Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Is there any way to objectively measure things like soundstage? (Read 2091 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Is there any way to objectively measure things like soundstage?

So, I'm in the middle of setting up my Raspberry Pi streaming server.

Doing homework on DACs, and there are a LOT of claims about what constitutes good audio on a Pi.

Reading discussions online, many people claim improved sound quality by adding a Pi HAT that does TOSLink vs using the USB ports when connecting to a DAC.  Many claim that TOSLink vs USB to the SAME amp, TOSLink produces a better result.

And they almost always claim that there is:
1. Better soundstage
2. Better instrument separation
3.  Better high end
4. "More definition" to the music

So, being a PITA about these things, I ordered a SPDIF HAT for my Pi to do an actual comparison.

So, some questions.  Can things like soundstage, instrument separation, and "more definition" be objectively measured in any way?

And, my plan was to go USB to my DAC and then DAC to my PC and capture the audio as a FLAC.  The go TOSLInk to my DAC, and then DAC to my PC and capture the audio.  Once I have that, I can ABX them.  Or even better, and I can use Audacity to get the delta between them and see if there is a difference at all between them, and what exactly the difference will be.

Re: Is there any way to objectively measure things like soundstage?

Reply #1 –
Quote from: apastuszak on
Or even better, and I can use Audacity to get the delta between them and see if there is a difference at all between them, and what exactly the difference will be.

I’m afraid this will be incredibly hard.
Simply because no clock without drift.
I expect (and recommend) to do a null test first, say 2 times capture exactly  the same stream without any change in any setting or hardware. Simply exactly the same thing twice.
If this test already yield differences, it is probably of no use to compare USB and Toslink this way.
Maybe the software mentioned here is of use: https://www.thewelltemperedcomputer.com/SW/AudioTools/NullTest.htm

TheWellTemperedComputer.com

Re: Is there any way to objectively measure things like soundstage?

Reply #2 –
Pardon my ignorance, but why are double blind tests not enough?
Quis custodiet ipsos custodes?  ;~)

Re: Is there any way to objectively measure things like soundstage?

Reply #3 –
Quote
So, some questions.  Can things like soundstage, instrument separation, and "more definition" be objectively measured in any way?
No.  Pretty-much "audiophile nonsense" with no scientific or engineering definition...   Take a look at Audiophoolery.   Different people might define these things differently or experience them differently.  Some people get a good soundstage illusion (and it is an illusion) with headphones but I never perceive a "soundstage" no matter what headphones I'm using. 

With speakers I get a better (I'll call it a "stereo image") but the center is not as defined as the hard-left & hard-right unless I'm listening to a surround system with a separate center channel and center speaker.   Stereo separation is perfect in the digital domain any left-right leakage in the analog electronics is way better than human perception (unless something is seriously defective).    "Soundstage" is more related to room acoustics, speaker placement, the recording, and psychology/perception.
 
ABX is the way to go.   A null test can prove there is no difference, but if the files don't null that doesn't prove there's a difference in sound.  The most obvious "failure" is a few-millisecond time delay which makes zero difference in sound but it creates a loud difference-file in a null test.

If you hear any difference it will probably be noise and you may not need an ABX test to hear the difference.  Most-likely frequency response and distortion are better than human hearing.  But of course an ABX test is required if you report your results here.

Re: Is there any way to objectively measure things like soundstage?

Reply #4 –
Common sense tells me the ABX test should show that I am guessing.  But I am curious.  One person said the difference between the USB and TOSlink was "immediate and apparent."  I somehow doubt that.  But I bought the hardware.  We'll see how it goes.

Almost every Pi review for audio says that the power on the Pi is "dirty" or "noisy" and this affects the sound quality of the PI.  But nobody that I have found ever points to any proof the Pi's power is dirty or noisy.  It's one of those audiophile myths that seems to perpetuate, because someone said it was so years ago.

Re: Is there any way to objectively measure things like soundstage?

Reply #5 –
Quote from: apastuszak on
Common sense tells me the ABX test should show that I am guessing.  But I am curious.  One person said the difference between the USB and TOSlink was "immediate and apparent."  I somehow doubt that.  But I bought the hardware.  We'll see how it goes.

Almost every Pi review for audio says that the power on the Pi is "dirty" or "noisy" and this affects the sound quality of the PI.  But nobody that I have found ever points to any proof the Pi's power is dirty or noisy.  It's one of those audiophile myths that seems to perpetuate, because someone said it was so years ago.


You may find this interesting.
http://archimago.blogspot.com/2018/12/measurements-raspberry-pi-3-b-as.html

Re: Is there any way to objectively measure things like soundstage?

Reply #6 –
Quote from: garym on
Quote from: apastuszak on
Common sense tells me the ABX test should show that I am guessing.  But I am curious.  One person said the difference between the USB and TOSlink was "immediate and apparent."  I somehow doubt that.  But I bought the hardware.  We'll see how it goes.

Almost every Pi review for audio says that the power on the Pi is "dirty" or "noisy" and this affects the sound quality of the PI.  But nobody that I have found ever points to any proof the Pi's power is dirty or noisy.  It's one of those audiophile myths that seems to perpetuate, because someone said it was so years ago.


You may find this interesting.
http://archimago.blogspot.com/2018/12/measurements-raspberry-pi-3-b-as.html
Thank you.  That was an interesting read.

Re: Is there any way to objectively measure things like soundstage?

Reply #7 –
Quote from: DVDdoug on
Quote
So, some questions.  Can things like soundstage, instrument separation, and "more definition" be objectively measured in any way?
No.  Pretty-much "audiophile nonsense" with no scientific or engineering definition...   Take a look at Audiophoolery.   Different people might define these things differently or experience them differently.  Some people get a good soundstage illusion (and it is an illusion) with headphones but I never perceive a "soundstage" no matter what headphones I'm using. 

With speakers I get a better (I'll call it a "stereo image") but the center is not as defined as the hard-left & hard-right unless I'm listening to a surround system with a separate center channel and center speaker.   Stereo separation is perfect in the digital domain any left-right leakage in the analog electronics is way better than human perception (unless something is seriously defective).    "Soundstage" is more related to room acoustics, speaker placement, the recording, and psychology/perception.
 
ABX is the way to go.   A null test can prove there is no difference, but if the files don't null that doesn't prove there's a difference in sound.  The most obvious "failure" is a few-millisecond time delay which makes zero difference in sound but it creates a loud difference-file in a null test.

If you hear any difference it will probably be noise and you may not need an ABX test to hear the difference.  Most-likely frequency response and distortion are better than human hearing.  But of course an ABX test is required if you report your results here.


That's a good point.  It's going to be really hard to make sure the two files aren't time shifted.  Even being off a little will cause an issue.

Re: Is there any way to objectively measure things like soundstage?

Reply #8 –
Quote from: apastuszak on
So, some questions.  Can things like soundstage, instrument separation, and "more definition" be objectively measured in any way?
Both the soundfield and the beholder can be measured. Not the same thing.

cheers,

AJ
Loudspeaker manufacturer