml my_midi_export.mid > my_midi_export.xml $ xsltproc -o my_midi_export.xml /usr/share/xml/gsequencer/stylesheet/ags-xsl/midi-xml/ags.xsl my_ags_file.xml $ ags --filename my_ags_file.xml Note, not all notation may be visible since the lack of a full Grand Piano in Soundfont2 files provided by GNU/Linux. To fix it I would recommend Swami. Please visit: http://www.swamiproject.org bests, Joël Krähemann

10 years, 5 months

Re: [LAU] Kurzweil PC3/K KB3 setup for x42-whirl

by Robin Gareus

On 06/19/2015 07:46 PM, Christopher Arndt wrote: > Am 19.06.2015 um 19:40 schrieb Robin Gareus: >> The speed (chorale/off/tremolo), yes, you can. Press 'm' to view current >> midi CC assignments. Ctrl+middle-click (same as ardour) to re-bind. >> >> Other Leslie parameters can currently only be bound to midi-CCs using a >> text config file (see setBfree --help or `man setbfree`). > > I was talking about x42-whirl not setbFree. Does it work the same? How > can you pass a config file to it? > > Chris > Nope. It does not have a Midi-input to begin with. It's a pure audio processor. I'm sorry for talking x-purpose here. best, robin

10 years, 5 months

Re: [LAU] Behringer and Linux

by mark hadman

Len: > I was looking at the X(R)18. Which acording to Sweetwater has 16 i/o > but on the Behringer site says 18/18. I would hope the manufacture > is the right one. Actually, the sweetwater site says 16 USB channels > in one place and 18 io half a page away... so I can see both numbers > at the same time. :P Fons: > just a quick remark on USB and Behringer. At my new workplace our > boss, in a fit of enlightment and for reasons only known to himself, > decided to buy a cheap Behringer mixer with a USB interface. Nobody > here needs it and it remained in its box for months. > Some week ago I had some spare time and decided to have a look at > it. Connected the USB to my Linux workstation and connected some > headphones to the mixer. Result: with all faders down and just > the HP volume turned up there was a high level 50/100 Hz in the > headphones. Removing the USB connection was enough to stop it. Patrick: >> That is not bad considering that only a few years ago the idea of an >> affordable plug and play digital hardware mixer on Linux was just that, >> an idea. Fons: > It's not a digital mixer, nor did I suggest it was. Just re-editing and recapping this, as I too was confused. Fons was not talking about the XR18 or any of Behringer's new digital range. mH

10 years, 5 months

Yoshimi V1.3.5 (out now) - more details

by Will Godfrey

A recap of my earlier post, but with much more detail. In response to suggestions at LAC 2015 we have made the title bars of all editing windows display both the part number and the current name of the instrument you are working on. In the addsynth oscillator editor you also see the number of the oscillator you are editing. Also, in response to suggestions, horizontal as well as vertical mouse dragging can be used to set rotary controls. Additionally, the mouse scroll wheel can be used, and if you hold down the 'ctrl' key you can get very precise setting. Another request we had was for the part effects window to have the same layout as System and Insertion effects. This has been done and it is now almost identical to Insertion effects. The most noticeable GUI enhancement is colour coded identification of an instrument's use of Add Sub and Pad synth engines, no matter where in the instrument's kit they may be. This can be enabled/disabled in the mixer panel. It does slow down yoshimi's startup, but due to the banks reorganisation (done some time ago) it causes no delay in changing banks/instruments once you are up and running. Some saved instruments seem to have had their Info section corrupted. Yoshimi can detect this and step over it to find the true status. Also, if you resave the instrument, not only will the PadSynth status be restored, but Add and Sub will be included, allowing a faster scan next time. Requests that don't look so likely: Oscillator sync: Not possible on SubSynth - no oscillators! - or PadSynth (Reset the wavetables every cycle?) and debatable on AddSynth due in part to the variable nature of what oscillators you might set up and the complexity of providing an interface for all of this. Subsynth filter levels: Changing apparent sensitivity would be potentially visually confusing as well as possibly changing the the resolution. Further confusion would be caused by the fact that these are not absolute levels but normalised ones, so changing the one with the greatest amplitude changes the relative level of all the others. Instrument kit note range: A cross fade overlap looks interesting until you consider 5 or 6 kit items all crossing over at different points, and different ranges. This is assuming you only want to control and maintain a reasonably consistent volume. In Yoshimi V1.3.5 a number of existing, as well as new features have come together to give much greater flexibility (especially for automation) using standard MIDI messages. These are: NRPNs. ZynAddSubFX controls. Independent part control. 16, 32 or 64 parts. Vector Control. Direct part stereo audio output. NRPNs can handle individual bytes appearing in either order, and usually the same with the data bytes. Increment and decrement is also supported as graduated values for both data LSB and MSB. Additionally, ALSA sequencer's 14bit NRPN blocks are supported. Zyn. System & Insertion Effect controls are fully supported with extensions to allow you to set the effect type and (for insertion effects) the destination part number. Independent part control enables you to change instrument, volume, pan, or indeed any other available control of just that part, without affecting any others that are receiving the same MIDI channel. This can be particularly interesting with multiple layered sounds. There are more extensions planned. With 32 and 64 parts it helps to think of 2/4 rows of 16. When you save a parameter block the number of parts is also saved, and will be restored when you reload. By default each *column* has the same MIDI channel number, but these can be independently switched around, and by setting (say) number 17 taken right out of normal access. In tests, *compiling* for 64 parts compared with 16 parts increased processor load by a very small amount when Yoshimi was idling, but this becomes virtually undetectable once you have 8+ instruments actually generating output. In normal use, selecting the different formats makes no detectable difference but using the default 16 reduces clutter when you don't need the extra. Vector control is based on these columns giving you either 2 (X only) or 4 (X + Y) instruments in this channel. Currently the vector CCs you set up can (as inverse pairs) vary any combination of volume, pan and filter cut-off. More will be added. To keep the processor load reasonable it pays to use fairly simple instruments, but if you have sufficient processing power it would be theoretically possible to set up all 16 channels with quite independent vector behaviour! Direct part audio is Jack-specific and allows you to apply further processing to just the defined parts, which can still output to the main L+R if you want to. This setting is saved with parameter blocks. Currently it is only set in the mixer panel window, but it will also eventually come under MIDI direct part control. In this release (again, to reduce unnecessary clutter) part ports are only registered with Jack if they are both enabled, and set for direct output. However, once set they will remain in place for the session to avoid disrupting other applications that may have seen them. Yesterday someone reported a build problem with fltk 1.1, so in our sourceforge tickets section there is a patch to fix it. I think most distros have been on 1.3 for some time. There is another patch to deal with a minor annoyance connected with gui refreshes when doing a state reload - something I've never used myself :o -- Will J Godfrey http://www.musically.me.uk Say you have a poem and I have a tune. Exchange them and we can both have a poem, a tune, and a song.

