MP3 vs FLAC

Barzini said:
foo_abx 1.3.4 report
*snip*
Click to expand...

Nice :)

As a query, are you able to pull off the same level of accuracy analysing different sections of the same track?
Out of interest which programme did you create your CBR 320KBps MP3 with?

I'm not saying this to be a pain, I'm saying it so it's down on record.

I'll admit I took my mp3s from the FLAC to test, rather than from the original CD...it shouldn't make any difference but I wonder if it ever does :)
(Actually I also used different sections of the same song)
 
Last edited:
So where are the guys who insinuated that I was full of **** and unwilling/unable to take that test and prove myself? This thread got real silent real quickly lol.
 
Last edited:
probedb said:
Personally I say let them be, they probably also pay £400/m for speaker cable and think What Hi-Fi? is a great, impartial magazine.
Click to expand...

Say what you will about FLAC/ALAC/WAV users but don't suggest we read WHF, it is a shocking excuse for a hifi mag full of AV kit and ipod docks.

I have FLAC on my NAS as space is not an issue and I like the tagging and album art options, WAV is just to big and crude a format to be worth the hassle. I do use 192kbps mp3 for my ipod as I couldn't get my music on in any lossless format as it wouldn't fit. It sounds fine through the relatively cheap (sub £20) headphones I use with it. Don't really want to spend loads on headphones as they live in my pocket/bag and have a relatively short life span.

To my ears FLAC is better though my sytstem, life's to short for arguing on the interweb and doing endless ABX tests. Hifi is a hugely subjective thing, which is whats so great about it.

To those who want to continue arguing about this; step away from the keyboard and go play some music as it really doesn't matter.

Dave
 
Sure is /mu/ in here...

Seriously though, there are audiophiles who can tell the difference. But most people [as in a very large majority] can't. I'm happy with VBR 32-256 and I'll keep it that way.
 
Seriously, some people here are nuts. Like i said, i don't have the time right now to go into such extensive tests. Only when i got round to ripping my CD's this summer into FLAC and Mp3 i did a quick sound test between the two to see which was better. FLAC was more dynamic and clearer. But took up far too much space. Maybe in the future i will get a 128Gb Mp3 player and i will re-rip everything again, But until then i'll stick with mp3 320
 
Robbie G said:
Or asking that you compare different sections of the track like alexrose requested
Click to expand...

I wasn't particularly trying to change the goalposts, just asking a query, as I know thats how I tested myself [I got the impression thats what we were meant to be testing rather than a random selection of tracks], and I wondered if it made any difference. (Different subject matter, some music has a LOT more going on at once!)

As it stands, I've already stated I've failed the 100% test, but think perhaps theres more to it than that, as you'll have both hardware limitations coming into play, different source material, and then perhaps people who won't 100% get it right, but doesn't mean that they can't ever hear any difference, just because they didnt get xyz/zyx, doesn't mean they're always guessing, and it could be on some tracks or portions of a track tested they could easily hear the difference and yet on other bits not, for whatever reason, be it a particular frequency range they just happen to notice more keenly....whatever :)

I'm fairly flexible on this subject.
 
Last edited:
I did a test recently with that, and I listened to a FLAC, a 320kbps mp3 and a 160kbps mp3 of the same song.... I don't have the BEST in the world audio setup but I do have a xonar dx, and some nice head phones as I do like it to sound good. Average ears too, probably damaged over the youthful years of clubbing and going to gigs etc but other than that and being slightly tone deaf I can hear just fine. I couldn't tell the difference AT ALL between any of them.

So I tend to stick to high quality VBRs now although I do have a lot of 160kbps stuff which I'm not that bothered about after that test.
 
Barzini said:
So where are the guys who insinuated that I was full of **** and unwilling/unable to take that test and prove myself? This thread got real silent real quickly lol.
Click to expand...

Most people can't tell the difference, so it wasn't an unreasonable assumption...

Not that I said you were full of poop, but even so.
 
Barzini said:
Thats the same CD, ripped once to FLAC, and then again with LAME to 320kbps, all correctly identified without errors. It was actually a fun test and some were easier to tell than others.
Click to expand...

So you ran one test on each track? Isn't that the 1/1 means? You need to do it rather a lot more than that for each track as otherwise you've got a 50/50 chance of guessing it right anyway.

Did you just use standard LAME settings btw? I.e. what command line switches did you use?

I don't think anyone said you're full of **** either. The fact is most people that post saying they can easily tell the difference refuse to test themselves.
 
probedb said:
So you ran one test on each track? Isn't that the 1/1 means? You need to do it rather a lot more than that for each track as otherwise you've got a 50/50 chance of guessing it right anyway.

