Head-Fi.org › Forums › Equipment Forums › Sound Science › BLIND TEST: 128kbps mp3 vs Lossless
New Posts  All Forums:Forum Nav:

BLIND TEST: 128kbps mp3 vs Lossless - Page 2

Poll Results: Which file is Lossless and which is mp3?

 
  • 16% (44)
    A is Lossless, B is 128 kbps mp3
  • 68% (181)
    B is Lossless, A is 128 kbps mp3
  • 15% (40)
    They both sound the same
265 Total Votes  
post #16 of 180

B is Lossless

A is Lossy

 

Yes, I did see the answers before I listened. That made me listen better.

 

The beginning's drums and bass are heavier on B. At :25, A is more closed in and brighter. Also, the bassdrum is heavier on B. B was more enjoyable to listen to. It was only slightly better, though. Akin to an improved cable. In-fact, the more I listened to them, the more they sounded the same. I can hear the differences better on other songs with flac vs 320 mp3.

 

Production sucks either way. Next time, try a better song like Rihanna or something.


Edited by Deverica Wolf - 2/4/12 at 7:56pm
post #17 of 180

I got it right while listening, but didn't read the poll options carefully and voted the wrong one :D

 

I haven't trained my ears at all for this kind of thing. Usually i can tell when it's something like a Youtube transcode. It's sad how often those come up in friends' collections. Anyway, i can usually tell by distortion in the cymbals, so that's what i tried to listen for here.

I'm sure the volume difference influenced my decision, though :P

post #18 of 180

Voted and thanks for the test!  It could be interesting if you would create another one using 192 kbps or 320 kbps and the result may be very different.

post #19 of 180
Thread Starter 

Ya, I have a good song to do it with also.

post #20 of 180

I believe the test is a little bogus in the first place...  I have a Goo Goo Dolls album with Black Balloon on it.. and I can say that the original recording is utter junk in the first place, making it difficult to distinguish differences between the lossless and down-sampled version.

 

I'd like to see the same test with a different band/musician as many of the Goo Goo Dolls tracks are of bad recording or mixing quality.

post #21 of 180
Quote:
Originally Posted by Mak333 View Post

I believe the test is a little bogus in the first place...  I have a Goo Goo Dolls album with Black Balloon on it.. and I can say that the original recording is utter junk in the first place, making it difficult to distinguish differences between the lossless and down-sampled version.

 

I'd like to see the same test with a different band/musician as many of the Goo Goo Dolls tracks are of bad recording or mixing quality.



That was kind of why I mentioned that it would help to know the song. Even if you start with a less than recording, it's not unreasonable to expect a difference between a low quality and quality file of the recording. There will be a more pronounced difference if you start with a high quality recording.

 

When I replaced my medium quality MP3s with FLAC, I accidentally left the MP3 files for 2 albums on my player's playlist. I had listened to those a couple of times and wondered if those CDs just had bad recordings or if the perceived benefit of the higher quality file was in my head. I eventually realized that I hadn't been using the FLAC files. Not exactly the most scientific test, but I did believe I was using the FLAC and noticed a difference of some sort.

post #22 of 180

This was easier than I expected. Since being at work I didn't even use my higher-end setup, just the ATH-AD900 and the crappy headphone jack from my Thinkpad X220. Even then it's not that difficult to tell. It was becoming pretty obvious in the lower frequencies I think. 128k mp3 tends to sound a bit hollow and tinny. 

post #23 of 180

Cool test, vote in spoiler below 

 

Vote (Click to show)

B is lossless

A is lossy

 

I'd propose we do this test again, but with no poll. Someone merely posts the files, then votes are sent through PM to the user and clocked without everyone being able to see. The only posts in the thread would be to acknowledge you voted, not how you voted until the test was completed.

 

Does this sound like a reasonable idea?

post #24 of 180

I voted both sound the same.

 

apparently, going from A to B, at first B seemed a little... how can I say... fuller

 

however, doing the same without looking/knowing what file was playing they sounded exactly the same. (test done by using foobar +speakers)

 

it would be quite something if these files are actually the same... lol

 

post #25 of 180

This test was slightly harder than expected but it was still obvious enough which was which.  Probably had to do with the recording quality of the song like someone mentioned.  I did a 320 cbr vs 128 and that one was much much easier.  I can even tell with my crappy laptop speakers which is hilarious.

post #26 of 180
Thread Starter 

I purposely did not choose a hi-fi recording because after all, is that all we listen to?

 

Most would agree that the changes are pretty small with 128 to flac, and even smaller with 320.  I personally always try and get mp3 VBR V2 or better.  128 is pretty useless today with the cost of space.  If you want to find the real diminished spot, go listen to a 64 then a 128.  It's like the difference between ibuds and HD600's - Massive.  I personally cannot hear a difference, but then again I hate to listen critically.  I just like the overall clear sound.

 

And yes, A is the mp3.  Thanks for participating.


Edited by Crazy*Carl - 2/5/12 at 7:37am
post #27 of 180
Quote:
Originally Posted by Crazy*Carl View Post

I purposely did not choose a hi-fi recording because after all, is that all we listen to?

 


Maybe so, but even then, is it a fair test if you intentionally choose music that you know isn't going to sound much different? You can't just compare 128 kbps vs lossless with one track whether it's a good or bad recording, either. Anyway, thanks for posting. Was pretty interesting to see the results. 

post #28 of 180
Imported both tracks in fb2k, scanned with replay gain, played both for ~10 secs. The cymbals etc. sounded less clearer and slightly clipped on A => mp3.

My abx test setup: some realtek onboard + px100. Yes, I'm serious.
post #29 of 180

Subtle difference, but definitely there. I got it right, before reading the rest of the thread. For me, the bass drum impact in the first seconds revealed which one was lossless.

post #30 of 180

Fairly easy TBH. The lossless, B sounds less compressed in the dynamic range, the soundstage is more airier, the highs have better sounding resolution and airiness to them and the kickdrum's impact in beginning was more clear and defined and realistic (especially the "after-impact" fading sound in the lossless one.

 

For me it was like the lossy didn't engage me with the music, felt like I was left out due to the a lot more compressed dynamic range especially so it sounded more streamlined and less vivid while the lossless made me more interested and I got better engaged with the track as the small details described above came out much more clearly so it made it more interesting to listen to.


Edited by RPGWiZaRD - 2/5/12 at 9:34am
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › BLIND TEST: 128kbps mp3 vs Lossless