MP3 vs. WAV

michaelab

michaelab

desafinado
Joined
Jun 19, 2003
Messages
6,403
Reaction score
1
Location
Lisbon, Portugal
http://www5.head-fi.org/forums/showthread.php?s=&postid=742941#post742941

Apologies to "Ross" if you read this forum (I saw your post on this on PFM).

It's worth reading the post above because, quite apart from the interesting conclusions (which my own iPod listening supports) about high bit-rate MP3 vs. uncompressed WAV it shows just how convincingly you can believe there's a difference in the way two things sound when there really isn't one.

Michael.
 
Well Michael, what can I say :D :D :D

I've never believed there to be a huge difference between a well encoded MP3 and the original, although some people will automatically say there is without trying it properly for themselves. Sure theres a difference, if there wasnt, there'd be no need for other codecs (other lossy codecs sound almost completely perfect to 99.9% of audiophiles using even very expensive equipment).

This really backs up what I already knew, and I know I have a rep for having a go at people about MP3 and suchlike when they make blanket statements, but there is a good reason why I do that IMO, because it gets an undeservedly hard time from some people. Sure I can understand those who would rather have the best quality available to them, and I have nothing against this, but when people say that there is a MASSIVE difference between even the highest quality MP3 and CD, I really do have to wonder what they are smoking, or what they are doing wrong.
 
I see trees of green, red roses too
I see them bloom for me and you
And I think to myself what a wonderful world.

I see skies of blue and clouds of white
The bright blessed day, the dark sacred night
And I think to myself what a wonderful world.

The colors of the rainbow so pretty in the sky
Are also on the faces of people going by
I see friends shaking hands saying how do you do
They're really saying I love you.

I hear babies crying, I watch them grow
They'll learn much more than I'll never know
And I think to myself what a wonderful world
Yes I think to myself what a wonderful world.

There is hope for us all when we understand
the power of the Art of Critical Listening.
 
I find the Mp3s versus CD versus other formats argument interesting really, at times I find it hard pressed to tell a high bitrate mp3 apart from a wav file, other times its quite obvious. Not all mp3s are created equal, even using the same bitrate.

If I'm ripping music to my PC, these days I'll use OGG in preference to mp3s. For some cds I've ripped I have found the mp3s sound a bit harsh and tinny compared to the OGG, though for other recordings I've found OGG and mp3 sound very similar (the Ramones anthology recorded to a quality 10 OGG to my ears doesn't sound massively different to say a 160kbit Mp3, where as say, a rip of Dido - Life For Rent, was noticeably worse in mp3 than OGG to my ears).

Some music I listen to really doesn't sound any different at all (128kbit mp3 can sound just as good as say, OGG), drum'n'bass isn't the most musically subtle of things though :D

What its played through makes as much difference to whats better as well, I find mp3s can sound more exciting through my cheap Sony walkman headphones or my cars cd player than the original cd, where as mp3s can at times sound really shallow and lifeless compared to the original on my good hifi or PC (both of which use the superdac), bearing in mind the way the recording and the sort of music seems to make a difference to my ears as well.

Bottom line for why I use OGG when ripping though - Its the closest to the CD sound I've found, and I judge the CD as the benchmark as to what it should sound like...sort of the 'safe' option as I know it shouldn't sound too different from the cd, where as mp3 can sometimes sound worse or sometimes better.

My 2p's worth
 
Originally posted by penance
ffs
we going to get dumb ditties everytime a subject like this appears are we?
Click to expand...


Sorry about that. I over react after reading the link. I missed the whole thread about Eupen thingy. I thought it would be yet another boring cable thread ...... :D
 
LAME seems to be a very high quality encoder for MP3's, a lot of the writer-bundled plug-ins or device embedded encoders might not be. These appear to be the source of the few internet tracks I've tried. If I'd only heard those, indeed I'd be saying "MP3 sucks"...

I originally encoded my CD's at 256Kbps high-quality mode MP3's. I noticed Mozart piano sounded wrong at 192KBps, hence the jump higher. I then adopted lossless APE encoding (650-950Kbps) because I thought I could hear the difference. Well I think on one track, a cymball sounded slghtly less sweet once!

foobar2000 has a few great features for trying this yourself, first it has a converter, which will allow you to write a second copy of a track as MP3 (LAME) or FLAC or others.

You then can run the ReplayGain tagger against them to ensure they play back at the closest possible level to each other, as some encoding can change the level by a few bits of a dB.

Then you can select any two tracks, right -click and choose ABX compare. It then runs a blind test comparison where you can listen to A, B , X, or Y tracks, where they are randomly assigned and you have to say if A=X or if A=Y.

I'm amazed at how hard this is to tell, even at 160Kbps through a relatively well tuned AV/hi-fi system. I do plan to read critical listening notes and re-test as I and a few others I've subjected it to have real trouble telling lots of digital sources apart, incl 24-bit 44.1Khz v.s 24-bit 96KHz upsampled, which really surprised me.

There also now apprears to be an auto-rate setting in LAME which takes the bitrate it needs for a given quality level, which at the 4 default works out to about 140-180Kbps on the few I things I tried, I suspect this is a great compromise for a really close to original MP3 library, at reasonable average bitrates, but a little higher if some material needs it.

