[WildStyle-R11]

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

Gives you 3 different quality files, you listen to them and you determine which one is the highest quality one.

I am destined to pick 128kbps files. Like seriously I only once or twice picked 320 over 128.

They all seem pretty much the same to me to be fair. So the question is it just me or the quality is as good on all of them? Cause I can tell the difference between a good and a bad file, but these ones don't seem to have any quickly spottable issues. But then again that is just playing the track once , can't get much of a opinion on that one.

 

[ProtegeManiac]

I think there was a thread on that same page already. In any case, that's a slightly better test than previous ones, which used cuts from tracks like vocal audiophile recordings, or instrumentals. The problem with those is that those don't cover a wide enough range to be affected by MP3 compression, plus most being relatively less familiar with them, it's harder to hear any difference. In my case metal and jazz tend to take a hit with anything below 320kbps, as the bass notes are more audible.

 

That said, I wasn't familiar with those tracks either, but I got nearly all of them correct, except there's a flaw that is only relevant for those with crappy internet: whenever I click on a track to play one of them doesn't play immediately, so it might have colored my perception of which one is lossless, although I consistently heard bloated bass. If anything, it's actually the one that I got wrong that validates my hearing. I got it wrong because I momentarily forgot that the task was to choose lossless, and had in mind to choose which Jay-Z track I preferred, which was 320kbps because the other tracks had hollow bass or had bloated bass (this is likely the lossless one). It might be the same case as yours, with yours more severe - the compressed track seems more clear because the enhanced bass response on the headphones we use is mucking up the lossless track. Basically, Jay-Z was using an accurate monitor system when he was mastering the tracks; and then in the case of my music, the upper bass hump on the Sennheiser HD600 is compensating for how you don't get that "kick in the chest" sensation that I'm used to with speakers.

 

In any case, as much as test tones show that with the right speaker or headphone I can still hear a little past 20khz, I can't recall any instance where it wasn't the bass notes that I can tell were affected by compression. The only times treble seemed drastically different wasn't with the file types (not even SACD/DSD), but with driver design - non-dynamic driver designs tend to be a lot smoother than dynamic drivers, save for a few like the Vifa XT25, or Sonus Fabers.

 

[dsf3g]

I think part of the problem is that, to the extent that different compression schemes color the music (as opposed to producing bad sounding artifacts like skips or crackle), you personally might actually *prefer* the sound of the lower bitrate encoded file.
 
Years ago, I boldly proclaimed on one of these threads that I could always distinguish 96kb encoded music from 128kb encoded, and that it was only higher bitrates that all pretty much sounded the same to me (I'm rather modest in my hearing claims). Well, a member of the forum asked me to put up or shut up, as it were and perform an ABX test. So I dowloaded an ABX plugin for a music player I had, and gave it a go. My results stunned me. Because, while I was able to consistently distinguish between 96kb and 128kb files, I *wasn't* able to say which one sounded better, which one I preferred. They just sounded a little bit different... one was a little bassier, while the other seemed a little clearer. And in all honesty, depending on my listening mood and the sort of music I'm listening to at any given moment, I might prefer the one over the other.

 

[LNuneek]

I did the Tidal test awhile back and I was able to pass it whenever I really wanted to, but I had to focus on aspects of the sound that didn't really matter. I mean really focus too. It wasn't easy. The differences were so minute. If I went with what I thought sounded better and did the tests quick I failed almost every time.
 
After that I realized I'm perfectly fine with Spotify and MP3's.

 

[cjl]

  I did the Tidal test awhile back and I was able to pass it whenever I really wanted to, but I had to focus on aspects of the sound that didn't really matter. I mean really focus too. It wasn't easy. The differences were so minute. If I went with what I thought sounded better and did the tests quick I failed almost every time.

That was a flawed test too - it intentionally added differences that would not be there if the only difference was the encoding. The actual difference in quality is much smaller.

 

[kar13]

I do constant A/Bing with 256 kbps aac,320 kbps mp3, 16 bit and 24 bit flac audio.Truth be told A/Bing with 16 bit flac and 24 bit flac audio is a little difficult.Otherwise, with rest of them one can easily make out.You can make out that latest mp3 codec standard is heavily v shaped as the algorithm truncates samples of the mids way too much, in comparison to aac 256 kbps.
Atleast, that's the way I identify mp3 320 kbps,aac

 

[-bynx-]

You know how at a certain resolution, the eye can no longer tell individual pixels apart on an LCD screen?... ('retina displays')
At what resolution are your ears no longer able to discern the difference between the recorded signal and the actual sound source? (assuming you could blindly A/B between the two)

 

[Brooko]

  You know how at a certain resolution, the eye can no longer tell individual pixels apart on an LCD screen?... ('retina displays')
At what resolution are your ears no longer able to discern the difference between the recorded signal and the actual sound source? (assuming you could blindly A/B between the two)

 
Each person is different.  Personally, my threshold for transparency is about aac256.  The good news is that the software to test yourself is freely available, and all it takes is time, an open mind, and a willingness to be completely honest with yourself.
 
Here - this should help get you on your way : http://www.head-fi.org/t/655879/setting-up-an-abx-test-simple-guide-to-ripping-tagging-transcoding

 

[-bynx-]

Thanks, I'll look into it. I'm curious if there's a statistical average though.

 

[cjl]

For most people, and with most samples of music, AAC256 and 320 LAME MP3 should be completely transparent.