Jump to content
LogicprObe

For those who think MP3 is great.

Recommended Posts

Every blind or double blind test I've seen, informal or properly conducted and controlled on a large scale, has shown people can only tell to a degree of chance whether the sound is a 128-160kbps MP3 or an uncompressed WAV of a CD track. They're usually done with audiophile equipment and with people who claim to be able to tell the difference.

 

I've also seen tests where people are deliberately led to believe something is superior when it's randomly either worse, the same or better quality (in terms of compression)... and regardless of the actual change, people will to an incredibly high degree claim they can tell it's better (even when it's worse).

 

I don't think I'm ever going to bother updating my 128kbps collection. But I'll tell all my guests that it's 640kbps and pumping out of $4,000 speakers from a $1,000 sound card.

Edited by tantryl

Share this post


Link to post
Share on other sites

I agree to an extent tantryl. A good 128kbps track is plenty for most consumer audio gear. Dynamic range compression, however, is where the biggest issue lies. You can hear the difference with the shittiest of equipment, and it's quite horrific. My main gripe with MP3 isn't so much its perceived quality, but the fact that you're introducing artifacts for no good reason. Internet quota and storage capacity is not longer a hindrance, so it would be great if online stores offered lossless formats. Then you could at the very least have a archive copy for your collection. If you decide you want to conserve space, then your media player can store MP3 tracks, and your larger capacity PC can have the lossless files. Win win.

Edited by .:Cyb3rGlitch:.

Share this post


Link to post
Share on other sites

I don't think I'm ever going to bother updating my 128kbps collection. But I'll tell all my guests that it's 640kbps and pumping out of $4,000 speakers from a $1,000 sound card.

You could tell people that, but anyone with half a clue that isn't half deaf, will forever be sceptical of anything you say from that point on :P

Share this post


Link to post
Share on other sites

I'm wondering if I should explain dynamic range better. I explained it in a generalised manner because it involved math. The ratio doesn't exactly represent 'loudest' to 'softest', it's really 'loudest' to 'RMS amplitude'. Then again, you could classify the RMS amplitude as 'soft'.

Edited by .:Cyb3rGlitch:.

Share this post


Link to post
Share on other sites

Every blind or double blind test I've seen, informal or properly conducted and controlled on a large scale, has shown people can only tell to a degree of chance whether the sound is a 128-160kbps MP3 or an uncompressed WAV of a CD track. They're usually done with audiophile equipment and with people who claim to be able to tell the difference.

Huh? I'm totally against most audiophile pseudo-science bullshit, insisting anyone that says otherwise should try some blind AB tests.

 

But not being able to tell the difference between 128kbs and FLAC? I'm happy to rise to that challenge.

 

After exams, that is.

 

Hardware will be custom USB DAC (self build, value ~$50), Gilmore Dynalo (self built, value ~$250) and Alessandro MS-1s (value ~$150). Or, in other words, audiophile equipment on a budget.

 

Can anyone suggest some software?

 

[edit]: Excellent, foobar2k comes with an ABX comparator :)

 

Rob.

Edited by robzy

Share this post


Link to post
Share on other sites

Mate, the simpler the better.

It's hard enough explaining the difference between audio and data compression!

I added a quick mention in there anyway. Saves me hearing from people that I "got it wrong". :P

 

I did some quick tests between NiN's free track '1,000,000' encoded in 128kbps and FLAC. I'll admit I struggled to hear any difference. Then again, the song sounds like noise in the first place...

Edited by .:Cyb3rGlitch:.

Share this post


Link to post
Share on other sites

These tests also depend on the quality of the playback gear, as has been previously mentioned in this thread.

tantryl specified "audiophile equiptment" :P

 

But yeah, all I can prove is that on my equipment I can hear a difference between 128kbs and FLAC.

 

Now the question is, what song is best for a test like this?

 

Rob.

Share this post


Link to post
Share on other sites

Use 'Lights in the Sky' by Nine Inch Nails. Purely because it's rated 15 for dynamic range, free, and you can get a studio quality version.

Edited by .:Cyb3rGlitch:.

Share this post


Link to post
Share on other sites

These tests also depend on the quality of the playback gear, as has been previously mentioned in this thread.

tantryl specified "audiophile equiptment" :P

 

But yeah, all I can prove is that on my equipment I can hear a difference between 128kbs and FLAC.

 

Now the question is, what song is best for a test like this?

