Blind test on human perception of THD

[AltoVariago]

I read this forum with a lot of interest for the objectivity and the importance given to numbers and measures.

I have a question: did anybody measured (in a blind test, obviously) what is the amount of THD that the trained human ear can detect by perception of a deterioration of sound quality?

Thanks.

 

[RayDunzl]

I don't have the numbers, but...

With a pure tone (sine wave) - not much. You begin to hear additional tones.

With music - much more, due to the harmonics of voice or instrument. You won't be able to dissociate added from inherent "distortion".


Here is a single note of an electric bass guitar (clean electrical signal only), gently plucked with my thumb, to keep the second harmonic from overwhelming the fundamental, so the "distortion" numbers apply correctly.

There's software available to apply measured distortion to music...

Beta Test: DISTORT - audibility of distortions

DeltaWave beta-test was a huge success, thanks to all of you! This new beta-test is again, only for the brave and the curious :) Introducing DISTORT - free software designed to help determine the audibility thresholds of various audio distortions with your own equipment, at your own pace, with...

www.audiosciencereview.com

 

[Sokel]

Do it and join the club:

Listening Test

www.klippel.de

 

[antcollinet]

Do it and join the club:

Listening Test

www.klippel.de

And in those tests if you look at the results (you have to choose a test setup to see the results for that setup)

about 98% of people who took the test were unable to detect distortion lower than about -50dB

 

[Sokel]

And in those tests if you look at the results (you have to choose a test setup to see the results for that setup)

about 98% of people who took the test were unable to detect distortion lower than about -50dB

View attachment 317818

And these are not just any people as you have to be informed somehow to get there and take the test,so someone has to have some good relationship with sound.
(is it about 15.000 people who took the test,or I read it wrong? )

 

[antcollinet]

And these are not just any people as you have to be informed somehow to get there and take the test,so someone has to have some good relationship with sound.
(is it about 15.000 people who took the test,or I read it wrong? )

I counted about 18K - so there or thereabouts.

Further - many people will have taken the test more than once (i have) and learn how to detect better each time. So a degree of training involved.

Oh, and I cheated once and used earbuds instead of speakers - doubled my score from around 20 to around 45. People doing this will also skew the results.

 

[BDWoody]

I read this forum with a lot of interest for the objectivity and the importance given to numbers and measures.

I have a question: did anybody measured (in a blind test, obviously) what is the amount of THD that the trained human ear can detect by perception of a deterioration of sound quality?

Thanks.

This might be a useful read.

Audibility thresholds of amp and DAC measurements

Introduction A recurring theme on ASR is whether or not the various measured qualities of the devices are audible. In this post, I'll present some clear and visual thresholds for when those imperfections can be considered a potentially audible concern. I will not explain the basics of amp/DAC...

www.audiosciencereview.com

 

[danadam]

Oh, and I cheated once and used earbuds instead of speakers - doubled my score from around 20 to around 45. People doing this will also skew the results.

Is it cheating when they expect you may be using headphones?

I read somewhere that there was some "tell" in one of the tracks (or maybe all). I don't know if it was fixed or not.

 

[AltoVariago]

Thanks to everyone!

 

[AltoVariago]

Do it and join the club:

Listening Test

www.klippel.de

Very interesting but the samples are inappropriate. A human voice or a violin or flute solo would allow a more appropriate evaluation

 

[Blumlein 88]

Very interesting but the samples are inappropriate. A human voice or a violin or flute solo would allow a more appropriate evaluation

Well it depends. A sine wave test tone gives the best results possible. A very simple recording of voice or violin is less good, but better than a multi-layer studio mix. Do you want to know what the limits are with the easiest signal, with the easiest music signal, with a normal music signal or with the most dense hardest to hear music sample. One obviously worthy approach is a sine wave test tone. You know the limits of that and you won't do any better. OTOH, with much music people listen to you'll not hear it until around 3%. One reason we can enjoy music on some pretty lo-fi setups.

Oh, and the Distort software mentioned earlier, you can dial in any level you wish using your preferred test track.

 

[AltoVariago]

Being a music lover, I know how a violin sounds but I have no experience of sine waves!

 

[Blumlein 88]

Being a music lover, I know how a violin sounds but I have no experience of sine waves!

Nevertheless, with no prior experience you could hear a slightly distorted sine wave as different vs a pure sine wave at much lower distortion than you could ever manage with the most pure of violin sounds.

 

[antcollinet]

Being a music lover, I know how a violin sounds but I have no experience of sine waves!

So pick whatever music you want. Distort it with Distort, and see what level you can hear.

 

[RayDunzl]

Being a music lover, I know how a violin sounds but I have no experience of sine waves!

Blow across the top of a beer bottle.

That's darn close to a sine:

 

[ahofer]

There was an online test somewhere. Klippel?

Listening Test

www.klippel.de