Linux Audio the way to go!?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
NS-320U USB Storage NAS Print File FTP BT DLNA server on eBay.ca (item 320602491060 end time 12-Oct-10 18:56:16 EDT)

Hey, I bought one of these units on ebay. It's a pretty handy little device.

I already put snake os, and installed mediatomb (on a usb stick plugged into the unit.)

If I were to rip my audio onto a usb hard disk, might I be able to stream music to a device like the squeezebox?

Thanks,
JG

Hi JG,
Do you know if in the snake-os kernel there is the module for usb audio?
Have you tried with some usb audio dac like those based on pcm270x or pcm290x for example?

Thanks
Andrea
 
Member
Joined 2004
Paid Member
For those of you interested in using a Linux based audio player but a little overwhelmed at the complexity my client (Auraliti) has decided to offer the entry level player as a kit needing only some screwdriver experience to assemble. It will be a little cheaper (but still more expensive than prowling the web for components). It would get you running quickly with a system to use. Its based on Voyage Linux and MPD. I may be persuaded to offer more insights into tweaks for it (as long as I don't need to fix the consequences of mangling the system).

It also seems we may finally be getting a review. Chris Connaker of Computer Audiophile tweeted that he is listening to the PK100.
 
Installing Hammerfall DSP 9632 on VortexBox?

In system information the sound-card is listed under hardware:

PCI Devices

- Multimedia audio controller: Xilinx Corporation RME Hammerfall DSP

But it doesn't list under vortexplayer.

Dos anybody on this forum have experience on how to install this audio-card on vortexbox? A detailed instruction is much appreciated.

Her is result from lsmod:
> lsmod
Module Size Used by
appletalk 27128 20
psnap 2708 1 appletalk
llc 4944 1 psnap
nfsd 206692 9
lockd 59416 1 nfsd
nfs_acl 2472 1 nfsd
auth_rpcgss 31020 1 nfsd
sunrpc 155680 11 nfsd,lockd,nfs_acl,auth_rpcgss
ipv6 235720 78
xfs 418100 1
exportfs 3620 2 nfsd,xfs
dm_multipath 14044 0
snd_usb_audio 75960 1
snd_hdsp 42380 0
snd_seq 45696 0
ppdev 6500 0
snd_pcm 63236 2 snd_usb_audio,snd_hdsp
i915 155204 1
snd_usb_lib 14348 1 snd_usb_audio
snd_rawmidi 18980 2 snd_hdsp,snd_usb_lib
drm 169328 1 i915
snd_timer 17760 2 snd_seq,snd_pcm
snd_seq_device 6216 2 snd_seq,snd_rawmidi
snd_page_alloc 7784 2 snd_hdsp,snd_pcm
snd_hwdep 6732 2 snd_usb_audio,snd_hdsp
i2c_algo_bit 4860 1 i915
snd 50292 11 snd_usb_audio,snd_hdsp,snd_seq,snd_pcm,snd_usb_lib,snd_rawmidi,snd_timer,snd_seq_device,snd_hwdep
video 18752 1 i915
output 2484 1 video
serio_raw 5096 0
r8169 28444 0
i2c_i801 10212 0
pcspkr 2176 0
parport_pc 22616 0
soundcore 5484 1 snd
mii 4052 1 r8169
i2c_core 25032 4 i915,drm,i2c_algo_bit,i2c_i801
parport 28936 2 ppdev,parport_pc
ata_generic 4344 0
pata_acpi 3696 0
 
Dear Soundcheck,

If you are still on this forum (despite some people who'd apparently rather see you off :) ) I'd like to thank you for a huge number of extremely informative posts!
You have managed to inspire my curiosity to try Linux even though it seems to be a gigantic pain in the butt.
As I'd like to minimize the pain in that place may I ask you the following:

- what distro do you recommend to someone who'd like to compare the BEST linux has to offer to best he already had heard on Windows? (in terms of audio playback only!) Is it still Mint or is it safe going with TinyCore etc (I don't care if I only have character-based input: I'm just interested in hearing at least 90%+ of what Linux can offer in shortest period of time and with minimal frustration - if I like it, I'm happy to invest more time in investigating/customizing in the future)
- as you seem to have moved to SB I'm curious as to whether you consider SB superior in terms of sound quality to your previous Linux setup?

Cheers
 
If you are still on this forum

I'm still around.

(despite some people who'd apparently rather see you off :) )

Who's that supposed to be!?!? ;)


- what distro do you recommend to someone who'd like to compare the BEST linux has to offer to best he already had heard on Windows?