10 years, 5 months

jack transport -> mtc, ssp or ltc

by Atte

Hi I'm trying to get reaper to sync with jack transport. However, reaper only understands mtcm, ssp and ltc. Is there a tool that will sync to jack transport and generate one of these sync signals? FWIW: I tried compiling mtc-tools, which depends on libtimecode, which fails like this: atte@skagen:~/software/libtimecode$ ./build-deb.sh <snip> Finished running lintian. Now signing changes and any dsc files... signfile libtimecode_0.0.0-1.dsc Robin Gareus <robin(a)gareus.org> gpg: skipped "Robin Gareus <robin(a)gareus.org>": secret key not available gpg: /tmp/debsign.etpTwBvY/libtimecode_0.0.0-1.dsc: clearsign failed: secret key not available debsign: gpg error occurred! Aborting.... debuild: fatal error at line 1278: running debsign failed gbp:error: Couldn't run 'debuild -i -I -rfakeroot': debuild -i -I returned 29 The code for mtc-tools is three years old, not sure if it's worth persuing this or there are other better/newer tools out there... Cheers -- Atte http://atte.dk http://a773.dk

10 years, 5 months

Rated Blue - Francis Underwood Blues

by Glen MacArthur

Hi, This is a track preview from an upcoming 'Rated Blue' band EP hopefully released by the end of the summer... An original tune written for pretty much any crooked politician that comes to mind but perhaps currently best embodied by the fictional president Frank Underwood from the Netflix series 'House of Cards'.. Recorded using a custom build of Ardour 4.0.0 with WinVST support with heaping portions of linuxDSP and Calf Plugins. https://soundcloud.com/glen-macarthur/rated-blue-francis-underwood-blues Pete Cox on bass, my son Connor on drums and me doing the rest.. Comments and Critiques welcomed..

10 years, 5 months

Re: [LAU] make a VU meter listen to jack output ports

by Kjetil Matheussen

Peter P: > Hi list, > > I am wondering how I could make a VU meter automatically receive all > audio that is sent to specific jack output ports by various > applications, also from those that I will start later on a given day. > > thank you for all ideas! > Hi, jack_capture is able to do exactly what you ask for. jack_capture is mainly made for recording audio, but to avoid recording anything, just run it in "timemachine" mode: $ jack_capture --meterbridge --timemachine --port system:playback_1 --port system:playback_2 (make sure "meterbridge" is installed)

10 years, 5 months

make a VU meter listen to jack output ports

by Peter P.

Hi list, I am wondering how I could make a VU meter automatically receive all audio that is sent to specific jack output ports by various applications, also from those that I will start later on a given day. thank you for all ideas! P

10 years, 5 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-user