PC music players

Status
Not open for further replies.
Now I got Linux Fedora with CCRMA low latency kernel and XMMS -> Jack -> ALSA up and running. Too bad BruteFIR does not include a nice GUI like the applications mentioned. How do I get Jack to identify BruteFIR as a possible I/O?
 
Ok, started playing around a bit with BruteFIR (using trial and error) and suddenly everything worked. XMMS - Jack - BruteFIR - Jack - ALSA. Now I need an output expansion board form RME digi 96/8.

By the way, is there an easy way to start everything up? Now I have to find Jack and XMMS in the start-menu-kind-of-thing, start up the jack server then open a terminal window and start BruteFIR before I am able to play anything.
 
JACK should be running before anything else and can be left always running.
Start it up on boot in rc.d.

As for the rest... easiest way is with a little bash script connecting the ports with "jack_connect".

Use "jack_lsp -c" to see the names of avaliable ports.
 
Progress

I've been busy with other stuff but have managed to improve the player as follows:

Tried Mark's suggestion of 2.6.4 and Con Koliva's patch but unfortunately 2.6.4 failed to probe my soundcard. I tried a few things including redoing the whole setup and the whole alsa setup and still no luck so then I decided to try the latest kernel upgrades so I downloaded 2.6.6, patched it to 2.6.7-rc2 and am using that now.

xmms hasn't been dropped once yet and I can even launch firefox browser, grip, games, terminals, etc without interrupting the player. P&C's (pops and clicks) are almost gone. I'm thinking a few more tweaks and it will graduate to the bulletproof awesome category. Just launched jackctl, pressed start, launched brutefir and launched xmms and the trio have been happily running uninterrupted for the last 5 hours.

Preamp arrived so next week it will be drc (I'm expecting miracles :cloud9: ).

I just finished listening to Emmylou Harris' Blue Kentucky Girl cd for the first time and it sounded sooooo good. Was never a big country fan, but probably 99% of humans will probably love this cd.

Will be out next few days ... hasta la vista!

-Robert

Oh almost forgot - DSP Geek, that link to the flash-disk pc was quite good. I may use it for the final implementation of my hardware player, but probably not for awhile since development is so fun! Oh - one more thing - I remember Santa Cruz fondly from racing sailboats there years ago before the whole town moved to Canada:joker: .
 
Another problem, I use xmms-jack for connection with jack. Every time I start up xmms it names the output ports something like "lib2jack_0_3224" where the four last numbers change every time I start it up. Therefore I must edit .brutefir_config every time and can not use jackconnect in a bashscript. This is a hassle, any ideas how to get around it?
 
Hi iffe. Good to see another swede here.

I'm also testing jack to see if I can do something useful with it. The number you see is probably some process number.

I found the command 'jack_lsp' which list all the currently active ports. I got this output when running it:
alsa_pcm:capture_1
alsa_pcm:capture_2
alsa_pcm😛layback_1
alsa_pcm😛layback_2
bio2jack_0_3441😱ut_0
bio2jack_0_3441😱ut_1

In this case 'bio2jack' correspond to the xmms output port. With some bash magic you should be able to get it

jack_lsp | grep bio2 | tail -n 1 | cut -f 3 -d '_' | cut -f 1 -d ':'

or for storing it in an environmental variable:
PNUMBER=`jack_lsp | grep bio2 | tail -n 1 | cut -f 3 -d '_' | cut -f 1 -d ':'`

--- Mikael R
 
Guys,

An incredibly interesting thread. I have also started to go over to a PC source for my music. It was only when I started archiving my discs using Exact Audio Copy (EAC) that I realised how bad some of my early CDs were! Some took ages to transcribe due to the amount of re-reading and error correction required. I know that a number of contributors on this thread require multi-output scenarios via (very) expensive sound cards. However, many people only require one output to feed into an external DAC. Buying a motherboard with in-built S/PDIF saves a hell of a lot of cash! If you are interested in noiseless PCs have a look at: http://www.quietpc.com I have supplied customers with a number of these products and all have been very satisfied.

Brian
 
Are those motherboards capable of sending bit-perfect information? For example Soundblaster soundcards always resample the information to 48kHz (i think). There should also be a problem that the motherboard digital outputs have large amounts of jitter. If you have a good DAC that is not a problem, but on the other hand you still have to choose between an expensive soundcard or an expensive DAC.