"BEST" !?!?!

I'd still go for Ubuntu/Mint. I've tried several other Distros incl. Vortexbox recently and those little ones ( some people seem to push special purpose Linuxes Voyage with MPD in pretty much every forum you might think of ;) ). I wouldn't go for that either.

99% of the Linux Distros around are more buggy, less flexible and suffer of poor
support compared to e.g. Mint.

The main issue with Ubuntu/Mint is Pulseaudio. Though there is a new Mint-Debian, which should run without Pulseaudio.

I just installed the new Mint 10 as desktop and server OS . Very nice.

However getting e.g. a HDSP9632 ( one of the supposedly well supported cards under Linux!!) to work properly even on a Mint 10 is IMO almost impossible for a non-hacker.

I'm still considering the Linux Sound Layer a pretty poor part of Linux in general terms.
( I just learned that the WLAN card support is at least as poor as the soundlayer)

In case your sound card works and you're able to manage that card properly. You can give Linux a try.

Until know my recipe for best sound was:

1. rt-kernel
2. ecasound as player with rt-priotities routed to the "hw" sound device of your card.
3. file stored 100% on /dev/shm
4. all this run in recovery mode ( no X active and the other 100dreds of apps stopped)


- as you seem to have moved to SB I'm curious as to whether you consider SB superior in terms of sound quality to your previous Linux setup?

Yep I do consider my Touch setup superior compared to my earlier Linux and Windows setups.
Finally no more hazzle with Alsa ( I do have some things to tweak here also on the Touch -- phofman likes the buffer tweaks best :D ) .
The SB setup is OS independent. The sound is better ( after tweaking that device) then ever. The database management and client control is just great.
MPD looks IMO like a little toy in comparison.
And I do get a streaming client plus an integrated soundcard at a price of 250$.

What else do I want? No idea.
Since half a year I 'm more then happy with my setup. I even presented tht setup with great success on the largest DIY-fair here in Germany some weeks ago.

There are a thousand different ways to achieve your goals. There is no right or wrong.
There is this or there is that subjective preference. Above are my current preferences
after 4 years digging into all that.

Cheers
 
I'd still go for Ubuntu/Mint.

Thanks - I read your earlier posts and installed Mint 10 as well. Before that I tried AV Linux - it is realtime (unlike stock Mint) but I was unable to install ecasound - tried aplay instead and was not impressed at all.....:(

Your Mint suggestion with ecasound (indeed - very easy to install!) performed MUCH better - and that's without optimizations you describe in your wiki!
Unfortunately it seems my USB stick (where I installed Mint) seems to have some problems as Mint won't boot anymore even in recovery mode so I'll have to try again.
However- I have to say that initial impressions were positive enough to warrant exploring your 'advanced' wiki settings - I'll keep you posted!

Until now my recipe for best sound was:

1. rt-kernel

I've read in your earlier posts that you preferred CK patch to 'stock' Molnar rt version - Does that still stand and have you tried BFS?
If not, do you have any pointers how BFS can be applied to Mint?

Yep I do consider my Touch setup superior compared to my earlier Linux and Windows setups.

Wait - I understand it is superior in terms of ease-of-use but are you saying it also sounds better? (and in what way?)
If so, that would be something really interesting - I guess I'll have to try that thing too!

Cheers!
 
Soundcheck,

I see that your SBTouch mods web pages are under maintenance. Is it possible to publish your older mods somewhere on the website, so that I can catch up on the original mods while your next rev of goodies is being prepared?

I'm already enjoying my SBTouch with an external DAC, fed via a server with brutefir EQ (waiting for calibrated mic to try out DRC). I am very keen to see (hear) how your various mods work with my config.
 
Hi,

After many years away from Unix / Linux environments (used freebsd and sun Sparc's on a basic level) I installed Ubuntu yesterday. Boy have things changed this really is a modern OS. Anyway before i go too far down this path and since my knowledge of such things is gone I have a couple of questions.

1. I installed Ubuntu 10.10 it looks like the hard realtime kernel for this is no longer supported so will I have to run an older distro and if so which one?

2. I Run a cut down streamlined xp os atm with vst plugins for active crossover duties on 3 way open baffles using a prodigy 7.1 hifi and asio. What do you suggest for active crossover duties that will mimic or be on a par with fabfilter pro?

before i start tweaking i'd like to make sure that capable software is available.
 
Hi there.

BFS-kernel:
Honestly I never tried it. It seems that Ingo Molnar integrated parts of that BFS logic in his newer schedulers.

