Per the suggestion of Pedro Lopez-Cabanillas on LAU (
http://www.spinics.net/lists/linux-audio-users/msg69260.html )
> In my opinion, if you run some ALSA sequencer applications with a modern
> Linux/ALSA and your motherboard provides the HPET, the module snd-hrtimer
> should be loaded. It is the alternative to the RTC timer present in older
> machines.
You can apply this patch to standard ALSA config file
/etc/modprobe.d/dist-alsa.conf (module-init-tools) and get the desired
behavior.
/////// /////// /////// /////// /////// /////// ///////
coggie-72-~> diff -c /etc/modprobe.d/dist-alsa.conf.~1~
/etc/modprobe.d/dist-alsa.conf
*** /etc/modprobe.d/dist-alsa.conf.~1~ 2009-10-20 21:25:39.000000000 -0700
--- /etc/modprobe.d/dist-alsa.conf 2010-05-29 14:34:42.000000000 -0700
***************
*** 6,9 ****
#
# Remove the following line if you don't want the sequencer.
! install snd-pcm /sbin/modprobe --ignore-install snd-pcm &&
/sbin/modprobe snd-seq
--- 6,33 ----
#
# Remove the following line if you don't want the sequencer.
! install snd-pcm /sbin/modprobe --ignore-install snd-pcm &&
/sbin/modprobe snd-hrtimer && /sbin/modprobe snd-seq
!
! ##NOTE 5/29/10: above, NPM added "&& /sbin/modprobe snd-hrtimer"
! ## (see https://bugzilla.redhat.com/show_bug.cgi?id=597354
! ## and http://www.spinics.net/lists/linux-audio-users/msg69242.html
! ## "Linux-Audio-User: Re: Is ALSA using hrtimer?" )
! ## before doing this, you get
! ## ROOT-coggie-60-~> cat /proc/asound/seq/timer
! ## Timer for queue 0 : system timer
! ## Period time : 0.001000000
! ## Skew : 65536 / 65536
! ## ROOT-coggie-61-~> cat /proc/asound/timers
! ## G0: system timer : 1000.000us (10000000 ticks)
! ## Client sequencer queue 0 : running
! ## ...
! ## Afterwards (also need to start/stop jackd or otherwise reset snd-seq).
! ## ROOT-coggie-66-~> cat /proc/asound/seq/timer
! ## Timer for queue 0 : HR timer
! ## Period time : 0.001000000
! ## Skew : 65536 / 65536
! ## ROOT-coggie-67-~> cat /proc/asound/timers
! ## G0: system timer : 1000.000us (10000000 ticks)
! ## G3: HR timer : 0.001us (1000000000 ticks)
! ## Client sequencer queue 0 : running
! ## ...
/////// /////// /////// /////// /////// /////// ///////
-- Niels
http://nielsmayer.com
PS: Isn't the current behavior, using the archaic system timer, even
on modern systems, a Fedora "bug"?
PPS: Or is enabling the HRtimer something better handled by a
PlanetCCRMA post-configuration script/package (kinda like 'firstboot'
after installing a new kernel) alongside settings in
/etc/security/limits.conf etc ?
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Last week, I sent an introductory message, about the Fedora Summer
Coding project that I'm proposing. The proposal is almost finished (see
it online at
https://fedoraproject.org/wiki/Summer_Coding_2010_proposal_-_Fedora_Musicia…),
but some serious work remains.
The most important part still remaining is that I need a mentor! I
would really like to have somebody from Fedora and Planet CCRMA, because
of the opportunities it would allow. This co-mentoring possibility has
been approved in principal by the Fedora Summer Coding SIG, and they
also recommend it for this particular proposal.
The responsibilities for a mentor from the Audio Creation SIG would
probably be like these:
- -helping sort through existing documentation to derive a 'recommended
practice' for optimizing Fedora PCs for audio work
- -giving tips on installing and using audio software
- -assist in getting algorithms tested on hardware that I don't have (esp.
multiple sound cards, including FireWire, and external MIDI devices)
- -assist in determining the success or failure of the project as a whole,
which determines whether I get paid
I can't say for sure, but this is my best guess. You can contact the
Fedora Summer Coding people for more information, or see the wiki page,
available at
https://fedoraproject.org/wiki/Summer_Coding_2010#You_are_a_mentor
Thank you for considering my proposal. Any comments that you may have
are welcome!
