Computer Audio Asylum

Music servers and other computer based digital audio technologies.

Return to Computer Audio Asylum


Message Sort: Post Order or Asylum Reverse Threaded

Raspberry Pi 3 as USB audio streamer - some audiophile settings to try...

172.98.67.74

Posted on January 7, 2017 at 11:41:11
Archimago
Audiophile

Posts: 755
Joined: January 18, 2002
Here you go folks.

Raspberry Pi 3 works great. Plus some configuration clock tweaks to try :-).

-------
Archimago's Musings: A 'more objective' audiophile blog.

 

Hide full thread outline!
    ...
RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 7, 2017 at 12:09:21
Cut-Throat
Audiophile

Posts: 11275
Location: Minneapolis - St.Paul Area
Joined: September 2, 2000
A very thorough report! .... I was told to stay away from USB on any device. So, on my Raspberry Pi 3s, I use the HifiBerry Digi+ Plus Board and run either Coax or Optical into my DAC. I am very pleased with the sound I am getting. A question though. What do you think is the best output of a Pi ... USB, Coax or Optical ?

Cut-Throat




 

RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 9, 2017 at 06:12:21
soundchekk
Audiophile

Posts: 1859
Joined: July 11, 2007
Hi.

Welcome to the PI club.

Most recent PcP is 3.10! Not 3.02.
Numerous audio related patches went into the optional audio kernel of 3.10.

Most recent Moode also comes with that specific audio kernel (and also squeezelite beside MPD)

Running a 44.1 flac at 2%CPU load doesn't reflect the most efficient
setup.

I have it down to 0.4-0.7%

Another hint:

Try a buffer setting of e.g.

-b 20000:600000

That'll give you a 600MB playback buffer. There'll be no
network streaming during playback.


And PcP also offers to turn off HDMI. Do that. It'll save you even more
power.

Enjoy.







 

RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 9, 2017 at 07:12:03
Cut-Throat
Audiophile

Posts: 11275
Location: Minneapolis - St.Paul Area
Joined: September 2, 2000
Have you changed your opinion of USB output? --- I thought I remember you did not like USB out of a Pi at all?

Cut-Throat




 

RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 9, 2017 at 07:22:49
soundchekk
Audiophile

Posts: 1859
Joined: July 11, 2007
Why?

I still don't use USB DACs anymore.

Most of the tweaks apply to both approaches though.

 

"And PcP also offers to turn off HDMI.", posted on January 9, 2017 at 07:47:41
Cut-Throat
Audiophile

Posts: 11275
Location: Minneapolis - St.Paul Area
Joined: September 2, 2000
I looked for this and was unable to find it. Which screen is this option on?

Cut-Throat




 

Two parts of your article seem to contradict each other., posted on January 9, 2017 at 12:03:01
Posts: 55
Joined: December 29, 2011
Either your measurements represent what you insist they represent (and clearly, you're pretty serious about that) - and therefore those tweaks are pure nonsense

OR

Those tweaks actually cause audible differences - and therefore your measurements are worth less than squat, because you have no idea what you're measuring, and what is the correlation between measured parameters and audible differences.

Which one is it?

 

Where is the contradiction?, posted on January 9, 2017 at 20:41:34
Ugly
Audiophile

Posts: 2542
Location: Des Moines, WA
Joined: August 22, 2006
I interpret the test results to indicate this test setup can't resolve the differences if there are any.... other than maybe the jitter result. I didn't see where the op said otherwise.

Though, I would note there does seem to be a small hint of a slight emi problem in this system. The components used in the test would be capable of slightly better resolution if those problems were addressed.

Of course, even then it still may not be enough to resolve any performance differences these configuration changes may or may not be having.

If these configuration changes indeed do happen to lie below the systems capability to resolve, then why should they be expected to be audible???

 

RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 9, 2017 at 21:56:41
Ugly
Audiophile

Posts: 2542
Location: Des Moines, WA
Joined: August 22, 2006
I've always speculated that systems with the biggest emi problems benefit the most from these tweaks aimed at reducing emi. Those systems with very small emi problems will achieve relatively fewer appreciable benefits I'd guess.

You have apparently been able to achieve fairly heroicly low levels of emi in your various loopback configurations considering at least part of the analog chain is single ended circuitry, ie the forte aaand you are apparently hooked to a pc somehow via usb in order to run RMAA...

Your capture systems decent apparent emi performance may be why you couldn't measure or hear much of a difference from these kinds of tweaks if that is indeed the case.

In other words, I'm wondering if your system may be too high performance at rejecting emi to really show why these tweaks may be so valuable in some systems.

You could possibly test my theory by trying to replicate the emi vulnerability extremes and rerun so called emi tweak tests.