Ubuntu/Mint rt-kernel:
The maintainer left the team. ( However it seems that the ppa for "Natty" is back. Currently available is a 2.6.36 kernel with preemption enabled or a 2.6.33 rt-kernel:

Install them according to here: Maverick rt-kernels.

Just click on the links at above site, download the packages and install them with "sudo dpkg -i xxx" in top to bottom order.

SB Touch:
You'll find the updates on my page soon. I've been quite busy during the last two weeks.


Multichannel&VST:
Check out Ubuntu Studio Wiki's and Jack. USTudio Prep

Mint vs. Ubuntu:
If you like Ubuntu Maverick, You'll love Mint 10. The new silver/gray theme and the preconfigured codecs make Mint 10 IMO look and feel much more serious then Ubuntu 10.10.


Cheers
 
Hi,

After many years away from Unix / Linux environments (used freebsd and sun Sparc's on a basic level) I installed Ubuntu yesterday. Boy have things changed this really is a modern OS. Anyway before i go too far down this path and since my knowledge of such things is gone I have a couple of questions.

1. I installed Ubuntu 10.10 it looks like the hard realtime kernel for this is no longer supported so will I have to run an older distro and if so which one?

2. I Run a cut down streamlined xp os atm with vst plugins for active crossover duties on 3 way open baffles using a prodigy 7.1 hifi and asio. What do you suggest for active crossover duties that will mimic or be on a par with fabfilter pro?

before i start tweaking i'd like to make sure that capable software is available.
Prior to ubuntu, I ran HP-UX, RISCos and IRIX systems. How's that for a _long_ time ago!

I use Liquorix kernels on Ubuntu 10.10, which include pre-emption and is currently testing out the hot new group priority patches which Linus Torvalds is saying good things about. They are great for interactive (gui) responsiveness, and so far work well on a oldish desktop that runs squeezebox server and brutefir eq. It's worth checking out Liquorix.
 
Hi there.

BFS-kernel:
Honestly I never tried it. It seems that Ingo Molnar integrated parts of that BFS logic in his newer schedulers.

Ubuntu/Mint rt-kernel:
The maintainer left the team. ( However it seems that the ppa for "Natty" is back. Currently available is a 2.6.36 kernel with preemption enabled or a 2.6.33 rt-kernel:

Install them according to here: Maverick rt-kernels.

Just click on the links at above site, download the packages and install them with "sudo dpkg -i xxx" in top to bottom order.

SB Touch:
You'll find the updates on my page soon. I've been quite busy during the last two weeks.


Multichannel&VST:
Check out Ubuntu Studio Wiki's and Jack. USTudio Prep

Mint vs. Ubuntu:
If you like Ubuntu Maverick, You'll love Mint 10. The new silver/gray theme and the preconfigured codecs make Mint 10 IMO look and feel much more serious then Ubuntu 10.10.


Cheers

Thanks! Ill give the rt package a go tonight if i have time.
 
Update

Hi Folks,

Thanks for the help I'm all up and running - to an extent.

Running jack with 32 frames latency 0.7ms and Aqualung for the tunes.

Realtime Ubuntu Studio kernel.

Listening to fatboy slim on fullrangers.

Now for the difficult part.

I'm currently running this using one of my sound cards - Audigy zs

I need help with a couple of things.

1. I Don't seem to be able to configure more than 2 channel output with jack
2. I need to configure 8 channels with their own crossovers and parametric equalizers.
3. I can use aqualung and can play about with certain filters etc but i'm having difficulty setting this up correctly.

Can anyone advise which patch program should i be using and how do i configure it to pass stereo from my music player to eight outputs via crossover filters and equalisation?

Thanks.
 
Hi,

Soundtech,

This topic has already 1736 posts, even with speed reading I don't think I can get through all of them in a short time.

I would appreciate if you can point/direct me to some posts where you have summarized the actions I have to take to set up a Linux music server, e.g. scale down Linux OS, what music player/server software to use, etc. BTW I have been to the Wiki website on Linux Audio, it is incomplete.

I have a Intel Duo core 1.2Ghz laptop and currently running XP Pro, I want make a dual boot with the Linux Ubuntu V10 OS. Will this do and should I run Ubuntu as a standalone OS?

Any help would be appreciated. Thanks.
 
Hi folks.

I've written a little ramplayer script (1st drop) based on ecasound.
I wanted a nice sounding app that I start from the filemanager
by clicking on an audiofile.

