How Well Can You Hear Audio Quality? An online blind test.

[Bhh]

Anyone seen or taken this test before?
https://www.npr.org/sections/therecord/2015/06/02/411473508/how-well-can-you-hear-audio-quality

Its basically a blind test between uncompressed wavs and 320 kbps and 192 kbps mp3s for several songs. It also randomizes the order each time so you can take the test multiple times and not memorize the sequence. Just tried it at work on a MacBook air with a pair of crappy Bose noise cancelling headphones and was about 50/50 between the uncompressed and the 320kps. I never picked the 192 but some of them definitely felt like more of a guess to be completely honest. Pretty cool.

I'm going to try this on my home system as well to see if it makes it a little easier to hear the differences but I thought it was a really cool way of testing the fidelity of a system in a "double-blind" way and figured I'd share if you guys hadn't seen it before.

Doing an "ASR" version of this test wouldn't be too difficult and it would be super interesting to track the results over time.

 

[A800]

Too easy.
But I cheated using an exciter.
The compression is very noticeable especially in the highs.

 

[MC_RME]

https://www.audiosciencereview.com/...w-well-can-you-hear-audio-quality&o=relevance

 

[Bhh]

https://www.audiosciencereview.com/forum/index.php?search/1445989/&q=https://www.npr.org/sections/therecord/2015/06/02/411473508/how-well-can-you-hear-audio-quality&o=relevance

Seems I'm not the first to stumble across this, lol. I'd still love to see the long-term results of this after a couple hundred thousand clicks. After writing code all day, it's generally the last thing I want to do at night or on the weekends but I may have to bust out a version of this with some logging capabilities and the ability to feed it better source material at some point.

 

[algarvehifi]

hi There, so interesting. With some medium quality (30€) headphones borrowed I got 3 out of 6 right and the other 3 chosen the 320kbs never the 128. This is a good thing because I once told and confirmed by blind tests after , to a friend I could detect 128kbs mp3 from 256 or 320 but very difficult to chose between those 2 unless with a music I deeply knew.
this forum was my best online find of the year!
Regards from Portugal and stay safe

 

[Jimbob54]

I failed miserably. My system must need more air, resolving, slam and lots of prats. Can't be my ears.

 

[Rick Sykora]

Yes, ran into this test before. Without any basis for what is better sound, seems to me that the test subject is likely to pick what they like (rather than the best quality). A year or so ago, I contacted NPR and offered to improve the test at my own expense. The pleasant NPR editor said it was 4 years old, doubted it was live and forwarded to the music section. Never heard back.

So yes, ASR could do better, but suspect not much to prove to the membership. At the time, I was worried that we would continue to see a decline in recorded audio quality. Now that we have higher quality streaming offerings from the likes of Amazon and Tidal, am more hopeful.

 

[Sgt. Ear Ache]

A/B/X testing is really a more interesting and meaningful way to do this afaic. It's not so much "which of these files sounds better" that matters. It's "can you actually identify any difference between these files such that you can tell which of A or B is X?" And there's not really a point to even including 128kb files anymore. Nobody really ever argues that low BR lossy isn't identifiable. It's 256kb and up that matters...

 

[Jimbob54]

Re did this a few times today. From my (still poor) results I have concluded the following:

1. The classical is the only one I can reliably get correct.
2. The Coldplay sounds just awful regardless of format.
3. The beat/ bass heavy ones (Jay-Z and Katy P) I most often get wrong and have a fair chance of picking the 128kbps.

 

[GD Fan]

I'd had serious doubts about the clarity of my system after reading here for the past several months, but truth be told this was extremely easy, a perfect 6/6. And I don't have any faith in my ability as a listener at all. It was played through my TV, so maybe 5.1 is 'cheating'?

 

[GD Fan]

Re did this a few times today. From my (still poor) results I have concluded the following:

1. The classical is the only one I can reliably get correct.
2. The Coldplay sounds just awful regardless of format.
3. The beat/ bass heavy ones (Jay-Z and Katy P) I most often get wrong and have a fair chance of picking the 128kbps.

Oh yeah, the Coldplay track was terrible...

 

[magicscreen]

I'd had serious doubts about the clarity of my system after reading here for the past several months, but truth be told this was extremely easy, a perfect 6/6. And I don't have any faith in my ability as a listener at all. It was played through my TV, so maybe 5.1 is 'cheating'?

No cheating. Anyway playing in a browser probably is low quality enough.
So big congrats for you! You have a good ear.

 

[GD Fan]

No cheating. Anyway playing in a browser probably is low quality enough.
So big congrats for you! You have a good ear.

It must have been the beer.