diyAudio (
-   PC Based (
-   -   BruteFIR stability problems (

a_tewinkel 12th April 2013 10:51 AM

BruteFIR stability problems
I'm using a PC with BruteFIR as a crossover for my speakers. It has an AMD Athlon 64 2.6 GHz processor, a 2 GB compact flash card as a hard disk and uses a Terratec DMS 6Fire with optical S/PDIF input. Besides BruteFIR it also runs a PHP/MySQL server with Webfilter, a web interface to generate filters for BruteFIR which was programmed by a friend of mine.

I've recently reinstalled it with the latest Ubuntu server (the old OS was version 9.10) and I have disabled updates on the system to prevent them from messing things up. With the new installation I'm having issues with the stability of BruteFIR. Sometimes I hear some very unpleasant buzzing distortion which sounds like BruteFIR isn't getting enough priority (like the output is getting interrupted for some samples).

CPU load seems to be fine (about 35% idle, even when the buzzing kicks in) and most of the time it's running at only 1.8 GHz. At first I thought the CPU speed step kicked in too late (it was running at about 13% idle @ 1 GHz) so I changed the threshold for upscaling from 95 to 85%. I've also installed a low-latency kernel. At first this seemed to help, but the problem persists (although it occurs less often). The only thing I've noticed which might be strange is that already 95% of the 2 GB hard disk is in use, but I'm not sure if this is a lot for the newer version of Ubuntu.

I know it's a long shot, but does anyone have a guess at what might be the problem? I'm pretty much clueless at the moment...:confused:

a_tewinkel 14th April 2013 09:51 PM

I've tried disabling the speed step but this doesn't help either. Any further suggestions?:headbash:

flexus 15th April 2013 10:39 PM

did you add brutefir to the audio group and the audio group to the realtime user group?
apart from that i got no idea. didnīt play around with it. i belive ubuntu studio got the realtime stuff preconfigured.

a_tewinkel 18th April 2013 12:01 PM

/usr/lib/brutefir/brutefir.real is owned by the root user and group. Is it still necessary to change this to audio/realtime?

qusp 19th April 2013 03:02 AM

if youve only got 5% left for memory swap space on the system drive, that is very likely your problem.

hochopeper 19th April 2013 03:16 AM


Originally Posted by qusp (
if youve only got 5% left for memory swap space on the system drive, that is very likely your problem.

Using CF as swap is a pretty bad idea generally, will not last long before data errors. In the config described I would set up with no swap partition, but ensure there is sufficient RAM.

High memory usage of the total OS is not of concern most of the time, linux will use available RAM when it can help. If the memory usage of the application grows too high, that is a concern.

How much RAM do you have in the PC?

a_tewinkel 19th April 2013 08:24 AM

I do not have configured swap space because the machine has 1 GB of memory, which should be more than enough.

phofman 19th April 2013 09:20 AM

IMO you should troubleshoot and investigate the very reasons for the glitches first, before taking any corrective actions.

I would start with checking for xruns as described in XRUN Debug - AlsaProject . Just enable the xrun dump and watch dmesg or /var/log/syslog. You can test that your xrun dump is enabled by e.g. running aplay at minimum buffer sizes:


aplay -v -D plughw:0 -F 1 -B 1 long.wav
This will give you a tool to monitor the symptoms. Afterwards we can troubleshoot each upper layer of your chain, step by step.

a_tewinkel 24th May 2013 10:32 AM

I haven't found the time to look at this yet. However, my machine doesn't boot up any more after what appeared to be a power surge. So I'll probably have to start from scratch again, perhaps even with new hardware.

All times are GMT. The time now is 03:26 AM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Copyright Đ1999-2016 diyAudio

Content Relevant URLs by vBSEO 3.3.2