or Connect
Head-Fi.org › Forums › Equipment Forums › Computer Audio › Bit Perfect Audio from Linux
New Posts  All Forums:Forum Nav:

Bit Perfect Audio from Linux - Page 16

post #226 of 438

Jack is not bit perfect from my understanding ?

I could get why jack would benefit, especially for recording instruments.  

post #227 of 438
Jack is.... Complicated. It can be bit perfect or not depending on your application chain. But yes rt kernels don't really help an already asynchronous USB connection.
post #228 of 438

Complicated is definitely the right word for Jack. It's always been way too much of a pain for me to try to implement properly on any chain I've had. I'm perfectly happy with Daphile as a server and a gmusicbroswer ALSA chain for listening when I need to multitask on the computer.

post #229 of 438

Well maybe I'm talking about non-exclusive mode of WASAPI and Shared mode is the name of when more than one audio stream is trying to access the audio output.  So I was lead to believe.  The person also use a DTS or AC/3 audio stream to prove that when it has to share with another applications active audio stream the decoder drops out during the two mixed audio streams.


post #230 of 438
What is that in reply to?
post #231 of 438

First, I would like to thank everyone on here, and Head-Fi.  Your information is terrific, as is your dedication to excellence in audio.  

​I used to have a very high end system.  It was a Linn system with 9 different speaker boxes, each box having either 3, 5 or 7 speakers in each box.  Every speaker had its own amp, and each amp had a dynamic filter in it for each dedicated freq range.  Every wire also was tuned for a certain freq range.  So the 7 speaker towers up front had 7 dual wires going to it, driven by 7 amps with 7 dynamic filters built into the amps, with selected wires for those freqs.  Also had the Linn player for the cds, which played the SACD and DVDA formats, and it had a DAC built in for the 24/192k sound.  So it was pretty awesome.  Linn even flew over 2 engineers to look at it and study it.  They then built an identical system for the Queen back home.  


But I lost everything in the crash in 2008, including house, cars, everything.  Left with clothes on my back.  So I miss having good audio.


​Since then I have acquired a good Linux computer and I am running Ubuntu 12.10 with a Sound Blaster SB400 and Grado SR325 headphones.  Using your info I think I have achieved 24/192 output.  I am using DeadBeef for the program.  I chose the output to be set at SB Audigy 2 Value (400), ADC Capture/Standard PCM device Playback Direct Hardware without any conversions.  Also set the secret Rabbit Code to automatic samplerate and a target sample rate of 192,200, and quality algorithm to since_best_quality.  For the ADplug I set it to prefer Ken emu over Satoh, and then set the ALSA output plugin to no ALSA resampling and release device while stopped.  Preferred buffer size is 20000 and preferred period size is 1024.


I do not know what some of these do, but the 24/192 recordings I do have on my hard disk do sound pretty good.  


Can you comment on what I have done and if there is an improvement, or have I done it correctly?  Your kind advice is appreciated, and I have the people on here so knowledgable that I take your advice to heart. 


​Also, would your recommend another program, or is DeadBeef good enough to do this.  I have noticed that some others like Gmusicbrowser.  Would this be a better program, or is the GUI better in people's estimation.


One more question.  I am thinking about getting an Emotiva DAC XDA-2 (wish I could afford the Stealth DC-1)  that uses USB input for my headphone driver.  Do devices like this need special drivers, or is that handled by the Linux OS?  Will the bits be passed directly to the USB bus and then to the external DAC for conversion, or is there some processing done before they are passed?  It seems like the programs will pass directly to the USB bus if the preferences are right.  I have emailed Emotiva with these questions but have not received an answer back yet. If Linux recognizes one USB device, will it recognize all the others that are also DACS?  That is, is one DAC the same as all the other DACs to Linux?


And the last question, if I do the above, would it benefit me to use the output of the DAC to go into a tube headphone amplifier?  Would I gain anything?  


Thanks again for any help that you throw my way.


Best Regards to everyone.

post #232 of 438

Here is the reply I received from Emotiva.  Any comments to help me?


The XDA-2 uses UAC2 (USB Audio Class 2), which is supported by some Linux dists and not others.

Check out the attached PDF, and the following link.....


The DC-1 uses the C-Media CM6631A USB interface chip.
Our implementation is UAC2 ONLY.
Other well known devices that use the same chip are the Asus Xonar Essence “high end external sound 
card” and many of the DACs made by Schiit Audio.

