jack hangs when exiting apps

2011-05-06 Thread bart deruyter
Hi all,

Running Natty now, skipped a version, so previously using 10.04 but having
the same issues as before.

Let me describe, it is weird, but it is very annoying:

System :  UbuntuStudio 11.04, 4 GB ram, intel 6700 @ 2.66GHz, drive space
enough, nvidia graphics GeForce 7600 GS, external firewire audio card Echo
Audiofire 12, usb2midi 2x2 .

Problem : exiting some audio apps causes jack to hang.

Really weird, no xrun problem, no startup problem, but exit problem. I have
to logout, login again, and then I can start jack properly again.

Apps involved : Rosegarden, Qsampler, PD (perhaps more, but I haven't time
to try them all.

Apps that don't cause the problem : Ardour, Rakarrack, Audio plugins in
general.

The message says something about not being able to stop jackdbus...

Anyone else with this problem, or maybe with a fix?

thnx,
Bart


http://www.bartart3d.be/
-- 
Ubuntu-Studio-users mailing list
Ubuntu-Studio-users@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users


Re: Natty and the Real Time Kernel

2011-05-06 Thread bart deruyter
Hi all,

I've got the same result as Brian David. The generic kernel works quite
well, but has xruns, strangely enough, mostly when doing 'nothing'. So far I
had no xruns because of recording, mixing, using rakarrack etc... The xruns
seem to happen at random.

I do use Unity though, maybe there is something in there that asks some
processes which cause the xruns.

My soundcard is an external one, Audiofire 12. jackd is setup at 48000, 256
frames/period and 3 periods/buffer, at a latency of 16 msec.

Running the low-latency kernel eliminates all xruns... I'll definatly keep
using it, and I do recommend it too.


http://www.bartart3d.be/


2011/5/7 Brian David 

> On Thu, May 5, 2011 at 3:46 AM, Victor henri  wrote:
> >
> > My personal experience is, since 2.6.33, many improvements have been done
> in
> > the low latency kernel; I, as several other people, have reported
> excellent
> > performances of the 2.6.37 and 2.6.38 low latency kernel, that seemed to
> get
> > much closer closer to the rt performance then before. That allows me,
> most
> > of the time to not use anymore the RT kernel and all its related
> problems...
> >
>
> I just installed Natty for the first time tonight.  This time around,
> I went with Xubuntu and installed the Ubuntu Studio packages on top of
> that.  As Scott mentioned, I needed to add myself to the audio group,
> but after this my equipment immediately worked.  Unlike previous
> versions, there was no need to change any configuration files to get
> access to my firewire device, which means that Ubuntu Studio is
> basically working out of the box for me now (or, it would be if I had
> installed from an Ubuntu Studio disc).  Yay!
>
> For a test run, I did some mixing on a recording I'm working on right
> now.  I started out using the generic kernel, and performance was
> surprisingly solid.  There were a few x-runs when starting up or
> switching between applications, but otherwise it was usable.
>
> I then installed Allessio's low latency kernel, and proceeded to mix
> for an hour and half without a single x-run, even when starting up
> applications and switching between them.  Rock on!  So, I can say
> definitively that the low latency kernel gives me better performance
> over generic.
>
> This test was run at 44.1 khz / 512 frames / 3 periods - getting about
> 34.9 msec latency.  The next time I get the chance, I'll set up some
> mics and do a recording test at lower latencies (I'll push it to 128
> frames, which will take it down to 8.71 msec latency.  This is
> something I have been able to do previously using rt kernels) and see
> how it performs.
>
> So far so good.  Thanks for the work, everyone!
>
>
> --
> -Brian David
>
> --
> Ubuntu-Studio-users mailing list
> Ubuntu-Studio-users@lists.ubuntu.com
> Modify settings or unsubscribe at:
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users
>
-- 
Ubuntu-Studio-users mailing list
Ubuntu-Studio-users@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users


Re: Ubuntu Studio Developer Meeting

2011-05-06 Thread Jorge G. Mare
Howdy,

On Wed, May 4, 2011 at 5:27 AM,   wrote:



> So, please join us this Sunday, May 8th @ 14:00 UTC for a Ubuntu Studio
> developer meeting. You can use the linked time zone convertor [1] to find a
> city/state/province near your to understand the local time.
>
> We will be meeting in the #ubuntustudio-devel channel on freenode.net for
> your IRC client. Alternately, you can join us via web browser using
> webchat.freenode.net [2].

