15 Sep
2007
15 Sep
'07
12:39 a.m.
Hi all,
Recently I've been working on 4 new tracks.
I've been trying new experiments.
http://www.archive.org/details/Coloured
More electro an crazyness than usual in:
- My super hero
- Telemorning bells
More feelings than usual, mainly using my voice:
- Put it away (this one is fresh from today :-) )
- Lights on
These are very early drafts, but I would appreciate your opinion, comments
in the very beginning so I can have an idea on how to continue them..
Thanks!
--
julien
15 Sep
15 Sep
2:04 a.m.
New subject: Important test:
Folks, I am on something of a quest here. I have been suffering clicks in the
audio stream here for sometime now. I have changed my audio HW from 3 x
ice1712 cards to a Firepod firewire, changed my Distro 3 times, changed
everything I know about kernels, jack, alsa etc etc...; all to no avail. The
click manifests when duplex recording from ardour back into ardour ( or
traverso, mhwaveedit etc... this is not ardour specific)
So what I would like to ask is can others please carry out this simple test.
Record some tone first. (you could just install Jaaa.) Generate 20 minutes
of tone into 4 Ardour (or whatever) tracks. Then playback the tone whilst
recording it back into another 4 tracks in Ardour.
Do you see any clicks ? They sound exactly like changing an I/O point within
ardour when program material is being played back.. They appear as a spike in
the waveform. On closer inspection I generally see a "flat-line" between a
broken waveform of about 70-110 samples. Sometimes it's smaller, and
sometimes it just a typical "distrotion" of the wave itself. I always get
them. Sometimes it goes for 3-4 minutes without one, but I always get them.
Traverso doesn't get them using the alsa driver. This may or may not be
significant. Jack is the standard way of connection here,in collaboration
with Qjackctl, so a test with jack is what I most seek. Tone is the best way
to test audio stability; i wonder how many people have tested with tone ? I
hadn't for some years; when i did i was shocked at the result.
any feedback will be most welcome
g.
HW: tried 32 bit Amd, 64 bit Amd, 64 bit Dual Core Amd, CCRMA, 64 Studio,
Musix, Gentoo, alsa 1.0.13 thru 15 rc1, Jack 103,107x, Jackdmp,Ardour 2.5,
Ardour 99.3, kernel 2.6.18,20,21,22 all the major RT,with/without
SMP,posix,RT,vst support; have i forgotten anything ? probably. but you can
bet I've tried it.
4:18 a.m.
New subject: Important test:
Geoff Beasley wrote:
HW: tried 32 bit Amd, 64 bit Amd, 64 bit Dual Core
Amd,
No Intel hardware?
CCRMA, 64 Studio,
Musix, Gentoo, alsa 1.0.13 thru 15 rc1, Jack 103,107x, Jackdmp,Ardour 2.5,
Ardour 99.3, kernel 2.6.18,20,21,22 all the major RT,with/without
SMP,posix,RT,vst support; have i forgotten anything ? probably. but you can
bet I've tried it.
pure:bolic? dyne:bolic?
--
David
gnome(a)hawaii.rr.com
authenticity, honesty, community
8:48 a.m.
New subject: Important test:
Samstag, 15. September 2007 Geoff Beasley:
Folks, I am on something of a quest here. I have been
suffering clicks in
the audio stream here for sometime now. I have changed my audio HW from 3
x ice1712 cards to a Firepod firewire, changed my Distro 3 times, changed
everything I know about kernels, jack, alsa etc etc...; all to no avail.
The click manifests when duplex recording from ardour back into ardour (
or traverso, mhwaveedit etc... this is not ardour specific)
Geoff, I'd be surprised if anyone could be helpful without knowing what
versions you run (jackd, ardour, kernel etc.), how your access to realtime
capabilities is set up, what your jackd invocation looks like etc.
So what I would like to ask is can others please carry
out this simple
test. Record some tone first. (you could just install Jaaa.) Generate 20
minutes of tone into 4 Ardour (or whatever) tracks. Then playback the tone
whilst recording it back into another 4 tracks in Ardour.
I can't believe I actually did this. 20 minutes sine tone out of jaaa into 4
ardour tracks and then another pass with 4 tracks recording from the first 4:
No clicks, no glitches. And I'll be hearing 300hz for the rest of the week.
This must be Dick Cheney's idea of Good Clean Fun.
Versions here are contemporary: Ardour/GTK 2.0.5 (built using 2234 and GCC
version 4.2.1) and jackd 0.103.0 and opensuse 2.6.22.3-7-bigsmp kernel with
CONFIG_PREEMPT_VOLUNTARY, mind you. I haven't encountered your problems with
different hardware and former versions either.