I have one comment that Linux Mint is “fully UAC2 compatible”. 

Especially read the attached “unofficial” help
(Note: we didn’t try it here)
The XDA-2 (which he used) has the CM6631 chip, while the DC-1 has the CM6631A, 
but they are basically the same and use the same drivers, so it should work similarly for both

Unfortunately, I'm not especially well-versed in Linux itself, so I can't offer you much support beyond that...



post #233 of 438
Originally Posted by ssedlmayr View Post

Here is the reply I received from Emotiva.  Any comments to help me?

As I seen on alsa site there is issues with linux and this chip - http://comments.gmane.org/gmane.linux.alsa.user/36935.

May be you will be lucky to solve it..:)

post #234 of 438

Hey ssedlmayr, UAC2 support only depends on 4 things:


1. USB 2.0 hardware.


2. Kernel with support for it or the correct modules (all modern distro's)


3. Modern Alsa install.


4. The manufacturer to stick to known standards.

Number 4 is where companies fall down, luckily its not super hard to program a way around it in alsa if the hardware is well documented.



Here is a starters guide I wrote to getting bit perfect audio on Linux. You seem to know a bit, but this will help you with the software setup side.


post #235 of 438

Awesome thread, just one question though.

Isn't bitperfect output supposed to render volume controls in the player useless?


I've disabled dmix via deadbeef using the settings: "M Audio Audiophile 24/96, ICE1712 multi Front Speakers" also disabled ALSA resampling, enabled Release device when stoped.

44.1 tracks play as 44.1 also 96 tracks play as 96Hz etc.


format: S32_LE
subformat: STD
channels: 10
rate: 44100 (44100/1)
period_size: 1024
buffer_size: 6553


SuperEQ and Resampling were also removed via the DSP section but SuperEQ keeps reappearing, although EQ isn't enabled.


But the volume controls in deadbeef still function.. Is it possible to disable them?

post #236 of 438

This is exactly true LowLatency.


You cannot have digital volume control AND bitperfect playback (well you can, but it gets SUPER complicated and needs special hardware). You can usually stand to lose a few bits before you will hear an audible difference that isn't placebo, but it does happen. Most people say they can hear a "normalisation" like effect at about 88% or lower, this is from the sound science forum if you would like to have a read.


May i ask why you have dmix if you don't intend on using it? A simpler system is, well, a simpler system. Always set a music player to the direct name of your device, not to a soft name a mixer gives it if you want to use bitperfect playback. If you want to get dmix working as a sole volume control on your system i can help you set that up if you would like. Otherwise just chuck it and set deadbeef to your hardware address.

post #237 of 438

Thanks yay101, thats what I figured..


Do you have a link to that info on the sound science forum? Seems like an intresting read.


Regarding dmix I thought that it was enabled as default for soundcards which do not support hardware mixing for later revisions?


I also have it generally disabled under .asoundrc via:


 type hw
 card 0


Although that has no effect on the volume controls in deadbeef. 

Without the settings in .asoundrc I can get playback without volume controls in cmus, although setting it to "hw:0,0" to bypass dmix and have the correct Hz disables any audio.


I've tried the other options for output in deadbeef and M2496 front speakers is the only one that outputs sound at the correct Hz.

I'd just like to remove dmix completely from the chain and have deadbeef without any soft volume control or dsp.

Edited by LowLatency - 2/25/14 at 9:18am
post #238 of 438

There is a much longer read on the subject I will try to find for you, as it was more in depth as far as usage scenarios go.




This is the one i was thinking of: http://www.head-fi.org/t/671220/effective-number-of-bits-or-why-you-have-to-keep-software-at-full-volume-is-nonsense

I am honestly not entirely sure what is default when you don't have hardware mixing as i haven't had a machine without it for many years, perhaps someone else can answer that question although i will be looking into it for my own personal interests. It sounds like you have a deeper issue going on however, by the lack of audio when you try to access the hardware directly. What version and system are you running?

Have you tried simply uninstalling dmix?

Edited by yay101 - 2/26/14 at 2:54am
post #239 of 438

How well do linux Distros handle USB Dacs?

post #240 of 438

You just plug them in and they work. Not sure if that's well enough for you. :D

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › Bit Perfect Audio from Linux