There is nothing like a real good ramplayer resp. realtime-kernel
optimized audioplayer under Linux.
OSX and Windows is pretty much ahead of Linux on this aspect.

The script will also decode flacs prior to playback.

You should try the script if you run after best sound.
You IMO won't find a better sounding app on a Linux system.
Once installed, it'll be started as soon as you click on an audio-file (flac/wav) in the file manager or by running it from commandline.
For now it just plays one file at a time. To stop it you just
double-click again on any audio-file.

The basic setting assumes your first soundcard ( index =0) as the interface of choice. You might have to configure that.

Note: I don't have any volume control implemented. You'll run the
stream at 0db into your soundcard. Don't blow your speakers!
You might choose a dmix outputdevice ( see man ecasound).
But that one won't sound pretty good.


Instructions how to install and run it you'll find inside the script:

Code:
#!/bin/bash
#
# RAM player based on ecasound written by Klaus Schulz aka soundcheck 
# Rev 1: 12/15/2010
#
# This ram-player script based on ecasound can be used to play single tracks (flac or wav) from filemanager by double click on a file or by commandline.
# Flacs are decoded prior to playback. And all files are played back 100% from RAM.
#
# Installation: 
# 1. save this code/file to a file called "eca". copy eca to /usr/local/bin: sudo cp eca /usr/local/bin
# 2. install apps: sudo apt-get install ecasound flac sox
# 3. make "eca" executable: sudo chmod 777 /usr/local/bin/eca
# 4. Assign script to filetype: In filemanager do right-click on a flac and later wav. 
#    Select Properties/Add/Use a Custom command ; Now you enter: eca %1
# 5. Install and boot a realtime kernel for best performance
#  
# Operate: Open filemanager; double-click will run the file, another double-click will stop it.
#          From commandline you enter: eca <your-audiofile.flac/wav> 
#          Different samplerates are recognized by ecasound. I also introduced an automatic sample rate detection 
#          and setup for an RME HDSP sound card
#          Hint: For commandline usage: try to boot into recovery mode for best performance
#
# Warning: There is no volume control implemented. By routing the data stream without volume control 
#          to the alsahw output you might damage your speakers. You won't be able to use the Gnome 
#          mixer. Only HW related mixers will work.
#
#


######################################################################################################
DEV=0          # change to device id of your soundcard 
SUBDEV=0       # change to sub device id of your soundcard
DEVTYPE=alsahw # change "alsahw" to "alsaplugin" in case of pcm problems see man ecasound
######################################################################################################


DEVICE=$DEVTYPE,$DEV,$SUBDEV  # change alsahw to "alsaplugin" in case of pcm problems 



SETHDSPSAMPLERATE () {  ## this function sets the samplerate on the RME HDSP 9632 mixer automatically

ACTRATE=$( soxi -r "$1" )

case "$ACTRATE" in
    '44100') 
       RATEID=2;
       ;;
    '48000')
       RATEID=3;
       ;;
    '88200') 
       RATEID=5;
       ;;
    '96000') 
       RATEID=6;
       ;;
esac
amixer -q -c $DEV cset numid=11 $RATEID # numid =11 is refering to sample clock source parameter
}


if [ "$( ps -edf | grep ecasound | grep -v grep )" != "" ] ; then
killall ecasound
rm /dev/shm/*.wav /dev/shm/*.flac 2>/dev/null
exit 0
fi 

rm /dev/shm/*.wav /dev/shm/*.flac 2>/dev/null
cp "$1" /dev/shm
cd /dev/shm

FILE="$( basename "$1" )"

if [ "$( ls "$FILE" | grep flac )" != "" ] ; then 
     flac -s -d --delete-input-file "$FILE"
     NEWFILE="$( basename "$FILE" .flac ).wav"
else NEWFILE="$FILE"
fi

if [ "$( aplay -l | grep HDSP | grep -v grep )" != "" ] ; then  
  SETHDSPSAMPLERATE  "/dev/shm/$NEWFILE" 
fi

ecasound -q -b:rt -r:90 -i:/dev/shm/"$NEWFILE" -o:$DEVICE &

exit 0


Note: Users who own an HDSP 9632 card will enjoy my
automatic samplerate configuration of the HDSP mixer. ;)






Enjoy
 
Hi,

Any chance of you answering my thread 1737?

Install the script I posted yesterday.

That'll summarize to a great extent what's been discussed
over here regarding SQ improvments.

My ambition level to maintain a Wiki about Linux Audio went down the drain at the time I introduced a Squeezebox Touch as the main audio transport to my system.

Good luck.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.