• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Raspberry pi3b usb output

Lucaxxaa

Member
Joined
Apr 6, 2019
Messages
19
Likes
3
Hello. I am newbie . I use now raspberry pi3b with hat dac IQaudio dac plus, connected to my home amplifier.

I like to improve dac, and try example or Khadas tone board or topping d30 or d50.

My question is, about usb output of raspberry.
It have same problem about clock and jittler , for sound quality ?

What is reason to use example , allo digione on raspberry that use spdif out ?

Thank
 

somebodyelse

Major Contributor
Joined
Dec 5, 2018
Messages
3,682
Likes
2,962
If you want to use BruteFIR on the Pi you'll almost certainly need to use i2s (hat dac or spdif) to avoid random glitches a bit like surface noise on vinyl. For most other uses it should be fine on USB, although some have reported similar glitches when streaming very high bit rate content. That content may be too high a bit rate for spdif anyway. IIRC the khadas and topping dacs have well engineered usb and spdif inputs as shown in their measurements, so clock and jitter aren't a problem. There's more detail about the possible cause of the glitches in this thread.
 

tw99

Senior Member
Joined
Mar 13, 2019
Messages
467
Likes
1,069
Location
West Berkshire, UK
I use an original RPI as a USB source for 16 bit material into an external DAC, and it works perfectly. Pi3 should be as good....
 

zalive

Active Member
Joined
Sep 4, 2018
Messages
263
Likes
38
Quality of the USB audio (from the listening/subjective perspective) oscillates from computer to computer. RPi 3 is not the worst USB streamer I have come across...but it's pretty far from being the best one as well. USB through its +5, mass and signal lines induces noise in DAC. Jitter itself should not be the big issue - modern USB protocols used are asynchronous anyway, and everything will be reclocked at DAC using its own clock (or converter).

For the stock RPi solution, Pi2Design 502DAC offers some real quality through its coax output (though I warmly recommend using real BNC instead of BNC/RCA adapter, as the adapter they ship with the card is not of highest quality - it's actually capable of degrading the sound. So it is an investment but IME it pays off in terms of sound. It will top Allo Digione (but don't ask me for their Signature as I never listened to it - all I know is that it's too expensive for my taste).
 
Last edited:

zalive

Active Member
Joined
Sep 4, 2018
Messages
263
Likes
38
There's no valid reason for facepalming. We do listen through our ears, not through measurement equipment.
Outside of this I'd like to see noise measurements of RPi USB output (of all three lines mentioned). Are there any?
 

somebodyelse

Major Contributor
Joined
Dec 5, 2018
Messages
3,682
Likes
2,962
The facepalm is because we know the mind plays tricks with our subjective interpretation. Can you repeatedly tell the difference between streamers when you don't know which streamer you're listening to?

See the link in post #3 above for a comparison of the same DAC on 3 different streaming devices (RPi3, odroid and Surface 3) - the measurements are all but indistinguishable, and probably within the margin of error on the measurements. Given the differences between different ports on the same machine shown in this post it's likely there are significant differences in noise between these streamers, but the DAC output doesn't change. Look through Archimago's other blog posts and you'll find a wealth of tests on RPi based systems among others, testing with different popwer supplies, usb cables and so on, and not finding differences. The one notable difference was increased IMD when using SPDIF instead of USB, but that was a known characteristic of the DAC used for the test. Amir's (and other) tests on USB conditioners have shown no improvement, and sometimes measurably worse performance, unless used with a DAC that has a known poor USB input.

A properly designed DAC will be immune to normal (ie quite high) levels of noise on USB because that's what the designers expect and design for. Thanks mainly to phones cramming lots of noisy stuff into tiny spaces we now have regulators and opamps with massively improved power supply rejection, especially at high frequencies. That's not to say that all DACs have properly designed USB inputs, but that's a fault in the DAC not the streamer.

There are some specific cases where the RPi fails to deliver the data in time over USB as noted above, and with the RPi3 B+ you should limit the wired ethernet to 100M for similar reasons.
 

BillG

Major Contributor
Joined
Sep 12, 2018
Messages
1,699
Likes
2,266
Location
Auckland, New Zealand
There's no valid reason for facepalming. We do listen through our ears, not through measurement equipment.
Outside of this I'd like to see noise measurements of RPi USB output (of all three lines mentioned). Are there any?

