27 Oct
2012
27 Oct
'12
10:18 a.m.
Has anyone else had this problem? For some pieces using live mic input, I
wanted to reduce latency by pulling the Jack IO buffer size down to 512
samples from the default 1024. (I also tried 256 samples, but that just
made Jack crash.)
But with the smaller buffer, the mic gets an additional 30-40 dB of gain,
making the input unusable. If I turn the preamp on the fast track down very
very low and speak quietly into the mic, the sound comes through without
obvious distortion or glitches, so it doesn't seem to be an xrun thing. But
it distorts very easily and very badly.
If I switch back to 1024 samples, the problem disappears.
Why would the driver's buffer size have an impact on the incoming signal's
gain?
Thanks,
hjh
29 Oct
29 Oct
3:45 a.m.
New subject: M-Audio fast track pro mic input fails with smaller buffer size
At Sat, 27 Oct 2012 16:18:43 +0800,
James Harkins wrote:
Has anyone else had this problem? For some pieces
using live mic input, I wanted to reduce latency
by pulling the Jack IO buffer size down to 512 samples from the default 1024. (I also
tried 256
samples, but that just made Jack crash.)
But with the smaller buffer, the mic gets an additional 30-40 dB of gain, making the
input
unusable. If I turn the preamp on the fast track down very very low and speak quietly
into the
mic, the sound comes through without obvious distortion or glitches, so it doesn't
seem to be an
xrun thing. But it distorts very easily and very badly.
If I switch back to 1024 samples, the problem disappears.
Just reconfirming this issue -- with 512 samples in the hardware buffer, the noise floor
was about -20 dB. At 1024 samples, it was -70 dB.
Baffled (and, I guess, resigned to 46.4 ms latency in my concert...),
hjh
--
James Harkins /// dewdrop world
jamshark70(a)dewdrop-world.net
http://www.dewdrop-world.net
"Come said the Muse,
Sing me a song no poet has yet chanted,
Sing me the universal." -- Whitman
blog: http://www.dewdrop-world.net/words
audio clips: http://www.dewdrop-world.net/audio
more audio: http://soundcloud.com/dewdrop_world/tracks
5:44 p.m.
New subject: M-Audio fast track pro mic input fails with smaller buffer size
On Sat, Oct 27, 2012 at 04:18:43PM +0800, James Harkins wrote:
Has anyone else had this problem? For some pieces
using live mic
input, I wanted to reduce latency by pulling the Jack IO buffer size
down to 512 samples from the default 1024. (I also tried 256
samples, but that just made Jack crash.)
But with the smaller buffer, the mic gets an additional 30-40 dB of
gain, making the input unusable. If I turn the preamp on the fast
track down very very low and speak quietly into the mic, the sound
comes through without obvious distortion or glitches, so it doesn't
seem to be an xrun thing. But it distorts very easily and very
badly.
If I switch back to 1024 samples, the problem disappears.
Why would the driver's buffer size have an impact on the incoming
signal's gain?
Thanks,
hjh
Hi James,
I've too have an M-Audio Fast Track Pro and I've found that it's audio
capture works unreliably under Linux. I only tested it using arecord
and I have not tried Jack.
I find that on random takes the audio will become heavily distorted.
Usually this sounds as though additional gain is being introduced and
there is what sounds like very audible quantization distortion. Once
the problem appears, it persists until the device is reset.
If you are able to successfully configure this device for recording in
any capacity you are doing better than me. I have found the device to
be unusable for capture which is a shame because it's the only device
I own. I'd be interested in testing your configuration on my own
device. I would love to get this thing working.
There don't seem to be many Fast Track Pro users out there; in fact,
you're the only other person I've heard from thus far.
-- Lewis
9:31 p.m.
New subject: M-Audio fast track pro mic input fails with smaller buffer size
On 10/29/2012 05:44 PM, Lewis Pike wrote:
There don't seem to be many Fast Track Pro users
out there; in fact,
you're the only other person I've heard from thus far.
http://joegiampaoli.blogspot.nl/2011/06/m-audio-fast-track-pro-for-debian-l…
Best,
Jeremy
31 Oct
31 Oct
1:20 p.m.
New subject: M-Audio fast track pro mic input fails with smaller buffer size
On Oct 30, 2012 12:44 AM, "Lewis Pike" <phaselocker(a)gmail.com> wrote:
I've too have an M-Audio Fast Track Pro and
I've found that it's audio
capture works unreliably under Linux. I only tested it using arecord
and I have not tried Jack.
I find that on random takes the audio will become heavily distorted.
Usually this sounds as though additional gain is being introduced and
there is what sounds like very audible quantization distortion. Once
the problem appears, it persists until the device is reset.
Thanks for this... so, 46 ms latency in my concert and count myself lucky
it isn't worse. And, replace the ftpro when I can...
