Noise as a test signal for audibility

[John Kenny]

I've read many time that pink or white noise are good test signals for audibility testing but never tried it in any serious way myself until inadvertently doing so recently.

Some favourite reference music for testing is Oscar Petersons' "We Get Requests" - all tracks are worth listening to & contain a nice variety of sounds from percussive drums, great double bass playing to some very light brushing on cymbals & drums not to mention the piano as an instrument is a great test of any system.

But the 3 track "My One & only love" has some very noticeable hiss on the lead in & in the very quiet parts. I presumed this was tape hiss as the recording dates back to 1964. In the WAV version I have of this track, the hiss is noticeable on all playback systems I've heard it on although the version on Youtube doesn't have this hiss

In a recent listening to some USB isolators there was a noticeable difference in the intrusiveness of this hiss - it was perceptually less intrusive with the USB isolator in-line.

I wonder how many others use noise signal to test for audible changes in a system?

 

[NorthSky]

If the noise was present in the original recording, then the noise is accurate. Any masking, noise removal will somehow alter the original music recording?
...Missing that "faithfulnes", no?

 

[John Kenny]

If the noise was present in the original recording, then the noise is accurate. Any masking, noise removal will somehow alter the original music recording?
...Missing that "faithfulnes", no?

No, I don't believe so. Firstly, how would using a USB isolator on a digital signal "mask" anything? Secondly, the low level detail of the HF was improved so it was possible to hear increased texture in cymbal brushing or drum brushing. Not just low level detail but also applause was rendered more realistically - with inner texture being evident. So, far from masking anything, the effect was to reveal information that was previously hidden or not evident

What seems to be happening is the removal of some noise (most likely common mode noise) riding on the digital USB signal is providing a cleaner. more stable operating environment for the USB audio devices handling of that signal. The CM noise was causing a distortion artifact most noticed when it was rendering the tape hiss on the recording.

If people are interested, I can set up a download of the first 30 secs or so of the recording for anyone to listen to & report their perception of the tape hiss.

 

[NorthSky]

That's cool John. Do you know how I can make my over few thousands CDs sound better? ...All the ones from the 80s and up to the mid 90s till almost late 90s.
Or should I go back to LPs and get a new TT with a USB port?

* And what is a "USB isolator"?

 

[Phelonious Ponk]

No, I don't believe so. Firstly, how would using a USB isolator on a digital signal "mask" anything? Secondly, the low level detail of the HF was improved so it was possible to hear increased texture in cymbal brushing or drum brushing. Not just low level detail but also applause was rendered more realistically - with inner texture being evident. So, far from masking anything, the effect was to reveal information that was previously hidden or not evident

What seems to be happening is the removal of some noise (most likely common mode noise) riding on the digital USB signal is providing a cleaner. more stable operating environment for the USB audio devices handling of that signal. The CM noise was causing a distortion artifact most noticed when it was rendering the tape hiss on the recording.

If people are interested, I can set up a download of the first 30 secs or so of the recording for anyone to listen to & report their perception of the tape hiss.

I guess I can imagine constant noise seeming to modulate with a signal as the signal's ability to mask the noise changes, but it's beyond me how the noise can "ride" on the signal...

Tim

 

[TBone]

I guess I can imagine constant noise seeming to modulate with a signal as the signal's ability to mask the noise changes, but it's beyond me how the noise can "ride" on the signal...

Dolby B/C?

 

[John Kenny]

That's cool John. Do you know how I can make my over few thousands CDs sound better? ...All the ones from the 80s and up to the mid 90s till almost late 90s.
Or should I go back to LPs and get a new TT with a USB port?

I don't know

* And what is a "USB isolator"?

It's a device that is put in-line on a USB cable & electrically isolates any upstream noise from getting downstream to the USB audio devices

 

[AJ Soundfield]

In a recent listening to some USB isolators there was a noticeable difference in the intrusiveness of this hiss - it was perceptually less intrusive with the USB isolator in-line.

This should be pretty easy for a DAC manufacturer if the noise/hiss reduction is of soundfield origin. Show the above threshold reduced noise in the analog output waveforms, with and without the "isolator". We would assume you measure these sort of things with your DACs on a regular basis.

 

[DonH56]

The two major USB noise coupling mechanisms are the power (supply voltage) and ground lines. Power is only an issue when the DAC is powered from the USB connector, a problem readily solved by an external supply (wall wart or whatever). Ground noise is a bit more insidious since noise from the source (like a PC) can be coupled to the output of the DAC by poor ground routing/routing. Keeping analog and digital grounds isolated is every bit as critical as isolating the voltage supplies (current is a loop, flows from voltage source to ground, and the amplifier does not know noise on the DAC's output ground is not signal).

 

[John Kenny]

The two major USB noise coupling mechanisms are the power (supply voltage) and ground lines. Power is only an issue when the DAC is powered from the USB connector, a problem readily solved by an external supply (wall wart or whatever). Ground noise is a bit more insidious since noise from the source (like a PC) can be coupled to the output of the DAC by poor ground routing/routing. Keeping analog and digital grounds isolated is every bit as critical as isolating the voltage supplies (current is a loop, flows from voltage source to ground, and the amplifier does not know noise on the DAC's output ground is not signal).

