Anybody Out There Who Hears a Difference Between 320 kbps MP3 and Red Book CD? What Differences Do You Hear?

[luft262]

I did several single-blind tests where my wife would select either an AAC 256, MP3 320, or lossless FLAC file using YouTube Music, Tidal, Qobuz, and/or Amazon HD music and I'm usually able to get about 9 out of 10 right when comparing lossy to lossless. First I listen for little echos, ticks, or other sounds that come from the musicians, their instruments, or their voices. If I can't find those then I simply clear my mind and ask which sounded clearer. To do this reliably I have to use headphones. I can't do it with speakers, but maybe my speakers just aren't good enough

 

[MediumRare]

I did several single-blind tests where my wife would select either an AAC 256, MP3 320, or lossless FLAC file using YouTube Music, Tidal, Qobuz, and/or Amazon HD music and I'm usually able to get about 9 out of 10 right when comparing lossy to lossless. First I listen for little echos, ticks, or other sounds that come from the musicians, their instruments, or their voices. If I can't find those then I simply clear my mind and ask which sounded clearer. To do this reliably I have to use headphones. I can't do it with speakers, but maybe my speakers just aren't good enough

Did you do the NPR test?

 

[Zensō]

Here’s the BBC white paper on the subject: https://downloads.bbc.co.uk/rd/pubs/whp/whp-pdf-files/WHP384.pdf

 

[Bjorn]

I did several single-blind tests where my wife would select either an AAC 256, MP3 320, or lossless FLAC file using YouTube Music, Tidal, Qobuz, and/or Amazon HD music and I'm usually able to get about 9 out of 10 right when comparing lossy to lossless. First I listen for little echos, ticks, or other sounds that come from the musicians, their instruments, or their voices. If I can't find those then I simply clear my mind and ask which sounded clearer. To do this reliably I have to use headphones. I can't do it with speakers, but maybe my speakers just aren't good enough

While the speakers obviously matters, the room and acoustics generally have much more impact on areas like clarity and resolution. You have to deal with specular reflections and room modes if one wants to hear details well. The lateral Toole approach with active side wall contribution that is often proclaimed here has a detrimental effect on accuracy.

 

[Digital_Thor]

In slower and more intimate music pieces with a bit of "room" sound, where voices and maybe a single instrument "rings out" slowly into the room. Here I sometimes hear the slight compression and maybe even soft clipping of lower resolution recordings. But if its well recorded - I mostly find 320kbps enough to keep me in the dark I'd much rather have well recorded music on youtube via pc, toslink and whatever, than badly recorded music in FLAC quality via all kinds of fanzy trickery

 

[solderdude]

I did several single-blind tests where my wife would select either an AAC 256, MP3 320, or lossless FLAC file using YouTube Music, Tidal, Qobuz, and/or Amazon HD music and I'm usually able to get about 9 out of 10 right when comparing lossy to lossless. First I listen for little echos, ticks, or other sounds that come from the musicians, their instruments, or their voices. If I can't find those then I simply clear my mind and ask which sounded clearer. To do this reliably I have to use headphones. I can't do it with speakers, but maybe my speakers just aren't good enough

Use 1 source, say 44.1/16 and encode this, for instance with LAME-XP in MP3 320 in highest quality setting. Throw in MP3 VBR in highest setting.
Compare those.

Do this with a few songs, maybe some songs that have been shown to give audible effects.

You will never know if any of the music services used poor encoding or watermarks or even used the same masters.

 

[Marc v E]

When I did a double blind test 320 kbps was where I would have to guess sometimes and actually preferred it on occassion over cd quality.

The only difference I could perceive is that the 320 kbps version could sound more pronounced in the bass.

 

[Zensō]

When I did a double blind test 320 kbps was where I would have to guess sometimes and actually preferred it on occassion over cd quality.

The only difference I could perceive is that the 320 kbps version could sound more pronounced in the bass.

Archimago’s blind test uncovered a preference for high bitrate MP3 over lossless:

http://archimago.blogspot.com/2013/02/high-bitrate-mp3-internet-blind-test_3422.html

 

[luft262]

Did you do the NPR test?

Yes. I can pass the NPR test. But I've done that test several times so I know exactly what to listen for in each track. It's harder if I don't know which tracks are coming.

 

[luft262]

While the speakers obviously matters, the room and acoustics generally have much more impact on areas like clarity and resolution. You have to deal with specular reflections and room modes if one wants to hear details well. The lateral Toole approach with active side wall contribution that is often proclaimed here has a detrimental effect on accuracy.

My living room has porcelain tile flooring, so that's probably a problem. However, there are lots of bookshelves, a large rug, and furniture to dampen reflections. The back of the room opens into the kitchen, which has higher 18 foot ceilings. I'm not sure how all of that would affect sound. I would assume the flooring and windows are bad, but the open back wall (or lack thereof) probably helps things?