Do you see any clicks ? They sound exactly like
changing an I/O point
within ardour when program material is being played back.. They appear as a
spike in the waveform. On closer inspection I generally see a "flat-line"
between a broken waveform of about 70-110 samples. Sometimes it's smaller,
and sometimes it just a typical "distrotion" of the wave itself. I always
get them. Sometimes it goes for 3-4 minutes without one, but I always get
them.
I suspect your board and/or an interrupt issue. Just a hunch.
Wolfgang
9:01 a.m.
New subject: Important test:
On Sat, Sep 15, 2007 at 12:48:50PM +0200, Wolfgang Woehl wrote:
Just a quick question: _how_ do you generate a tone with JAAA? I
can't seem to see any options for it.
James
So what I
would like to ask is can others please carry out this simple
test. Record some tone first. (you could just install Jaaa.) Generate 20
minutes of tone into 4 Ardour (or whatever) tracks. Then playback the tone
whilst recording it back into another 4 tracks in Ardour.
I can't believe I actually did this. 20 minutes sine tone out of jaaa into 4
ardour tracks and then another pass with 4 tracks recording from the first 4:
No clicks, no glitches. And I'll be hearing 300hz for the rest of the week.
This must be Dick Cheney's idea of Good Clean Fun.
9:18 a.m.
New subject: Important test:
Samstag, 15. September 2007 James Stone:
Just a quick question: _how_ do you generate a tone
with JAAA? I
can't seem to see any options for it.
You might have to increase jaaa's window-size to see the noise/sine buttons in
the lower right-hand corner.
8:08 p.m.
New subject: Important test:
thanks for the test Wolfy; sorry 'bout your ears !
btw, i ididn't expect you to listen to the code just to analyze the recorded
result !
i did mention the bit's and pieces on the bottom of the email though not very
thorough.
all interupts are the best they can be. i don't suspect HW; the fact that the
wave flatlines is proof the card is recording "time". if it were interupts
then you would lose time.it's the program material that's being interupted
here.... though i do see "broken" waves very occasionally which is what
you'd
see with interupt trouble. it's more like a jack dropout to me but i can't
see how they are happening. i wondered about qjackctl's patchbay's auto
function.... i need to do more tests.
best,
g.
8:55 p.m.
New subject: Important test:
thanks for all the comments.
John Rigg I'm testing now with 512, but the latency will be way too big for
midi/audio for my useage, as you rightly say. i'm sure i have tested this
before but we'll see.
John Emmas, i would strongly recommend the use of 3d accelleration in your
system. without it your cpu will be working too hard hence the drop-outs you
are hearing. Another one to be mindful of is that a lot of mobo's have "fast
write" "graphics turbo something" enabled in the bios by default. this can
cause lots of issues 'cause the graphics card is getting higher priority than
it should and it's pulling cycles away from your pci bus.
Fons, thanks for the tip; however that way you need to watch the readout all
the time instead of just scanning the results after the re-record cycle.
there'll be a solution i'm sure; just hope i don't die before i find it ;)
g.
1:51 p.m.
New subject: Important test:
On Sat, Sep 15, 2007 at 12:48:50PM +0200, Wolfgang Woehl wrote:
I can't believe I actually did this. 20 minutes
sine tone out of jaaa into 4
ardour tracks and then another pass with 4 tracks recording from the first 4:
No clicks, no glitches. And I'll be hearing 300hz for the rest of the week.
A simple test to verify jack + driver + hardware is to make an analog loopback
and feed the tone back into JAAA for display. If the spectrum occasionally
'jumps' with sidelobes appearing around the main line this clearly indicates
a problem.
This must be Dick Cheney's idea of Good Clean Fun.
Or an interrogation technique.
--
FA
Follie! Follie! Delirio vano è questo !
9:02 a.m.
New subject: Important test:
On Sat, Sep 15, 2007 at 02:04:05PM +1000, Geoff Beasley wrote:
Folks, I am on something of a quest here. I have been
suffering clicks in the
audio stream here for sometime now. I have changed my audio HW from 3 x
ice1712 cards to a Firepod firewire, changed my Distro 3 times, changed
everything I know about kernels, jack, alsa etc etc...; all to no avail. The
click manifests when duplex recording from ardour back into ardour ( or
traverso, mhwaveedit etc... this is not ardour specific)
What happens if you increase the latency?
James
9:45 p.m.
New subject: Important test:
On Saturday 15 September 2007, Geoff Beasley wrote:
HW: tried 32 bit Amd, 64 bit Amd, 64 bit Dual Core
Amd, CCRMA, 64 Studio,
Musix, Gentoo, alsa 1.0.13 thru 15 rc1, Jack 103,107x, Jackdmp,Ardour 2.5,
Ardour 99.3, kernel 2.6.18,20,21,22 all the major RT,with/without
SMP,posix,RT,vst support; have i forgotten anything ? probably. but you can
bet I've tried it.