But of course, people who i.e have an reciever with digital input and a motherboard with digital output who would like to use their computer as source this would be a nice budget solution.
 
A good way to check if the digital out is bit perfect is to hook it up to a receiver which can decode Dolby Digital or DTS, play a .wav file encoded in either format and listen to the output. If there is any type of resampling done you will just hear white noise.

There is an external USB soundbox, Creative Labs MP3+ which can be used in digital only mode. It passes the 'DTS-test' and should be bit perfect. However, no guarantees are given regarding jitter or latency. The CL solution has a latency of approx. 500ms and varies depending on USB traffic. But it shouldn't be a problem unless you want to syncronize output from several devices.

I've used an on board digital out also, but I havn't checked if it is bit perfect yet. It sounded OK though but that was before I got my new speakers. 🙂

--- Mikael R
 
iffe said:
Are those motherboards capable of sending bit-perfect information? For example Soundblaster soundcards always resample the information to 48kHz (i think). There should also be a problem that the motherboard digital outputs have large amounts of jitter. If you have a good DAC that is not a problem, but on the other hand you still have to choose between an expensive soundcard or an expensive DAC.

Whoops, I naturally assumed that on this list everybody would use either a home built or modified ‘off the shelf’ external DAC. :xeye: I seem to remember that some time ago there was a thread on the ‘Asylum’ that suggested that re-sampling Soundcards should be avoided. The posters suggested that this activity caused bit errors to be introduced. I was also assuming that one would insert ELSO’s Asynchronous Reclocker :nod: between the Motherboard and the DAC. I am using a modified ART DI/O which was considerably cheaper than some of the high-end Soundcards mentioned.
 
tiroth said:

That's neither here nor there. If the chipset destroys the signal by resampling it, it can never be fixed downstream whether you are using a DIY DAC or a toaster as your source.
Well yeah, hence feeding the signal directly out of the Motherboard to the DAC without going through a Soundcard. Also, the DAC is a downstream component not a source. The title of this thread is: ‘PC music players’. Or am I missing something here....
 
Most on board digital outputs comes from a cheap sound card controller. These tend to follow the AC97 specs which more or less demand resampling to 48000Hz. My experience is that resampling is the most common case and if you want 44100Hz out, you better verify that by experiments. Even in that case, it might be corrupted since many circuits use digital attenuator before sending the signal out.

--- Mikael R
 
Testing 2.6.7 with patch ck1

Hi all,

Summer, family and friends have reduced progress considerably but yesterday I patched linux kernel 2.6.7-rc2 with the Con Kolivas patch set and tested my gui-oriented xmms-qjackctl-brutefir-alsa setup. After reducing priority of X, we almost have a well-behaved dekstop player. Pops and artifacts were almost gone and very insensitive to launching browsers, word processing, etc. I'm sticking with this version for awhile and will try to improve scheduling a bit.

Mark,

I can't get schedtool to work properly. How important is it?

DRC is getting closer. My microphone is working well although I can't figure out yet how to adjust the recording volume of the M-audio 410.

-Robert
 
Re: Testing 2.6.7 with patch ck1

Summer, family and friends have reduced progress considerably but yesterday I patched linux kernel 2.6.7-rc2 with the Con Kolivas patch set and tested my gui-oriented xmms-qjackctl-brutefir-alsa setup. After reducing priority of X, we almost have a well-behaved dekstop player. Pops and artifacts were almost gone and very insensitive to launching browsers, word processing, etc. I'm sticking with this version for awhile and will try to improve scheduling a bit.

Good to hear your still going 😉
Havnt heard from you in a while now.

If you havnt noticed, the full 2.6.7 kernel is out, and con has a new patch for it (staircase7).

I can't get schedtool to work properly. How important is it?

Very... more so than using staircase.
What problems are you having?
 
Schedtool

Ok, I've downloaded the right version of schedtool and it works quite well. From the first few minutes of play, it appears to solve all remaining pops/clicks. I'll let you know if that's not quite true in the end.

Do you think I need to replace my patched 2.6.6 - 2.6.7-rc2 with the 2.6.7 release kernel?

How are you setting up the staircase scheduling on Con Kolivas' patch?

-Robert
 
Status
Not open for further replies.