In this case the DACs are not powered from USB - they were self-powered
The interesting thing is that noise wasn't originating on the ground connection but on the USB signal lines itself as when the ground was connected across the isolation barrier, it made no difference - the audible improvement was still evident.

As I said, I suspect it is common mode noise

 

[fas42]

John, I don't take much notice of relatively 'pure' noise as such, it doesn't disturb anywhere near as much as it appears to do for others - but artifacts directly triggered by the musical content irritate the crap out of me! If the rendition of the recorded event captures my attention then I don't notice any noise, subjectively - the playback is doing its job well ...

I have an Odetta at Carnegie Hall, Vanguard CD - the average level of this is incredibly low, max volume on a previous system got about as loud as a kitchen radio - I'm sure the tape hiss is pretty evident if one listens for it, but I never notice it.

 

[John Kenny]

I know - as I said this is a reference track we often use & listen to. This is the first time that any change to this hiss has been so noticeable.
Here's a download link to the first 30 seconds or so of the track - let me know what you think of the hiss.

This is the track on Youtube & there's no discernible hiss but there's also no comparison to the SQ of the WAV file

 

[NorthSky]

1. What is the audio resolution on youtube, 128kb? ...Less, 64kb?
2. And that download link?

If the res is higher, would it indeed be easier to hear the distortion than if it's not?

It's like hi-def pictures (UHD, 4K); they show more in prominence the flaws, because it's more detailed, more defined, more resolved in pixels.
But on DVD everything looks "soapy", unfocused, lacking details, definition, so the flaws are easily masked. But the picture is like looking through a polluted cloud.
In 4K, from a true 4K source, it's like looking through a clear day atop a mountain with clean air all around and no windows, even if cleaned with Javex window cleaner.

With music (hearing), it is similar; the higher the audio resolution, the more the details and ambianc, e and hiss from the tape machine.
With an iPod and 92kB audio everything is cloudy and you hear no flaws, and no life either...just a highly compressed audio signal obscured by a dark cloud.
...Music for the kids on the go; windsurfing, surfboarding, snowboarding, skydiving, water skiing, bungee jumping, motorcycling, snorkeling, ...

Yes, no?

P.S. Is it possible to make that download link easier to access?

 

[fas42]

I know - as I said this is a reference track we often use & listen to. This is the first time that any change to this hiss has been so noticeable.
Here's a download link to the first 30 seconds or so of the track - let me know what you think of the hiss.

This is the track on Youtube & there's no discernible hiss but there's also no comparison to the SQ of the WAV file

Thanks for that - but, you need to make it public, I think "share" is the term Google uses.

 

[fas42]

1. What is the audio resolution on youtube, 128kb? ...Less, 64kb?
2. And that download link?

Bob, you have to "play games" with YouTube to get the best quality capture online. My experience is that this is the best way:

1. Upload the clip in a video resolution which is at least 720p, and have the audio in WAV quality, probably 48k rate is best. Otherwise, you're feeding compressed audio to another compression process - duuhhh ...

2. Watch or download the clip at 720p, nothing else(!!), if you want the best audio playback; reason is, there is some peculiarity in the YouTube handling of your upload, which results in the 720p version having 192kB audio - all other resolutions, higher ones as well, are at 128kB!! Huuuuhhhh??!! Go ask YouTube - I just know what works ... at the moment!

 

[John Kenny]

Sorry, I thought it was shared - try this link

 

[NorthSky]

Sorry, I thought it was shared - try this link

Thanks, that works: https://drive.google.com/file/d/0BzRUtXo0Yuj5THc1cDlLRkg2NWs/view

And you were right; distortion from that link, and no distortion from youtube.
The hiss on that google link came within the first ten seconds, very very obvious.

 

[amirm]

What is the audio resolution on youtube, 128kb? ...Less, 64kb?

Here is the latest settings for audio (when you are streaming): http://www.h3xed.com/web-and-internet/youtube-audio-quality-bitrate-240p-360p-480p-720p-1080p

 

[NorthSky]

Bob, you have to "play games" with YouTube to get the best quality capture online. My experience is that this is the best way:

1. Upload the clip in a video resolution which is at least 720p, and have the audio in WAV quality, probably 48k rate is best. Otherwise, you're feeding compressed audio to another compression process - duuhhh ...

2. Watch or download the clip at 720p, nothing else(!!), if you want the best audio playback; reason is, there is some peculiarity in the YouTube handling of your upload, which results in the 720p version having 192kB audio - all other resolutions, higher ones as well, are at 128kB!! Huuuuhhhh??!! Go ask YouTube - I just know what works ... at the moment!

Frank, that particular Oscar Peterson youtube video is @ 360p (the highest), and I assume the audio is @ around 192kB @ the very best, but my guess is half of that...96 kilobytes. Besides, those res are not right, it's more like less than that...128 or 64kb.

P.S. Amir beat me up to it.
- Picture: 360p
- Audio: 128kb

 

[fas42]

Here is the latest settings for audio (when you are streaming): http://www.h3xed.com/web-and-internet/youtube-audio-quality-bitrate-240p-360p-480p-720p-1080p

View attachment 732

That's the theory - the reality at the moment is that above 720p the quality is 128k, at least for me. Bizarre ... I actually uploaded an effectively 4k video, which didn't help - 720p is the "magic" setting.