Although I am not a developer ( ;) ), I will try to attend. Although
if I were any of you, I would not expect much from me on a Sunday at
7AM. :)

Cheers,

Jorge

-- 
Ubuntu-Studio-users mailing list
Ubuntu-Studio-users@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users


Re: Natty and the Real Time Kernel

2011-05-06 Thread Brian David
On Thu, May 5, 2011 at 3:46 AM, Victor henri  wrote:
>
> My personal experience is, since 2.6.33, many improvements have been done in
> the low latency kernel; I, as several other people, have reported excellent
> performances of the 2.6.37 and 2.6.38 low latency kernel, that seemed to get
> much closer closer to the rt performance then before. That allows me, most
> of the time to not use anymore the RT kernel and all its related problems...
>

I just installed Natty for the first time tonight.  This time around,
I went with Xubuntu and installed the Ubuntu Studio packages on top of
that.  As Scott mentioned, I needed to add myself to the audio group,
but after this my equipment immediately worked.  Unlike previous
versions, there was no need to change any configuration files to get
access to my firewire device, which means that Ubuntu Studio is
basically working out of the box for me now (or, it would be if I had
installed from an Ubuntu Studio disc).  Yay!

For a test run, I did some mixing on a recording I'm working on right
now.  I started out using the generic kernel, and performance was
surprisingly solid.  There were a few x-runs when starting up or
switching between applications, but otherwise it was usable.

I then installed Allessio's low latency kernel, and proceeded to mix
for an hour and half without a single x-run, even when starting up
applications and switching between them.  Rock on!  So, I can say
definitively that the low latency kernel gives me better performance
over generic.

This test was run at 44.1 khz / 512 frames / 3 periods - getting about
34.9 msec latency.  The next time I get the chance, I'll set up some
mics and do a recording test at lower latencies (I'll push it to 128
frames, which will take it down to 8.71 msec latency.  This is
something I have been able to do previously using rt kernels) and see
how it performs.

So far so good.  Thanks for the work, everyone!


-- 
-Brian David

-- 
Ubuntu-Studio-users mailing list
Ubuntu-Studio-users@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users


Re: [OT] tar command

2011-05-06 Thread Gustin Johnson
This worked because the MBR, partition table, and file systems were
intact.  What you were doing was a simple file copy (tar can preserve
file permissions) so that the only real obstacle was disk I/O.

If you have a more serious failure, a simple tar will not be enough.

On Wed, May 4, 2011 at 6:00 PM, Mac  wrote:
> tar has been around for as long as *ux.
>
> So it's had plenty of time to mature.
>
> As with many of the vintage *ux cammand line utilities it can be cryptic
> and not necessarily intuitive to use.
>
> Mac
>
> On Thu, 2011-05-05 at 01:44 +0200, Ralf wrote:
>> Does anybody know something about tar?
>>
>> Very interesting, I deleted a broken Maverick-Natty-Upgrade-Install by
>> # rm -r *, then I copied a backup archive and restored Maverick from the
>> backup by tar xzf *, restoring Maverick takes between 3 and 4 minutes
>> only.
>>
>> How is this possible?
>>
>> The restored Maverick seems to be ok, at least Email etc. is ok, the
>> size, 6.76 GiB is correct.
>>
>> An amazing voodoo trick.
>>
>> I didn't know that it's that fast.
>>
>>
>
>
>
> --
> Ubuntu-Studio-users mailing list
> Ubuntu-Studio-users@lists.ubuntu.com
> Modify settings or unsubscribe at: 
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users
>

-- 
Ubuntu-Studio-users mailing list
Ubuntu-Studio-users@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users


RE: Spectrum3d : software that displays the harmonics of the sound in 3D

2011-05-06 Thread Ralf
OT regarding to Spectrum3d.