With the -rt kernels, have you also tuned your priority [jack/irq's] setup?
Flo
--
Palimm Palimm!
http://tapas.affenbande.org
16 Sep
16 Sep
1:08 a.m.
New subject: Important test:SOLUTION !!!
On Sunday 16 September 2007 09:45, Florian Schmidt wrote:
With the -rt kernels, have you also tuned your priority [jack/irq's] setup?
Flo
done and dusted !
no, the actual culprit is qjackctl/jack ... if you
enable "auto-refresh-connections" whilst re-recording tone as in my exaple
sit back and watch the droputs ! jack surely shouldn't click when
adding/subtrackting channels....but what do i know ...
I'll be reporting to paul & rui immediately.
thanks to all
g.
(TF; i've got me daw back yyyaaahhhoooo)
2:19 a.m.
New subject: Important test:SOLUTION !!!
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sun, Sep 16, 2007 at 01:07:32PM +1000, Geoff Beasley wrote:
On Sunday 16 September 2007 09:45, Florian Schmidt
wrote:
Congratultions!
Yes, JACK does in fact glitch whenever adding/subtracting "graphs" (devices) to
the system. If I have jack transport running, the glitches can even screw up the timing,
or cause Xruns.
I'm told that jackdmp doesn't have this limitation, but I haven't played with
it myself.
I try to launch everything I might need, before starting a song and/or show, so that there
aren't audible clicks or other artifacts.
- -ken
With the -rt kernels, have you also tuned your priority [jack/irq's] setup?
Flo
done and dusted !
no, the actual culprit is qjackctl/jack ... if you
enable "auto-refresh-connections" whilst re-recording tone as in my exaple
sit back and watch the droputs ! jack surely shouldn't click when
adding/subtrackting channels....but what do i know ...
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
iD8DBQFG7K6ye8HF+6xeOIcRAs9kAKD+dofwuEFRfKF4MTyHmiph/G7FagCg2pf3
EoVur8bQW++sii3fvAEurr0=
=DT60
-----END PGP SIGNATURE-----
11:06 a.m.
New subject: Important test:SOLUTION !!!
On Sat, 2007-09-15 at 21:18 -0700, Ken Restivo wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sun, Sep 16, 2007 at 01:07:32PM +1000, Geoff Beasley wrote:
> On Sunday 16 September 2007 09:45, Florian Schmidt wrote:
>
> >
> > With the -rt kernels, have you also tuned your priority [jack/irq's]
setup?
> >
> > Flo
> done and dusted !
>
> no, the actual culprit is qjackctl/jack ... if you
> enable "auto-refresh-connections" whilst re-recording tone as in my exaple
> sit back and watch the droputs ! jack surely shouldn't click when
> adding/subtrackting channels....but what do i know ...
jackd will always stand a good chance of causing a glitch whenever
anything is done that requires taking a lock on the "graph" (the set of
ports it is currently managing). this includes adding/removing ports,
but it also includes checking current connections.
i don't know why geoff has this option selected - its not on by default,
and for sure, every time you check the current graph state you stand a
good, though not guaranteed, chance of causing an xrun. qjackctl will
refresh its view of things every time jackd notifies it of a graph
change, which is basically what everybody wants. there will be a click
then too, but since this only happens when adding/removing ports or
clients, its not really out of the realm of the expected.
i've never even seen this option till i went to look for it, and sure
enough on my system, which runs at 128 frames/period without an RT
kernel and almost no xruns, it causes audible clicks every time the
auto-refresh happens. its a pointless option as far as i can see, and
should probably be removed/made insensitive unless qjc knows that
jackdmp is in use (which i think (but am not sure) is free of this
problem).
--p
15 Sep
15 Sep
3:22 p.m.
On Sat, Sep 15, 2007 at 04:39:15AM +0200, julien lociuro wrote:
"My super hero" ... I don't know what to say.
"Telemorning bells" is cool. I think you should try to make it
agressive and LARGE (wide space)!
"Lights on" is a nice start, except for the dee-dee ;)
"Put it away" should become more smooth ... could benefit from
a lower tempo, perhaps
--
Thorsten Wilms
Thorwil's Design for Free Software:
http://thorwil.wordpress.com
6331
days inactive
6333
days old
linux-audio-user@lists.linuxaudio.org
14 comments
10 participants
participants (10)
-
david -
Florian Schmidt -
Fons Adriaensen -
Geoff Beasley -
James Stone -
julien lociuro -
Ken Restivo -
Paul Davis -
Thorsten Wilms -
Wolfgang Woehl