Christopher Antila.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
iQEcBAEBAgAGBQJL7bC7AAoJEInCktGVqZ8V2KUH/RQU02rzKU5YnCiFep0i/Ega
68GXjuBlJVGF7MOleYlRPTmPmHEdHzY9hZ6HZCZVGMOtQvHG+wICUefp0IEvwWZc
7GlCABU7xSkmNlKqe8PT2WrRCUpcbfcsQ+h5U9LDTvzwnlgVHA2tFKqCoCmg6jLu
JJNDbESRnxUxBzUc4uspgmmoAvGe+W94RUtmznf/thJ0IBic3b/R/UHStIAm8n8+
YImFc6B5aKag3Wraz2EXXYsXkfla4GmRQzzSMWc+wxPjIHQeJwb6/Is85RW0h8x3
qydGwaou1M1KAELs1+W3HLq6/ughxQ8bFKbGiHOLmI0+wm5RqddFUbrV3CBMPwg=
=bVOG
-----END PGP SIGNATURE-----
Hello Christopher
Aria Maestosa (http://ariamaestosa.sourceforge.net/) has a groovy gui if
you prefer to enter notes as sheet music and would like to save as midi
file. Also, it outputs to timidity so you can hear what you are composing...
The latest stable release is 1.2.1 but you will need a tricky rpm
specfile to get it up and running. The attached srpm has been tested on
x86_64 and should build on i686. The source code does not support ppc.
Regards, Simon
Am 15.05.2010 21:08, schrieb Christopher Antila:
> Hello Simon:
>
> Thank you for your comments. These are precisely the kinds of tips that
> would be required of a mentor!
>
> I've been thinking about your Qtractor/Rosegarden comment already, and I
> feel ready to substitute Qtractor for Rosegarden.
>
> I've never heard of AriaMaestosa, and I haven't explored Phasex or
> Yoshimi in great detail.
>
> As for PulseAudio, I think that I'll still cover it. What I should have
> put is that it's difficult to escape, and that most desktop users will
> not want to escape it. This is an issue that will be addressed in a
> chapter about optimizing your computer for use with audio applications.
>
> Again, thank you for your comments. I will add information to my
> proposal page right now.
>
>
> Christopher.
>
> On 05/15/2010 02:32 AM, Simon Lewis wrote:
>
>> Hello Christopher
>>
>> I hope you don't mind me making a couple of comments to your proposal...
>>
>> The most important sequencer is Qtractor - despite Rui's cautionary
>> notes it's a very solid and advanced app that allows you to work with
>> both midi and audio. Rosegarden is perhaps interesting from a
>> historical approach but it's midi only.
>>
>> You noted that you can not escape Pulseaudio in Fedora - you certainly
>> can! Simply remove everything except the pulseaudio-libs... Your PC will
>> feel like it's turbo-charged after you done this!
>>
>> Synthesizers? Yoshimi - the best sounding synth I know. Phasex - very
>> clean sound. Alsa Modular Synth - fun.
>>
>> Typesetting AriaMaestosa
>>
>> One overlooked but very important aspect when creating a pro-audio spin
>> of a distro is to set up common directory locations for plug-ins and
>> make sure that all apps automatically point to them.
>>
>> The two areas that need a lot of effort and code changes is multiple
>> sound cards and getting all apps to work on x86_64 (lib64 savvy).
>>
>> Regards, Simon
>>
>> PS: Versions...
>>
>> AriaMaestosa.x86_64 1.2.1-1
>> ams.x86_64 2.0.1-2
>> phasex.x86_64 0.12.0-0.2
>> qtractor.x86_64 0.4.5svn1517-2
>> yoshimi.x86_64 0.055.3-1
>>
>>
>> Am 14.05.2010 22:21, schrieb Christopher Antila:
>> Last week, I sent an introductory message, about the Fedora Summer
>> Coding project that I'm proposing. The proposal is almost finished (see
>> it online at
>> https://fedoraproject.org/wiki/Summer_Coding_2010_proposal_-_Fedora_Musicia…),
>>
>> but some serious work remains.