On Fri, 2011-05-06 at 11:03 +0200, Victor henri wrote:
> I t seems that having a rt kernel working is already very difficult; I
> believe there is already really hard work behind this... Furthemore,
> in my humble opinion, the preempt kernel becomes more and more
> interesting...

Building the kernel-rt ourself isn't difficult! AFAIK the only known
issue is regarding to the proprietary nvidia driver, which anyway could
cause issues for harder real-time.

It's correct, that more and more of the rt-patch occupies the vanilla
kernel, anyway, it's untrue that a real-time kernel can be replaced by a
preemtion only kernel.

It's correct, that just doing audio work with Ardour2 is without
problems, regarding to hard real-time issues, but if you wish that MIDI
event recordings should have a proper timing, like an analog recording
of a really good musician, than you need a kernel-rt + jack from svn
etc.. A lot of people think that it's impossible that a MIDI event
recording could preserve the groove of a very gifted musician, well,
that's a half-truth. MIDI nearly is possible to do that. Most people
aren't able to hear that MIDI isn't able to do it perfectly. What people
are able to hear, as a less good timing, is jitter. They guess that the
sequencer does play too correct, human touch should be missing, but
that's wrong. You don't need to quantize MIDI events and if hard
real-time works, it's very close to a real musician or an audio
recording. Unfortunately most of the PCs are unable to do hard real-time
and jitter will break the groove. For less gifted musicians it might
sound like the timing is to perfect, but this is wrong, it's bad timing
caused by jitter, that makes MIDI breaking grooves.

I don't like the statement that a PREEMPT kernel could replace a PREEMPT
RT. Even the PREEMPT RT should do harder real-time. Note that on
multi-tasking OS real-time seldom is real-time. The kernel-rt, ASIO etc.
don't do, what's called hard real-time. Hard real-time usually is done
by old computers and stand alone sequencers, e.g. the C64, Atari ST.

Btw. JACK2 from svn is able to do MIDI hard real-time, when using the
kernel-rt. Without JACK2 from svn, there's no chance to get this! Of
cause, I'm talking about hw MIDI and not about soft synth (Or do you
guess Hexter does sound like an old DX7 in the brown metal case? Any
analog synth emulation does sound like e.g. an Oberheim Matrix-1000? I
do use those and other old faithful synth. Btw. I'm a guitarist not a
keyboarder ;).

2 cents,

Ralf


-- 
Ubuntu-Studio-users mailing list
Ubuntu-Studio-users@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users


RE: Spectrum3d : software that displays the harmonics of the sound in 3D

2011-05-06 Thread Victor henri

Hello Aradnix

> O_O wow Congratulations:

Thank you

> I think the develpoment for Ubuntu Studio is very hard, because I see there 
> is really few people working on it, and because is very difficult to follow 
> each release that Cannonical launches. Perhaps it's only my imagination.

I'm not involved in ubuntu Studio develepment, though I'm willing to help; but 
i can imagine that's a hard work indeed


I> n the last FLISOL's edition a friend mine who is musician presented a custom 
distro he did for musicians and talking about Ubuntu studio he thinks the rt 
kernel is not so pulished and debugged as should and he talk about the audio 
and the control did it whit the same tool that Ubuntu, something he considers 
not helpful for audio production and well it makes me thinks a lot of things 
about it.

I t seems that having a rt kernel working is already very difficult; I believe 
there is already really hard work behind this... Furthemore, in my humble 
opinion, the preempt kernel becomes more and more interesting...


> And all this is because I think tools like yours and others should take part 
> of this distro, developed for multimedia creation. I'm not sure how many 
> distros for this task exist, I know this one and fedora Design Suite that 
> it's for me not so complete as Ubuntu studio, but I feel this is a sector 
> where we need to work a lot yet for can compete with professional solutions 
> in other SOs.

Thank you for your comments

Victor
  -- 
Ubuntu-Studio-users mailing list
Ubuntu-Studio-users@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/ubuntu-studio-users