 

[luft262]

Use 1 source, say 44.1/16 and encode this, for instance with LAME-XP in MP3 320 in highest quality setting. Throw in MP3 VBR in highest setting.
Compare those.

Do this with a few songs, maybe some songs that have been shown to give audible effects.

You will never know if any of the music services used poor encoding or watermarks or even used the same masters.

Yeah, there are a lot of potential variables between music streaming services. I was mostly just trying to find out if I prefered one service over another or if I even noticed any audible difference at all. They might all be using different masters, like you said.

 

[luft262]

Archimago’s blind test uncovered a preference for high bitrate MP3 over lossless:

http://archimago.blogspot.com/2013/02/high-bitrate-mp3-internet-blind-test_3422.html

Thanks for sharing. I just read this for the first time! Crazy! But I believe it. It's hard to tell!

 

[Bjorn]

My living room has porcelain tile flooring, so that's probably a problem. However, there are lots of bookshelves, a large rug, and furniture to dampen reflections. The back of the room opens into the kitchen, which has higher 18 foot ceilings. I'm not sure how all of that would affect sound. I would assume the flooring and windows are bad, but the open back wall (or lack thereof) probably helps things?

Helps some, but there are likely still a lot of high gain specular energy and low frequency decay and resonances that hinders you from hearing the recorded signal with accuracy.

 

[Kegemusha]

For what I can experience, I think with high DR music, you can hear the difference but has to be music you know and you really need listen. Mostly Jazz, classical or well recorded CD, like I can hear the difference in Radioheads CDs, but you really need to listen, if you have the music like background listening, I dont think it matters.

If you would then listen to music with very low DR, like 6 (for example, some CDs I listen have these DR) or something like that it would be difficult to discern the difference.

My wife for example, can not tell the difference between a cassette and a CD

 

[daedalus]

I couldn't hear a difference using the test from this site, so much so that I didn't even complete the test. Fiio BTR5 via USB on a DT1990.
http://abx.digitalfeed.net/

Does this mean that there is even less of a difference between CD quality lossless vs. bluetooth at 990 kbps with the same lossless source ?

 

[HRTF_Enthusiast]

I couldn't hear a difference using the test from this site, so much so that I didn't even complete the test. Fiio BTR5 via USB on a DT1990.
http://abx.digitalfeed.net/

Does this mean that there is even less of a difference between CD quality lossless vs. bluetooth at 990 kbps with the same lossless source ?

This test is flawed. It uses music you don't know and it's not possible to swap between short time slices.

FLAC vs. 320 kbps MP3 using Amir's method with a real ABX tool.

For the test, I used the song, "Signs of Love," from the Persona 4 soundtrack. As I converted directly from flac to mp3, there was no reduction in volume.

I can easily pass this test by listening to the note at the beginning of the song and quick switching between FLAC and MP3. The attack sounds blunted on the MP3.

 

[flipflop]

This test is flawed. It uses music you don't know and it's not possible to swap between short time slices.

FLAC vs. 320 kbps MP3 using Amir's method with a real ABX tool.

For the test, I used the song, "Signs of Love," from the Persona 4 soundtrack. As I converted directly from flac to mp3, there was no reduction in volume.
View attachment 172173View attachment 172174
I can easily pass this test by listening to the note at the beginning of the song and quick switching between FLAC and MP3. The attack sounds blunted on the MP3.

Please share the files if you still have them.

 

[Jimbob54]

I'm perfectly willing to believe people can discern between the two. I also understand getting the best source for your files when possible.

The real question is, if you only ever had ready access to the mp3, are the differences enough to change ones enjoyment of the files in their reduced form? Or if the FLAC was 3 or 4 times more expensive?

Or, put another way, is the "blunted attack" etc that @HRTF_Enthusiast notes a problem if you'd never A/B with the FLAC?

 

[Zensō]

I'm perfectly willing to believe people can discern between the two. I also understand getting the best source for your files when possible.

The real question is, if you only ever had ready access to the mp3, are the differences enough to change ones enjoyment of the files in their reduced form? Or if the FLAC was 3 or 4 times more expensive?

Or, put another way, is the "blunted attack" etc that @HRTF_Enthusiast notes a problem if you'd never A/B with the FLAC?

Yup, agreed. Fast switching to pick out minute differences in transients or reverb tails only tells us about fast switching. The more relevant questions are whether or not these differences are noticeable in normal listening, and whether they keep one from enjoying music.

Interestingly, in his blind test, Archimago uncovered a clear preference for high bit rate MP3 over lossless Redbook, which is sort of hilarious.

High Bitrate MP3 Internet Blind Test: Part 2 - RESULTS

A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.

archimago.blogspot.com

High Bitrate MP3 Internet Blind Test: Part 3 - DISCUSSION

A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.

archimago.blogspot.com

 

[BoredErica]

Is Opus at 512kbps for stereo totally transparent even with killer tracks?