>>
>> The most important part still remaining is that I need a mentor! I
>> would really like to have somebody from Fedora and Planet CCRMA, because
>> of the opportunities it would allow. This co-mentoring possibility has
>> been approved in principal by the Fedora Summer Coding SIG, and they
>> also recommend it for this particular proposal.
>>
>> The responsibilities for a mentor from the Audio Creation SIG would
>> probably be like these:
>> -helping sort through existing documentation to derive a 'recommended
>> practice' for optimizing Fedora PCs for audio work
>> -giving tips on installing and using audio software
>> -assist in getting algorithms tested on hardware that I don't have
>> (esp.
>> multiple sound cards, including FireWire, and external MIDI devices)
>> -assist in determining the success or failure of the project as a
>> whole,
>> which determines whether I get paid
>>
>> I can't say for sure, but this is my best guess. You can contact the
>> Fedora Summer Coding people for more information, or see the wiki page,
>> available at
>> https://fedoraproject.org/wiki/Summer_Coding_2010#You_are_a_mentor
>>
>> Thank you for considering my proposal. Any comments that you may have
>> are welcome!
>>
>>
>> Christopher Antila.
>>
> _______________________________________________
> music mailing list
> music(a)lists.fedoraproject.org
> https://admin.fedoraproject.org/mailman/listinfo/music
>
>
Hi everyone. We have 1.5 updates:
guitarix-0.08.0:
F-13: https://admin.fedoraproject.org/updates/guitarix-0.08.0-1.fc13
F-12: https://admin.fedoraproject.org/updates/guitarix-0.08.0-1.fc12
F-11: none. sorry, gtk2 version of F-11 is too old to build
guitarix-0.08.0. gtk2 maintainer refuses to update it.
faust-0.9.22:
only updated in rawhide (F-14). I couldn't find any documentation
about the changes. Apparently this is a not-much-tested version just
to present at LAC. There will probably be a more documented version
soon. I am planning to build it on F-13 only. Does anyone need this on
earlier releases of Fedora?
As usual, please allow a day or two until these arrive at the mirrors.
To give positive or negative karma, please visit the above pages.
Orcan
On Fri, May 14, 2010 at 10:35 AM, Sean Beeson <seanbeeson(a)gmail.com> wrote:
>
> Which version of Fedora are you running Mixxx on in the case you
> mention using it? Curiously, I will be trying to setup Mixxx for a
> community radio station.
>
I'm using 2.6.32.11-99.fc12.x86_64 and
"mixxx.x86_64 1.7.0-1.fc12
@rpmfusion-free"
It has one major issue that needs to be worked around -- there appears to be
an X error that messes up the waveform display on the first deck when run
against its own local display. Running to a remote X server it dies with
this error:
Debug: [Main]: createWaveformViewer()
Debug: [Main]: WaveformViewerFactory :: Creating new visual waveform
Debug: [Main]: WaveformViewerFactory :: Making new GL context.
Warning: [Main]: X Error: BadMatch (invalid parameter attributes) 8
Major opcode: 1 (X_CreateWindow)
Resource id: 0x70
mixxx: xcb_io.c:176: process_responses: Assertion `!(req && current_request
> && !(((long) (req->sequence) - (long) (current_request)) <= 0))' failed.
Abort (core dumped)
You can run it locally ($DISPLAY=":0.0") with a simple work-around: each
time you start the app, you have to go
to Options->Preferences->Interface->Skin and RESELECT the desired skin. This
repaints the damaged waveform display area and allows the app to run as
expected.
The other important thing is that Mixxx isn't particularly stable driving
ALSA -- but works very well using JACK. In particular, when driving ALSA
directly, it doesn't like running with CPU-frequency scaling set so it's a
good idea to run "/usr/bin/cpufreq-selector -g performance" prior to
activating mixxx. I don't think it suffers from this issue when running
through Jack.
I guess I should file a bug report since the version supplied on fedora
isn't even the latest version. http://www.mixxx.org/download.php shows the
current stable version at 1.7.2 and a 1.8.0 experimental version available.