If you are able to successfully configure this device
for recording in
any capacity you are doing better than me.
My experience using it with Jack is: if audio input is correct when Jack
starts, then it stays OK as long as the Jack server is running. I have also
seen the problem with default Jack settings, but rarely, certainly less
often than you report. The difference is that I start Jack and keep it
running for a whole session, maybe several hours, while in your case,
arecord connects to the hardware fresh on every take, raising the
probability of a failed take.
I'm fairly certain if you use Jack and keep one connection to the Fast
Track alive for a longer time, the issue will not go away totally, but it's
manageable. Check for good input at the start of the session, and then you
should be OK.
hjh
1 Nov
1 Nov
7:41 a.m.
New subject: M-Audio fast track pro mic input fails with smaller buffer size
Hi!
On 29/10/2012 21:31, Jeremy Jongepier wrote:
On 10/29/2012 05:44 PM, Lewis Pike wrote:
I'm the commiter of the FTP driver in the Linux kernel. I know that the driver
is not perfect, but we, in my company, manage to get really good recording
performance even at ~4 ms latency.
Could you tell me what kind of setup you have (distro, kernel version, jack
version, etc..) ? Also do you use any special modprobe rules described in the
Joe's blog page above ?
cheers,
Yomguy
There don't seem to be many Fast Track Pro
users out there; in fact,
you're the only other person I've heard from thus far.
http://joegiampaoli.blogspot.nl/2011/06/m-audio-fast-track-pro-for-debian-l…
7:41 a.m.
New subject: M-Audio fast track pro mic input fails with smaller buffer size
Hi!
On 29/10/2012 21:31, Jeremy Jongepier wrote:
On 10/29/2012 05:44 PM, Lewis Pike wrote:
I'm the commiter of the FTP driver in the Linux kernel. I know that the driver
is not perfect, but we, in my company, manage to get really good recording
performance even at ~4 ms latency.
Could you tell me what kind of setup you have (distro, kernel version, jack
version, jack parameters, etc..) ? Also do you use any special modprobe rules
described in the Joe's blog page above ?
cheers,
Yomguy
There don't seem to be many Fast Track Pro
users out there; in fact,
you're the only other person I've heard from thus far.
http://joegiampaoli.blogspot.nl/2011/06/m-audio-fast-track-pro-for-debian-l…
5:22 p.m.
New subject: M-Audio fast track pro mic input fails with smaller buffer size
Tried to post this by gmane to keep the thread history accurate, but it
failed for some reason I don't care to discern at the moment. So I'll just
reply to an earlier message of mine.
This is in response to Parisson's request for more info on system
configuration.
Thanks,
hjh
---------- Forwarded message ----------
From: James Harkins <biz(a)dewdrop-world.net>
To: linux-audio-user(a)lists.linuxaudio.org
Cc:
Date: Thu, 1 Nov 2012 14:53:04 +0000 (UTC)
Subject: Re: [LAU] M-Audio fast track pro mic input fails with smaller
buffer size
Parisson <lists@...> writes:
I'm the commiter of the FTP driver in the Linux
kernel. I know that the
driver
is not perfect, but we, in my company, manage to get
really good recording
performance even at ~4 ms latency.
Could you tell me what kind of setup you have (distro, kernel version,
jack
version, etc..) ? Also do you use any special modprobe
rules described in
the
Joe's blog page above ?
Hi (and thanks for the driver!),
I'm on ubuntu 12.04 with lowlatency kernel, Jack 1.9.8.
$ uname -a
Linux dlm-A6200 3.2.0-23-lowlatency #31-Ubuntu SMP PREEMPT Wed Apr 11
02:24:03
UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
Synaptic GUI says jackd2 is 1.9.8~dfsg.1-1ubuntu1
CPU: Intel(R) Core(TM) i3 CPU M 350 @ 2.27GHz
Oddly, after claiming that I couldn't get it to work with a buffer smaller
than 1024, I did get good mic input today with a 512 sample buffer. But
then, SuperCollider started behaving oddly ???? It seemed like some of the
audio processing just wasn't happening, although mic --> reverb --> DAC
sounded fine. I've asked about that on the sc list.
I'm in China, where blogspot is not a happy place. I have one machine with
a proxy service, but that's been down for weeks. So I'm afraid I'm unable
at the moment to read that article. Are there some particular nuggets of
wisdom you could copy and paste to the list? (So, no, I'm not doing
anything special with modprobe. I did find that I have to remove
snd-usb-audio, and then re-probe it, or the FTPro doesn't work, but I
haven't configured any rules [nor do I know how to configure rules :-p ].)
Thanks again!
hjh
4362
days inactive
4367
days old
linux-audio-user@lists.linuxaudio.org
7 comments
4 participants
participants (4)
-
James Harkins -
Jeremy Jongepier -
Lewis Pike -
Parisson