[linux-audio-user] Re: latencies

tim hall tech at glastonburymusic.org.uk
Fri Jul 22 09:15:01 EDT 2005

Last Friday 22 July 2005 05:52, Wolfgang Lonien was like:
> Esben Stien wrote:
> > Wolfgang Lonien <wolfgang at lonien.de> writes:
> >>everything is nice until I hit that 'record' button
> >
> > It would help if you described your system, both hard and soft, to be
> > able to help you.
> Oh oh - sorry I didn't do that until now. Ok:

aadebug can be useful for system stats.

> An ASUS L8400 laptop with P3-750 and 384MB RAM, integrated graphics S3
> Savage (tested with savage or vesa driver)
> A M-Audio Midiman MidiSport 2x2 USB-MIDI interface (which should work ok
> - Dave Phillips from linuxjournal.com uses that same interface)
> A hardware MIDI keyboard (evolution MK-149)
> A Roland Sound Canvas SC-33
> Debian GNU/Linux 3.1 'Sarge' with additional repository from DeMuDi
> 'stable', Kernel 2.6.12-3-multimedia-686 with realtime capabilities
> enabled, Alsa 1.0.8-7, jackd 0.99.51-1, rosegarden4 1.0-1, alternatively
> muse 0.7.2pre1-0

Very similar to the system I'm running. I only get MIDI timing problems when I 
max out my CPU (more than a dozen simultaneous parts in Rosegarden).

Have you read all of this:

> No .asoundrc (maybe that is the cause?); jackd is started with -R -d
> alsa or with jackstart from qjackctl.

It's worth making a ~/.asoundrc - I doubt whether that's the cause though.

> It helps a wee bit if I use Fluxbox instead of Gnome, as does using Muse
> instead of Rosegarden.

Yep, it can also be worth disabling some services. 

> The lags and timeouts (xruns) start being worse when the screen starts
> scrolling in Rosegarden/Muse.

For complex pieces it can be worth turning off the playback follow option and 
various other bits of eye-candy in Rosegarden. It should run smooth for a 
four-part arrangement on your system.

> In the company I tried a dry run (without the MIDI interface) on an ASUS
> nforce2 machine with more or less the same results, tho there it helped
> to break the audio playback connection in qjackctl and leave the capture
> connection only (tho nothing was connected to the sound input). But that
> same test failed on the laptop here.
> The interrupt used for USB is heavily shared with 2 yenta sockets (don't
> even know what they're good for, I assume the cardbus (PCMCIA)?) and acpi:
>            CPU0
>   0:    4702299          XT-PIC  timer  0/2299
>   1:       3432          XT-PIC  i8042  0/3432
>   2:          0          XT-PIC  cascade  0/0
>   3:       2370          XT-PIC  0.0  0/2368
>   8:          1          XT-PIC  rtc  0/1
>   9:      35917          XT-PIC  acpi, yenta, yenta, uhci_hcd:usb1  0/35917
>  10:     229763          XT-PIC  Allegro  0/29763
>  14:      49090          XT-PIC  ide0  0/49090
> And lspci says:
> 0000:00:00.0 Host bridge: Intel Corp. 440BX/ZX/DX - 82443BX/ZX/DX Host
> bridge (rev 03)
> 0000:00:01.0 PCI bridge: Intel Corp. 440BX/ZX/DX - 82443BX/ZX/DX AGP
> bridge (rev 03)
> 0000:00:06.0 Multimedia audio controller: ESS Technology ES1988
> Allegro-1 (rev 12)
> 0000:00:06.1 Communication controller: ESS Technology ESS Modem (rev 12)
> 0000:00:07.0 ISA bridge: Intel Corp. 82371AB/EB/MB PIIX4 ISA (rev 02)
> 0000:00:07.1 IDE interface: Intel Corp. 82371AB/EB/MB PIIX4 IDE (rev 01)
> 0000:00:07.2 USB Controller: Intel Corp. 82371AB/EB/MB PIIX4 USB (rev 01)
> 0000:00:07.3 Bridge: Intel Corp. 82371AB/EB/MB PIIX4 ACPI (rev 03)
> 0000:00:08.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
> RTL-8139/8139C/8139C+ (rev 10)
> 0000:00:0a.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev 80)
> 0000:00:0a.1 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev 80)
> 0000:01:00.0 VGA compatible controller: S3 Inc. 86C270-294 Savage/MX-MV
> (rev 12)

Hmm, maybe worth disabling acpi in the BIOS? I don't really understand how USB 
works. Which bit of hardware uses it? the keyboard?


tim hall

More information about the linux-audio-user mailing list