If anybody wants to build and package the latest and greatest mixxx for
fedora, that would be awesome.
Niels
http://nielsmayer.com
http://lalists.stanford.edu/laa/2010/04/0017.html
> Remember that piece of junk that I rig on every desktop which still has
some PCI card slots anyway? If you don't, I'll tell you once more: the
Yamaha DB50XG piggyback. I know it's not easy find those nowadays, but
there's always that auction sites for you to try. Moreover the hosting
sound-cards which exposes a Creatine Labs Waveblaster daughter-board
expansion port, are no easier too. Ah, the good old nineties...
> Never mind, so here it goes again:
> QXGEdit 0.1.0 has fallen from the attic
Note these binaries probably work on Fedora too:
http://downloads.sourceforge.net/qxgedit/qxgedit-0.1.0-2.rncbc.suse112.i586…http://downloads.sourceforge.net/qxgedit/qxgedit-0.1.0-2.rncbc.suse112.x86_…
The latter "worksforme" in Fedora 12:
> sudo rpm -ivh qxgedit-0.1.0-2.rncbc.suse112.x86_64.rpm
Preparing... ###########################################
> [100%]
1:qxgedit ###########################################
> [100%]
This synth is a daughterboard card for
http://en.wikipedia.org/wiki/Creative_Wave_Blaster ... Well reviewed ...
http://www.soundonsound.com/sos/1996_articles/may96/yamahadb50xg.html and
still available
http://cgi.ebay.com/Yamaha-DB50XG-DB60XG-Sound-Daughter-Board-SW1000XG-/200…
.
This program looks nice enough that I almost want to find one of these cards
and daughterboard for sale cheap on ebay. I wish I had patch editors on
linux that look as nice as qxgedit.
Niels
http://nielsmayer.com
PS: Anybody know whether this can control other Yamaha GM chips -- like the
one in my Suzuki Q-Chord or Yamaha DD-55??
I did a quick test and although the lights on the midi box flashed, the
sound didn't change any. That's why I tested this out in the first place.
Hello:
I am a Canadian university student who would like to take part in the
Fedora Summer Coding process. My academic background is in Music, but I
have taken some courses in Compuer Science. My idea is to write a guide
for musicians using Fedora (see the idea page here:
https://fedoraproject.org/wiki/Summer_Coding_2010_ideas_-_Fedora_Musicians'_Guide).
The idea, basically, is to get people started with the various programs
available in Fedora and Planet CCRMA. The idea page is far from a
complete project, so any suggestions will be greatly appreciated. What
I need now is a mentor to guide me through the project! Ideally, that
task would be split between people from the Docs SIG, and the music
community.
As before, any comments will be helpful - thank you for reviewing this
project.
Christopher Antila.
N.B. I posted this to the docs, music, and summer-coding-discuss lists
to get what seems like the widest applicable audience
seen in LAU:
https://bugs.launchpad.net/ubuntu/+source/pulseaudio/+bug/178442
<https://bugs.launchpad.net/ubuntu/+source/pulseaudio/+bug/178442>
http://www.mail-archive.com/pulseaudio-discuss@mail.0pointer.de/msg01082.ht…
gives a solutiion (see all messages).
http://www.mail-archive.com/pulseaudio-discuss@mail.0pointer.de/msg01155.ht…
gives the cause:
> PA knows no suitable default channel map for devices that have 10
channels (in contrast to 1, 2, 4, 5, 6, 8). ALSA doesn't know either,
and we default to the ALSA channel maps.
> I am not really sure what I should be doing in this case.
> Does you device have any implicit channel mapping that we could adopt?
> Lennart
Comment: Seems like this is just a pain old "distro" bug. If the ICE 1712/24
gets supported by the "kernel", the distro needs to carry whatever
additional configuration files needed to give all it's channels proper ALSA
names and the same for pulseaudio. So if
kernel/sound/pci/ice1712/snd-ice1712.ko and
kernel/sound/pci/ice1712/snd-ice1724.ko ship with a distro, additional ALSA
and Pulse config files for them should ship with the distro as well.
Niels
http://nielsmayer.com
PS: I haven't tried reinstalling pulseaudio to find out if this is still an
issue in Fedora . Maybe it's already fixed?