Did you just use standard LAME settings btw? I.e. what command line switches did you use?

I don't think anyone said you're full of **** either. The fact is most people that post saying they can easily tell the difference refuse to test themselves.
Click to expand...

Dude, thats 14 different tracks in a row I correctly identified without error. The odds of guessing every single one consecutively are ridiculously unlikely. I ripped with LAME at 320kpbs CBR with the highest quality and best algorithms command line switches.
 
Barzini said:
So where are the guys who insinuated that I was full of **** and unwilling/unable to take that test and prove myself? This thread got real silent real quickly lol.
Click to expand...
OK, I'm one of the guys who insinuated you were full of ****, and I'll happily withdraw said insinuation, if for no other reason than you were happy to step up and put your money where your mouth is.

If those are genuine results (and I've no reason to believe otherwise), then you either have truly phenomenal hearing, or you've found an entire album (I assume it's this one?) on which LAME fails disastrously, in which case the results should be repeatable.

It would be better to run 16 iterations on the same track though - I take your point about the unlikelihood of guessing right on 16 different samples, but sticking to the established ABX protocol will help for consistency's sake when other people try to repeat the results independently.

Just to clarify, how did you rip/encode each track? Are you absolutely sure no gain or other processing was applied during the ripping process? Can you consistently repeat the results with different source material (different albums)?

Lakeland said:
They will probably be along shortly suggesting that you edited the test report ;)
Click to expand...
Well, it's very easy to do and not an unknown occurrence when someone's ego is at stake, but I'm sure we're better than that here, and I certainly wouldn't call someone a liar unless I were very sure of my facts. :)
 
Barzini said:
So where are the guys who insinuated that I was full of **** and unwilling/unable to take that test and prove myself? This thread got real silent real quickly lol.
Click to expand...

meh. your results weren't even worth replying to. i'll let the numbers speak for themselves. :)

Total: 1/1 (50.0%)
Click to expand...

Barzini said:
I ripped with LAME at 320kpbs CBR with the highest quality and best algorithms command line switches.
Click to expand...

say what now? :/ "-b 320" is all you need.
 
Last edited:
I use FLAC (or rather ALAC) on my iPod in the car, but I'm one of those "full of ****" guys with a decent sound system at home and in the car who goes to Sound & Vision shows, etc.
 
marc2003 said:
oh dear lord.... :eek: :D why don't you use a proper lossless codec with tagging support? you can save around 45% disk space and have a bit for bit identical copy. the clue is in the name.... lossless.

i hope you're not as deluded as this muppet....
Click to expand...

:D I was trolling.

I use Apple lossless for my CDs.
 
Durzel said:
I use FLAC (or rather ALAC) on my iPod in the car, but I'm one of those "full of ****" guys with a decent sound system at home and in the car who goes to Sound & Vision shows, etc.
Click to expand...

I was wondering to myself whether an iPod even has the fidelity to play back a FLAC (ALAC, whatever) to it's full potential. Plenty of the iPod decoders have come under fire in the past for poor sound quality.
 
marc2003 said:
anyway, the point is moot. i know no one is going to bother even attempting the test.
Click to expand...

This would be true, I got to nine before getting sick of listening to the same track over and over and decided I was doing this just to prove to some faceless internet person that I could hear a slight difference.

In fairness, though you can choose not to believe me, I identified the track correctly for those nine trials and I was confident that I could accurately discern between the formats. Was there a large difference, no, and I would not have been able to notice it had I not been actively listening for it.

Originally I had prepared two tracks for the test, "Un bel di vedremo" from Madam Butterfly and "Mountains Beyond Mountains" by Arcade Fire and I went into the test fully epecting to be able to identify the different formats for the Madam Butterfly track but not for the Arcade Fire track. I ran the nine tests using "Mountains Beyond Mountains" and was very surprised.
 
Barzini said:
Dude, thats 14 different tracks in a row I correctly identified without error. The odds of guessing every single one consecutively are ridiculously unlikely. I ripped with LAME at 320kpbs CBR with the highest quality and best algorithms command line switches.
Click to expand...

You've not understood the purpose of ABX then. You don't ABX on 14 different tracks, once on each. I'll give you credit for attempting it but you really need to do it properly to give the results any meaning.

You still haven't said which switches or which version of LAME. Saying 'best ever mega settings' isn't helpful or what I asked for clarification of.

I'm just giving you the same response/aggro you'd get over on hydrogen audio ;) Actually, you'd get far worse over there.
 
You must log in or register to reply here.
Back
Top Bottom