• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Can you really hear the lossy codec? - a public blind test (RESULT)

Which file is better

  • A is Better

    Votes: 6 37.5%
  • B is Better

    Votes: 6 37.5%
  • No difference

    Votes: 4 25.0%

  • Total voters
    16
  • Poll closed .

JohsonChou

Member
Joined
Aug 14, 2018
Messages
50
Likes
102
Location
California
Poll is now closed.
Answer is: A is lossless

6 people voted for A, and 6 people voted for B. So basically people are just blindly guessing...
Conclusion is that we generally cannot tell the difference between AAC 256kbps and Lossless.
The pursue of 24bit+ audio and higher bitrate than 48khz is probably not worth it as well.

If you voted for A, you likely only guessed it right
If you voted for B, don't feel bad, I might just saved you from buying that $1000 DAC.

On a side note, This experiment has a sample of 16 and a track downloaded from internet archive. Any "truth" drawn from this experiment is only as real as the invisible pink unicorn.
--------------------
Here is an interesting experiment I always wanted to do
I have an uncompressed wav file, then I compressed it to AAC 256kbps and then decoded it back to wav. I was careful during the conversions so both file is the same size

A: https://drive.google.com/file/d/1GqAUy26nnMXANCO5DSH87HrGvv8hdsCS/view?usp=sharing
B: https://drive.google.com/file/d/1-oJLTa5YOcB839fANwgIUcL0N_LQGAY9/view?usp=sharing

Please pick which one do you think sounded better in the poll
Please ask your friend to cast a vote so we get more data.
Poll will be closed in 7 days and I will post result to see if people can actually hear the difference.

Please do not use any software out there to "see" which file is compressed before casting a vote.
Please do not discuss what you have voted for or what the status of the poll is

To prove that one of these is compressed significantly, I attached a spectrum view where top 2 channels is the uncompressed file and the bottom two is the compressed one.

I got the music from Internet Archive https://archive.org/details/EFJAI1003/06-IfRightIsWrongLeftMustBeRight-Edit.flac
 

Attachments

  • difference.jpeg
    difference.jpeg
    219.8 KB · Views: 533
Last edited:

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
Interesting experiment. Though I would suggest next time you pick a different kind of music as the electronic music with only few instruments playing is not really a challenge for compression algoritm. For that reason I suggest next time you think of something like Pink Floyd - Time as the sound of those 2 clock alarms and the guitar solos will present much harder job for compression than electronic music of this kind.

P.S. I can't hear the difference

P.P.S. If I really have to decide I would vote that A is better :)
 
Last edited:

FrantzM

Major Contributor
Forum Donor
Joined
Mar 12, 2016
Messages
4,337
Likes
7,731
Interesting experiment. Though I would suggest next time you pick a different kind of music as the electronic music with only few instruments playing is not really a challenge for compression algoritm. For that reason I suggest next time you think of something like Pink Floyd - Time as the sound of those 2 clock alarms and the guitar solos will present much harder job for compression than electronic music of this kind.

P.S. I can't hear the difference

P.P.S. If I really have to decide I would vote that A is better :)

+1 .. A seems more "open"
 

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
Interesting experiment. I can't hear an obvious difference between the two after a couple of minutes of back and forth, but my money would be on A being the lossless file.

PS. It's hard to do a test like this though when the transition between the two is not immediate and seamless. Also have to +1 @Krunok on the choice of music, it's a well-mixed track but maybe not ideal for this kind of test.
 
OP
JohsonChou

JohsonChou

Member
Joined
Aug 14, 2018
Messages
50
Likes
102
Location
California
Interesting experiment. Though I would suggest next time you pick a different kind of music as the electronic music with only few instruments playing is not really a challenge for compression algoritm. For that reason I suggest next time you think of something like Pink Floyd - Time as the sound of those 2 clock alarms and the guitar solos will present much harder job for compression than electronic music of this kind.

P.S. I can't hear the difference

P.P.S. If I really have to decide I would vote that A is better :)

Yeah... I wish i can freely pick music... because of copyright reasons, I can only use royalty free music since I am pretty much offering it for download. This is the best I can find in 1 hour. If you have any royalty free lossless music, please let me know, I have been looking for a 24 bit high sampling rate one.
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro

cjfrbw

Senior Member
Joined
Mar 5, 2018
Messages
410
Likes
472
Could be expectation bias from reading previous posts, but A sounds a little less stressed and more reverberant. I wouldn't bet the bank on it.
 
OP
JohsonChou

JohsonChou

Member
Joined
Aug 14, 2018
Messages
50
Likes
102
Location
California
If it turns out that sample B is losless there may be absolutely no need to do that. :D

Well, to correct the bias in the comment section a bit:
Lossy codec would lose a lot of detail in high frequency (since most people can't hear it or won't pay attention to it). And it may cause the music to sound better since it is easier on the ear.
 

Soniclife

Major Contributor
Forum Donor
Joined
Apr 13, 2017
Messages
4,500
Likes
5,417
Location
UK
Playing with them in Foobar ABX quietly on my desktop speakers one of these does seem slightly smoother than the other, but my focus kept wandering off and there is no way I could pass a blind test right now. Even if I was passing the blind test I could not pick the lossy version, as it is I have no real preference.
 

FrantzM

Major Contributor
Forum Donor
Joined
Mar 12, 2016
Messages
4,337
Likes
7,731
Your systems are not revealing enough. That’s exactly what happens when you don’t use Kevyatta Power Cords and AmZon Woods isolation platforms. The music is homogenized. It loses its texture and organic PRATT. Hence everything sounds the same.
 

cjfrbw

Senior Member
Joined
Mar 5, 2018
Messages
410
Likes
472
Now, Frantz. Don't go all WBF on everybody.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,202
Likes
16,982
Location
Riverview FL
PS. It's hard to do a test like this though when the transition between the two is not immediate and seamless.

The files are aligned,

Drag them into Audacity, and play using the Solo button on one track then the other. I get seamless switching that way.
 

FrantzM

Major Contributor
Forum Donor
Joined
Mar 12, 2016
Messages
4,337
Likes
7,731
On a more serious side, this shows how difficult it really is to discern lossy from lossless. I thought my threshold was mp3/320. It seems to be lower. . My ears’ Golden platina is flaking.
 

Soniclife

Major Contributor
Forum Donor
Joined
Apr 13, 2017
Messages
4,500
Likes
5,417
Location
UK
What AAC encoder was used for this? The last time i played with lossy vs lossless the apple AAC encoder was a lot better than the others I tried, especially in it's iTunes plus vbr setting.
 
OP
JohsonChou

JohsonChou

Member
Joined
Aug 14, 2018
Messages
50
Likes
102
Location
California
What AAC encoder was used for this? The last time i played with lossy vs lossless the apple AAC encoder was a lot better than the others I tried, especially in it's iTunes plus vbr setting.

Just the regular aac codec in ffmpeg, highest quality, 256kbps VBR
 
Top Bottom