As someone who works professionally in Computer Science, I've ever reason to face-palm when dealing with subjective experience: That lump of mostly fat and water between our ears is easily manipulated into believing its own delusions, and those of others who are skilled at manipulating us for profit.
 

zalive

Active Member
Joined
Sep 4, 2018
Messages
263
Likes
38
As someone who works professionally in Computer Science, I've ever reason to face-palm when dealing with subjective experience: That lump of mostly fat and water between our ears is easily manipulated into believing its own delusions, and those of others who are skilled at manipulating us for profit.

Listening is solely subjective experience. There's nothing objective when it comes to creating impressions by listening.
Enjoyment comes from subjective impression. If it comes from watching graphs than it has little with listening to sound and music ;)
 

BillG

Major Contributor
Joined
Sep 12, 2018
Messages
1,699
Likes
2,266
Location
Auckland, New Zealand
Listening is solely subjective experience. There's nothing objective when it comes to creating impressions by listening.
Enjoyment comes from subjective impression. If it comes from watching graphs than it has little with listening to sound and music ;)

It's only subjective if one doesn't understand the sciences behind it, which is what a number here attempt to due being scientists, engineers themselves - I'm one myself in a related field.

You'll never find me agreeing with a solely subjective take on audio, even though I am a melophile, and have been since my teens. I'll measure first, or review the independent performance analyses of those I trust, and enjoy afterwards - Firm in the knowledge that my mind/brain hasn't played tricks on me, and caused me to spend money unnecessarily... :facepalm:
 

zalive

Active Member
Joined
Sep 4, 2018
Messages
263
Likes
38
It's only subjective if one doesn't understand the sciences behind it, which is what a number here attempt to due being scientists, engineers themselves - I'm one myself in a related field.

You'll never find me agreeing with a solely subjective take on audio, even though I am a melophile, and have been since my teens. I'll measure first, or review the independent performance analyses of those I trust, and enjoy afterwards - Firm in the knowledge that my mind/brain hasn't played tricks on me, and caused me to spend money unnecessarily... :facepalm:

Listening impression can only and solely be subjective. You don't command yourself to enjoy the music, it cannot be done by your will or by your decision. Either you like it and enjoy it, or not. And if you do, the level of your enjoyment is again of subjective nature. Beauty/enjoyment criteria are subjective.

I don't expect anyone to agree on solely subjective take on audio, it's personal approach and it should be such. However I expect everyone to agree on those few sentences above (if sound logic is being used). It's about understanding the subjective nature of listening to the music and sound, and the subjective nature of finding personal enjoyment in it :)

As for cheap and good sounding hifi equipment according to my subjective listening evaluation, I definitely find it. It's just that I'd like to be able to buy anything with great objective measurements and cheap besides it - it saves me money and I don't care much about status, brands, even the visual aesthetics of components are not crucial to me. Only how it sounds to me. I'm sorry, but I doesn't happen always, regardless of good measurements.

Recently I bought a cheap (<70 USD), no name DAC circuit board based on NOS AD1865 chip and made myself a nice DAC according to designer's recommendations for transformers power, with surprisingly good results when it comes to listening experience. AFAIK, NOS multibit DAC chip designs don't have perfect measures in general, mostly because of lack of oversampling and lack of filtering. But to my ears it sounds beautifully. Total cost was perhaps 120 USD with the casing, so it's definitely not about the money or the brand - there's no money, there's no brand.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
There's no valid reason for facepalming. We do listen through our ears, not through measurement equipment.
Outside of this I'd like to see noise measurements of RPi USB output (of all three lines mentioned). Are there any?

Except yours ears dont tell you the truth. Measurements tell us what is actually happening as opposed to what you think, or would like to be happening, or are biased/marketed into believing..

This post/ thread shows the difference between in DAC output between multiple computers and SBCs. None whatsoever. I will caveat that by saying it is possible that a deficient dac may not isolate or decouple the power supply adequately, but this is dac design issue not a computer issue.

https://www.audiosciencereview.com/...end-points-are-they-any-good.5707/post-127194
 

zalive

Active Member
Joined
Sep 4, 2018
Messages
263
Likes
38
You're missing the point. Subjective listening is very easily influenced by confirmation bias.

If you really want to pick something using subjective testing, then make sure you use a blind testing regime.

Of course. But I'm not missing the point.

