Summary: | hydrogen-0.9.0 OSS output only produces noise on PPC | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Michael Kiermaier <michael.kiermaier> |
Component: | New packages | Assignee: | PPC Porters <ppc> |
Status: | RESOLVED FIXED | ||
Severity: | major | CC: | sound |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | PPC | ||
OS: | All | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- |
Description
Michael Kiermaier
2005-01-02 08:02:08 UTC
When I run hydrogen on JACK as root (the above report was as user), I get: # hydrogen [WARNING] PreferencesMng [readXmlString] Using default value in QTStyle Hydrogen 0.9.0 [Jan 2 2005] - Copyright 2002-2004 Alessandro Cominu Compiled modules: (Alsa seq) (FLAC) (Jack) (OSS) (LRDF) Hydrogen comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions. See the file COPYING for details [LadspaFX::getPluginList] reading directory: /usr/lib/ladspa [LadspaFX::getLadspaFXGroup] [LadspaFX::getPluginList] reading directory: /usr/lib/ladspa [ERROR] JackDriver [JackDriver::init] Jack server not running? (jack_client_new) [ERROR] Hydrogen [audioEngine_startAudioDrivers] Error starting audio driver [audioDriver::init()] [WARNING] Hydrogen [audioEngine_setupLadspaFX] m_pSong=NULL [ERROR] Hydrogen [audioEngine_setupLadspaFX] nBufferSize=0 [ERROR] NullDriver [setBpm] not implemented yet and I get only 1 error window: "Error starting audio driver" The OSS behaviour seems to be the same independent from user/root. another update: running hydrogen/jack as root works. (I was stupid enough to run jackd as user and hydrogen as root in my last posting.) So concerning the jack output it seems to be some permission problem. Finally I learned how to handle realtime capabilities on a 2.6 kernel. I got hydrogen + jack working for normal users. There is still the noisy ALSA output. check your sample rates. The term ALSA should have been OSS in my last comment, sorry. In my opinion a wrong sampling rate would result in too low/high and too slow/fast sound, but not in white noise. Nevertheless I checked all available sampling rates (22.05, 44.1, 48) in the OSS output configuration of hydrogen, and get noisy sound all the time (as expected). My guess is still that we have some PPC endianness problem. when you say OSS, I assume you are using the kernel's oss emulation drivers. What happens when you use alsa-oss useland oss emulation? The OSS plugin was passing Little Endian data to OSS which was expecting Big Endian (Or vice versa...). I've added a patch that swaps the data and had it apply for ppc only. |