 

Rob.

 

I can hear a difference when they play MP3 files on AM radio.

Share this post


Link to post
Share on other sites

It'd be interesting if you grabbed some high, low, and ok DR tracks to see if it has an affect on the perception of artifacts.

 

If you can get your hands on a copy of this album, you've got a wide range right there.

Share this post


Link to post
Share on other sites

Thanks for the DR range suggestions, I'll grab a copy of the NIN ones on offpeak.

 

[edit]: The NIN website only has MP3. While 320kbps is all well and good, I was hoping for FLAC.

 

I can hear a difference when they play MP3 files on AM radio.

Hrm, I don't believe I've ever heard an MP3 file on AM radio before... but I only listen to AM when I'm driving a car after my old man, and it's always talk back :P

 

Rob.

Edited by robzy

Share this post


Link to post
Share on other sites

Just wanted to add in that NIN are awesome in any format :P

 

As for preference on media or delivery for some reason I still prefer vinyl... and I think it is all down to the music listening ritual that accompanies it... completely off topic but it did remind me of the arguments about analog vs digital when CDs first came out, and thus reminded me of vinyl

 

(unfortunately my son killed the stylus on my record player a while ago... was a cheap turntable but was still nice to spin up the occasional oldie or not so oldie)

Share this post


Link to post
Share on other sites

NiN site has 24/96 wav, FLAC etc. too.

 

It even says to on the sign-up page.

Ah, excellent, thank you.

 

The download sign-up page only lists MP3.

 

[edit]: I really want to grab the FLAC and WAVE, and see if there's even a binary difference between the two's audio :P

 

Rob.

Edited by robzy

Share this post


Link to post
Share on other sites

But yeah, all I can prove is that on my equipment I can hear a difference between 128kbs and FLAC.

 

Now the question is, what song is best for a test like this?

One song with only MP3 and FLAC? No, I'm afraid that proves approximately dick all in either direction, except maybe that when given two choices someone will often choose one of them.

 

I'd say at bare minimum four songs and three different compression settings. How are you intending on making the test blind? I'll assume double blind is a bit hard considering the circumstances.

Share this post


Link to post
Share on other sites

I'd say 9 songs with differing amounts of dynamic compression. Have a copy of each encoded in FLAC, 320kbps MP3, and 128kbps MP3. Then get someone to select and play each set of tracks at random, and record your order of preference.

Edited by .:Cyb3rGlitch:.

Share this post


Link to post
Share on other sites

One song with only MP3 and FLAC?

Well, it depends what you're saying. Do you mean to suggest that only for some songs 128kbps MP3 / FLAC makes no difference?

 

How are you intending on making the test blind? I'll assume double blind is a bit hard considering the circumstances.

http://en.wikipedia.org/wiki/ABX_test

 

Rob.

Share this post


Link to post
Share on other sites

One song with only MP3 and FLAC?

Well, it depends what you're saying. Do you mean to suggest that only for some songs 128kbps MP3 / FLAC makes no difference?
Uh. I think if you don't truncate my quote it's fairly clear I was referring to the fact if you only do one song with two formats all your showing is that when you flip a coin, it's likely to come down heads or tails. It's proving that you can tell that there was more than one thing played and have the ability to choose one of them... and that's all. Whether you choose correctly or incorrectly as to what format they are, you prove nothing.

 

Unfortunately this problem also even extends to the 4 songs, 3 formats style I suggested, because the odds aren't long enough that you won't choose correctly by pure random chance. If you choose one correct format on the first song, which you have a 33% chance of doing, then you've got a 50% chance of getting the other two formats for that first song correct. The odds are obviously worse than a simple coin toss, but it's hardly out of the realm of possibility to get it right on three or even all four of the songs just through pure chance.

 

But we've gotta work within realistic boundaries here, and I don't see anyone bothering to go further.

 

How are you intending on making the test blind? I'll assume double blind is a bit hard considering the circumstances.

http://en.wikipedia.org/wiki/ABX_test

 

I'm sorry, I didn't mean what type of test, I meant how are you going to make it blind. As in how are you going to play them without knowing which is which.

 

*EDIT* Ah, I see some programs support the test format and will make it blind for you. Handy.

Edited by tantryl

Share this post


Link to post
Share on other sites

I'm sorry, I didn't mean what type of test, I meant how are you going to make it blind. As in how are you going to play them without knowing which is which.

This is what he'll be using:

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×