The point of listening to the music is enjoying in listening to the music.
The point of listening to a HQ sound (music) is enjoying further more in a music being played with a HQ sound.
How I got to this personal enjoyment, if I got there, is completely irrelevant. I'm enjoying it.
What the objective measurements show is also not relevant to my enjoyment.
Should I enjoy more if I see the graphs and measurements are nice? Or should I enjoy less if I see the graphs and measurements are not as nice?
Now if this happens, that's a bias as well ;) seeng those graphs should not affect how I hear it and how I enjoy it, but it can happen...right?

Everyone has to set his reference point in search for that perfect sound. If one wishes to go with graphs and measurements feel free, but to me, it's my ears. Everyone should be free to choose the reference, well...this one is mine, personal. I can't advocate it since it's subjective. What I how I hear and what and hwo you hear is not necessarily the same...right? So there's no comparison. No result.

Only personal enjoyment. And no matter how we approach it, in the end you and me and everyone else will listen through our ears.

Why so? Because what I see and what I hear don't align perfectly. And I'm pretty much sceptic whether objective guys measure it all and if they perfectly described the human hearing mechanism with those set of measurements which is usually being done. Otherwise, I think this is not the all there is guys.

It would be great if seing measurements would suffice to hypnotize me to a perfect enjoyment through listening experience, but it just doesn't happen.
 

zalive

Active Member
Joined
Sep 4, 2018
Messages
263
Likes
38
Except yours ears dont tell you the truth. Measurements tell us what is actually happening as opposed to what you think, or would like to be happening, or are biased into believing..

This post/ thread shows the difference between in DAC output between multiple computers and SBCs. None whatsoever. I will caveat that by saying it is possible that a deficient dac may not isolate or decouple the power supply adequately, but this is dac design issue not a computer issue.

https://www.audiosciencereview.com/...end-points-are-they-any-good.5707/post-127194

Listening is not about the truth, it's about subjective personal impression.

Subjective personal criteria within my brain cells will tell me 'oh, that vocal sounds almost like in my living room', od 'oh, that saxophone sounds almost like live experience'. It's not necessarily the truth. But who cares?

Thanks for the link.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
Of course. But I'm not missing the point.

The point of listening to the music is enjoying in listening to the music.
The point of listening to a HQ sound (music) is enjoying further more in a music being played with a HQ sound.
How I got to this personal enjoyment, if I got there, is completely irrelevant. I'm enjoying it.
What the objective measurements show is also not relevant to my enjoyment.
Should I enjoy more if I see the graphs and measurements are nice? Or should I enjoy less if I see the graphs and measurements are not as nice?
Now if this happens, that's a bias as well ;) seeng those graphs should not affect how I hear it and how I enjoy it, but it can happen...right?

Everyone has to set his reference point in search for that perfect sound. If one wishes to go with graphs and measurements feel free, but to me, it's my ears. Everyone should be free to choose the reference, well...this one is mine, personal. I can't advocate it since it's subjective. What I how I hear and what and hwo you hear is not necessarily the same...right? So there's no comparison. No result.

Only personal enjoyment. And no matter how we approach it, in the end you and me and everyone else will listen through our ears.

Why so? Because what I see and what I hear don't align perfectly. And I'm pretty much sceptic whether objective guys measure it all and if they perfectly described the human hearing mechanism with those set of measurements which is usually being done. Otherwise, I think this is not the all there is guys.

It would be great if seing measurements would suffice to hypnotize me to a perfect enjoyment through listening experience, but it just doesn't happen.

If under controlled blind conditions I would wager that you would not recognise "what you enjoy" about one dac over another you supposedly didnt.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
Listening is not about the truth, it's about subjective personal impression.

Subjective personal criteria within my brain cells will tell me 'oh, that vocal sounds almost like in my living room', od 'oh, that saxophone sounds almost like live experience'. It's not necessarily the truth. But who cares?

Thanks for the link.
As I said, try it under controlled blind conditions and your view will change. Your (anybodies) sighted and uncontrolled subjective opinions/impressions are unreliable and inconsistent.
 
Last edited:

BillG

Major Contributor
Joined
Sep 12, 2018
Messages
1,699
Likes
2,266
Location
Auckland, New Zealand
I purchased my most recent system based solely on measurements alone - they told me everything I needed to know. And, I've yet to be disappointed by it, nor do I wish to upgrade it soon due to being attracted to cosmetically nicer gear in the same price range... :)
 
Top Bottom