I found it hard to distinguish between 320 kbps and uncompressed. On one occasion I thought the 128 kbps sounded the best. It is really hard to distinguish the difference if you aren't paying attention.

http://www.npr.org/sections/therecord/2015/06/02/411473508/how-well-can-you-hear-audio-quality



[Edited on January 14, 2016 at 12:43 PM. Reason : .]

1/14/2016 12:42:51 PM

I would imagine a lot depends on the quality of speakers or headphones you are using as well.

1/14/2016 1:10:07 PM

Using my Logitech G930 wireless headset, I scored 3/6. On two of my wrong answers, I chose the 320 kbps MP3. On the last of my wrong answers, I chose the 128 kbit MP3, and it was actually on the sample you might guess would be the easiest to get right.

1/14/2016 1:10:35 PM

^^ Yes of course. I used Audio Technica Noise ATH ANC7 Noise Canceling Headphones and a Creative XFi DAC. I turned off all enhancements on the DAC. With the enhancements on it is even harder to tell the difference.

1/14/2016 1:54:03 PM

I've heard quality audio many times

1/14/2016 1:59:26 PM

so on my macbook pro with my iphone headphones i got it right 4 out of 6 times. When I was wrong, I selected the 320kbps.

That's good enough for me to say I can tell a difference (guessing would be 1/3, and if it were truly random the 320kbps and 128 would be chosen equally).

That said it wasn't easy. I had to sit, close my eyes, and focus.

1/14/2016 3:26:42 PM

I got 3/6 with my cheap but well reviewed Monoprice headphones plugged into my desktop computer. I was not at all confident in any of my choices. I picked 128 kbps on the Katy Perry track. I was using differences in the high frequency sounds as my discriminator.

1/14/2016 4:40:04 PM

oh more importantly i gave a shit about the difference in quality in 0/6

1/14/2016 4:43:26 PM

'pends on the number of instruments a lot of time

I notice a huge difference in 160/192 and V0

inb4 treetwista?

1/14/2016 10:39:00 PM

ok, the first two tracks are unbearable to listen to, so I have officially dropped out of the challenge

1/14/2016 10:42:46 PM

got 4 out of 6

the two I got wrong were 320 (Jay Z and Vega)

had to jack up my earphones a bit but I finally got adjusted to which ones were the 128

320's usually pretty solid, but I've noticed that it doesn't work quite as well as V0 (and definitely not FLAC) for classical stuff

other stuff I don't really care I guess - can deal with 192 or better


yeah that's right... triple post and flip floppin' everything in dis bitch

1/14/2016 10:49:20 PM

I picked 320 for Katy Perry and Coldplay. I was kicking myself on the katy perry thing because I got cocky and only listened to them once each.

[Edited on January 15, 2016 at 4:17 AM. Reason : Monoprice headphones out of laptop.]

1/15/2016 4:16:02 AM

I got 3 out of 6, one them being the song they used to build/calibrate mp3 around.
Using a super nice sound card, receiver, and speakers.

Jay-z's track has such garbage production quality that it's tough to tell. Mozart and the accapella track are actually what MP3 does best, compression of single source audio.

1/15/2016 2:37:47 PM

Quote :

"oh more importantly i gave a shit about the difference in quality in 0/6"

1/15/2016 2:51:34 PM

I took this test since I'm considering different paid audio streaming services. Some services claim that they are better since they offer higher bitrates. If the difference is nearly imperceptible above 128k I'm not going to worry about bit rates. I currently use soundsurf.io (which is free) and of course the quality isn't that great.

1/15/2016 2:59:31 PM

all bitrates aren't equal. codec and encoding process matters just as much as the genre and mastering

1/15/2016 5:41:51 PM

5/6 but on a few I couldn't tell a difference and just guessed right

1/15/2016 11:32:47 PM

4/6

i picked 128kbps once

I basically tried to pick the ones with most crisp treble

1/16/2016 10:42:32 AM

Yep, crisp treble is one of the places I really hear the difference...little stuff like the details of enunciation of words (which is probably why they used the Suzanne Vega song for engineering mp3 compression). Percussion is another place it's more audible. Uncompressed stuff sounds like a film has been removed from the recording.

I got 4/6. I picked 128 on one, and 320 on one.

I listened to it send over AppleTV Airplay (lossless as far as I know, unlike Bluetooth), played over a $2000 receiver into $6000 worth of electrostatic speakers, plus a decent Klipsch subwoofer. (I got it all for about half-price. It's really, really good high-end stuff, although I wouldn't say exotic level, other than being electrostatic, which is kinda off the beaten path)


I have done a test like this in my car a couple of times with people who are pretty casual music listeners (once with mom, once with brother). I switched back and forth between a CD in my CD player, and the same song ripped to my iPhone at 256 bitrate. I think one song was Dire Straits - Tunnel of Love; I can't remember which other song I used. At any rate, they both got it right. I have decent audio in the car (Memphis Audio, 4 coaxial 6.5s and a Memphis amp...and a little powered sub tucked away by the spare tire), but nothing really high end like in my house.


I'm curious as to which songs people chose incorrectly on. A couple of these are good illustrations of compression; a couple are really poor examples to illustrate it, in my opinion (the one I chose 128 on, I can't really tell the difference even when I know the answer!). On the other hand, I feel like if we compile that and put it into the threat, that might taint people's responses. I will say that the 2 that I missed are kinda peas in a pod from a production standpoint.

1/16/2016 11:16:06 PM

I missed the Katy Perry one (128kbps) and I can't remember the other one

My PC is hooked up to
http://tinyurl.com/gwgpg2a
Sony center swapped in: http://tinyurl.com/zzs8b9p

I was pretty sure about Tom's Diner and Mozart. The other two I got by sheer luck. 128 kbps would be fine by me if it was recorded like on that website. A lot of 128 sounds like trash (yo, lemme jack up the equalizer for this one!)

[Edited on January 17, 2016 at 11:19 AM. Reason : ]

1/17/2016 11:14:51 AM