1. You'd need an "ideal" test system. Perhaps your existing system is already good enough for this. Ensure optimal emi hardening via proper grounding, etc. I'd suggest you could possibly be able to achieve even better emi results than you are getting now with a fully balanced analog signal path in your loopback.

2. create a "worst" case system to test. Easy. Get some single ended gear and throw in some massive ground loops.

 

RE: "And PcP also offers to turn off HDMI.", posted on January 10, 2017 at 02:07:32
soundchekk
Audiophile

Posts: 1859
Joined: July 11, 2007
It's available in the admin or beta mode under tweaks if I recall it correctly.

I don't use PCP anymore. I can't be of much help here.


There's also another way of accomplishing hdmi = off afaik.

Add to /boot/config.txt

"hdmi_blanking=2"

Good luck.

 

RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 11, 2017 at 00:54:00
soundchekk
Audiophile

Posts: 1859
Joined: July 11, 2007
His DAC filters most of the stuff above the audio band.
His ADC filters most of the stuff above the audio band.


His conclusion only means, that no matter
what source he's using his DAC/ADC combo measures
the same way (on his questionable tools) in the audio band.

This has nothing to do with the noise originated from different
sources. Nor it can lead to any conclusions related to the source.

His measurement methods are pretty much ridiculous.
And he keeps spreading his crap (doing it for years).

As a starter - I recommend to read this:

http://nwavguy.blogspot.de/2011/02/rightmark-audio-analyzer-rmaa.html


Enjoy.

 

RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 11, 2017 at 06:59:22
Ugly
Audiophile

Posts: 2542
Location: Des Moines, WA
Joined: August 22, 2006
His Dac and ADC presumably respond in the audio band. We as audio enthusiasts presumably care most about what is within the audible ranges.

If the theory us that these system changes are responsible for making changes in the audible noise profile then presumably having more ability to resolve noise within the audible band will help the test results be more useful regardless of whether or not his results are calibrated to some standard.

To say that the RMAA SNR testing results in this case has absolutely nothing to do with noise would seem a bit short sighted by my estimate.

The high res RMAA plots show evidence of what is almost certainly unintended emi creeping into the audible signal being sampled. Getting rid of it could make it easier to identify low level changes in noise profile which are within this systems ability to respond to it. Having the other noise sources gone from the picture could make it easier to identify low level changes if thy happen to cause system response that is made visible by the test.


Though the results certainly mean the most within the context of this test setup and can't easily be directly correlated to other systems, the relative differences could certainly be meaningful if they actually showed appreciable difference.

Meaningful in that if changes are detected by the test, other users could more reasonably wonder if similar changes might have some effect on their own systems.

 

Yes Thanks! - Beta Tab)Tweaks)HDMI Power .......nt, posted on January 11, 2017 at 07:09:32
Cut-Throat
Audiophile

Posts: 11275
Location: Minneapolis - St.Paul Area
Joined: September 2, 2000
nt

Cut-Throat




 

RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 11, 2017 at 08:18:57
soundchekk
Audiophile

Posts: 1859
Joined: July 11, 2007
Look.

Since years Archimago tries to explain "Computer Audio" with his
way of measurements.

He is pretty immune against any feedback/criticism.

And he seriously believes that he can prove everything, where
professionals with measurement equipment worth ten thousands of $ fail.


His measurements are OK for basic checkups. And that's about it.

 

And, hopefully, his readers are immune to his conclusions, too., posted on January 11, 2017 at 11:38:40
Posts: 55
Joined: December 29, 2011
Those who are not, have only themselves to blame.

His measurements are suitable, to determine whether a piece of equipment is broken, or grossly misrepresented by manufacturer - that's it. NOTHING beyond that.

Do I personally think his work is in any way useful? No, I do not - but hey, it's as good of a hobby as anything else out there.

 

RE: Raspberry Pi 3 as USB audio streamer - some audiophile settings to try..., posted on January 11, 2017 at 22:06:28
Ugly
Audiophile

Posts: 2542
Location: Des Moines, WA
Joined: August 22, 2006
"And he seriously believes that he can prove everything,"

It would be a shame to try and read too much into any results, maybe I missed where the OP did that in this case.

"where professionals with measurement equipment worth ten thousands of $ fail."

If these tests are already being performed by more qualified individuals on more accurate gear and the results are being submitted somewhere into the public domain for public discussion similar to how the OP is doing these articles I would sure love to learn where this is going on so I could see it for myself.

My impression has been the guys with the good test equipment are either working for manufacturers and therefore paid to paint their gear in the best light and so therefore a questionable source of reliable comparisons OR work for the media and always too afraid of biting the hand that feeds them to truly pit one manufacturers gear against anothers in truly scientific shootout form.

 

Page processed in 0.029 seconds.