17 Dec
2003
17 Dec
'03
4:51 a.m.
Dear all,
as a user of a "low end" machine (duron 1Ghz, 256MB) I am trying to get
some reverb on the snare drum (living as a soundfont in the Emu10k1
wavetable) being triggered by my MIDI-Sequencer (MusE).
Unfortunately, the hardware reverb on SBlive rev.4 is afaik not
supported, so I wondered whether there is a possibility to route the
wavetable output through JACK into JACK-Rack or something similar?
Simply connecting the MusE output to the JACK-Rack ins in the qjackctl
connection bay does not work (and would btw also put reverb on the
keyboards).
I tried fluidsynth with the jack output, but I can't really get it on
time (too slow, esp. the rolls sound very bad).
Is there any other way to do this?
Thank you,
t.
--
Thorsten Mika mailto: tmika(a)t-online.de
Hamm / Germany TM5173-RIPE
Public Key ID: 41338C37 -- http://www.keyserver.net
20 Dec
20 Dec
10:50 a.m.
On Wednesday 17 December 2003 08:49, Thorsten Mika wrote:
as a user of a "low end" machine (duron
1Ghz, 256MB) I am trying to get
some reverb on the snare drum (living as a soundfont in the Emu10k1
wavetable) being triggered by my MIDI-Sequencer (MusE).
So 1G processors are 'low end' now? Heaven help the rest of us!
I assume you have a channel dedicated to the snare sound (i.e. it's not part
of a drum kit). I don't suppose editing the soundfont is going to 'be your
friend' here. I have often wondered how to do this with MusE.
(tbh: I don't have a 'danny' :-).
sorry
tim hall
20 Dec
20 Dec
12:10 p.m.
On Fri, 19 Dec 2003 12:12:13 +0000
tim hall <tech(a)glastonburymusic.org.uk> wrote:
On Wednesday 17 December 2003 08:49, Thorsten Mika
wrote:
> as a user of a "low end" machine (duron 1Ghz, 256MB)
So 1G processors are 'low end' now? Heaven
help the rest of us!
I do not really think they are "low end". But my experience with jack,
fluidsynth and timidity tells me that it reliable low latency is not
possible. Best thing I can achieve is a latency of 92,9ms (as I said,
with Duron 1Ghz and 256MB RAM). Setting the laten cy lower leads to JACK
"throwing out" apps like ZynAddSubFX.
Or is anybody able to get better results with hardware like this?
t.
--
Thorsten Mika mailto: tmika(a)t-online.de
Hamm / Germany TM5173-RIPE
Public Key ID: 41338C37 -- http://www.keyserver.net
1:32 p.m.
On Saturday 20 December 2003 10:08 am, Thorsten Mika wrote:
On Fri, 19 Dec 2003 12:12:13 +0000
tim hall <tech(a)glastonburymusic.org.uk> wrote:
The Jack implementation in ZynAddSubFX is problematic, and is being
worked on. It is known to "zombify" very easily. I love it anyway :)
Do you have this problem frequently with other Jack applications?
|)
|)enji
On Wednesday 17 December 2003 08:49, Thorsten
Mika wrote:
I do not really think they are "low end". But my experience with jack,
fluidsynth and timidity tells me that it reliable low latency is not
possible. Best thing I can achieve is a latency of 92,9ms (as I said,
with Duron 1Ghz and 256MB RAM). Setting the laten cy lower leads to JACK
"throwing out" apps like ZynAddSubFX. as a user of a "low end" machine (duron
1Ghz, 256MB)
So 1G processors are 'low end' now? Heaven help the rest of us!
2:05 p.m.
On Sat, 20 Dec 2003 11:31:11 -0600
Benjamin Flaming <lau(a)solobanjo.com> wrote:
The Jack implementation in ZynAddSubFX is
problematic, and is
being
worked on. It is known to "zombify" very easily. I love it anyway :)
So do I. It sounds _so_ good.
Do you have this problem frequently with other
Jack applications?
Depends. Sometimes ReZound (whose JACK implementation is said to be not
too good) zombifies, but MusE and QARecord never do. As these are the
only ones I use, I cannot say nothing about other programs.
Do you know whether there is a schedule for ZynAddSubFX and when it
could be made more stable?
t.
--
Thorsten Mika mailto: tmika(a)t-online.de
Hamm / Germany TM5173-RIPE
Public Key ID: 41338C37 -- http://www.keyserver.net
5:15 p.m.
On Saturday 20 December 2003 11:57 am, Thorsten Mika wrote:
Do you know whether there is a schedule for
ZynAddSubFX and when it
could be made more stable?
There was some serious investigative/experimental work done a week or two
ago. The CVS version now has two Jack drivers - the original (which is very
responsive, but zombifies easily) and a new one which is slightly slower to
respond (due to additional buffering), but which doesn't zombify. I've spent
a little bit of time looking over the code in an effort to come up with a
better implementation, but the author is currently making some big changes to
the internals of the synth, so I'd better wait until things stabilize a bit.
If you're interested, I could try sending you my slightly hacked
ZynAddSubFX binary (offlist, of course). It coughs and sputters a bit when
it gets pushed too hard, but it uses the older, more responsive Jack driver,
and I've managed to use it for a couple hours with Jack buffer size set to
64! I could send you the source code, but that might take a few days, since
I'd have to dig around in order to figure out which source tree built the
binary I put into my /usr/bin directory (assuming I still have a copy of the
source tree from which I built it.... ;)
|)
|)enji
21 Dec
21 Dec
7:03 a.m.
On Sat, 20 Dec 2003 15:14:15 -0600
Benjamin Flaming <lau(a)solobanjo.com> wrote:
On Saturday 20 December 2003 11:57 am, Thorsten Mika
wrote:
> Do you know whether there is a schedule for ZynAddSubFX and when it
> could be made more stable?
If you're interested, I could try sending you
my slightly hacked
ZynAddSubFX binary (offlist, of course).
Yes, please :-) The "From:"-Address works. This could improve things a
lot for me.
Thank you,
t.
--
Thorsten Mika mailto: tmika(a)t-online.de
Hamm / Germany TM5173-RIPE
Public Key ID: 41338C37 -- http://www.keyserver.net
7696
days inactive
7700
days old
linux-audio-user@lists.linuxaudio.org
8 comments
4 participants
participants (4)
-
Benjamin Flaming -
Guy Daniel CLOTILDE -
tim hall -
tmika＠t-online.de