I think it still makes sense to have a main media library as lossless in a high-quality hi-fi media-server, as a good, quiet 250GB (Western digital WD2500PB) hard disk is now about £120+VAT. If they were stored as lossey compressed then you'd have a quality drop whenever re-compressing them for any other device in future.

Also as you upgrade your kit around it, you never know if suddenly the lossey MP3 type encoding is limiting the audiable quality, it's one more thing in the chain that doesn't strictly have to be there.

regards,
Rob.
 
Last edited by a moderator:
Very interesting read! I shall have to do some listening before I buy my terrabyte hard drive - maybe I will not need it afterall. Hmmm... more money to spend on a better soundcard!

Dominic
 
Its good to see even just a couple of people taking notice at last. I have been saying this for ages, but would anyone listen to me back then ;) :D :p
 
Recently bought an Iriver 120 for use at work (me being a lazy arse got fed up taking cds in every day :) ). At first I just quite happily ripped to wav mainly for gapless playback. Then when I saw how much size the wavs were taking up I decided to re rip them (from the cd originals naturally). So ended up using Ogg Vorbis at 192. Sounds fine to my ears. Theres a fair bit of background noise, phones going, women yacking (:D ;) ) so having a higher bit rate probably wouldnt make much difference in those circumstances :)
 
Originally posted by PBirkett
Its good to see even just a couple of people taking notice at last. I have been saying this for ages, but would anyone listen to me back then ;) :D :p
Click to expand...

Some people are taking notice because they understood what seems to be logical. If you couldn't tell something apart once the comparison is done by even the most basic blind test the similarity is actually closer then you suspected. There is no sense of shame with this 'new' discovery that you make yourself. The correct response is you go and find out if this could be true. Either you repeat the comparison and read more about the subject. Maybe in the end you even find out that actually the original assumption was correct. This is only the start.
 
Originally posted by wolfgang
Some people are taking notice because they understood what seems to be logical. If you couldn't tell something apart once the comparison is done by even the most basic blind test the similarity is actually closer then you suspected. There is no sense of shame with this 'new' discovery that you make yourself. The correct response is you go and find out if this could be true. Either you repeat the comparison and read more about the subject. Maybe in the end you even find out that actually the original assumption was correct. This is only the start.
Click to expand...

Well, I am not saying there was any shame in this discovery, indeed, if there was then I certainly wouldnt have admitted this to myself long ago. I did a similar test, and found that I struggled to tell the difference, even when previously I felt I could. When I stated this on here, sure a handful of people listened *Kermit*, many more made no further comment (fair enough), but some basically said that there was a HUGE difference, and implied I had crap gear and cloth ears. So forgive me if I am slightly re-assured by the original post above, and am safe in the knowledge that there arent really *huge* differences., but huge imagined differences in many cases (or badly encoded files). People continued to take the piss saying I was some kind of crusader for compressed audio. Not really true, it was more just someone struggling to come to terms with a HUGE difference being inaudible to me. If it was a HUGE difference, then surely I'd hear it.

For the record, I am still saying there are small differences, and I am still saying there are people who will be able to tell the difference fairly reliablly, but what I am also saying is that the differences are subtle, nowhere near as big as imagined, and heavily dependant on the method used to encode them and the equipment used to play them back.

I am sure that this will still not convince some, as they are likely ignorant and will not give this technology a proper chance. It is interesting to see the number of people beginning to see the potential for computer audio on this forum nowadays though. :)
 
It might not be that your system or ears are dull!

I would also assume that a system with good bass extension that is not over bright would make it harder to detect the quantisation differences.

I'd say my previous front Linn Index II's were revealing and slightly bright, so revealed everything I didn't want to hear off most digital sources of the time. Perhaps why I don't hear as much difference now, even though the bigger Linn Keilidhs give more midrange detail.

regards,
Rob.
 
WAV = Exact (or should be) copy of the original CD track.
MP3 = A copy of the original file that is compressed to make it a smaller filesize. Encoder chucks out what it thinks you cant hear, result = easier to stream files over the net, more music on a single disc, etc.
 
Umm.

I have tried a variety of formats, including the original file put to disk and back out to CD. Each time its sounds worse, I can't explain that, but logic I suppose dictates that it has been on a drive jittering to disk, then back out to a cheap CDR jitterting all the way then back onto a CDP with more.

As for MP3, AAC, etc etc, they all sound worse. I suppose for headphone listening or ipod on the go it will make next to sod all difference. However through the dock direct into the hifi it sounds awful, I am confident this is measurable the cut off on bass and treble makes it sound like radio 1.

Not great at all, mabye it depends if you are just listening on the puter or what ever.
 
Pity you cant use Exact Audio Copy on the Mac isnt it....

If it cant even get the WAV file right, what chance does the MP3 have?
 
Originally posted by garyi
but logic I suppose dictates that it has been on a drive jittering to disk, then back out to a cheap CDR jitterting all the way then back onto a CDP with more.
Click to expand...
That might seem logical but it isn't because jitter can't be "stored" either on the hard disk, in memory or on a CDR. Jitter is timebase errors and when digital audio data is stored (on whatever medium) it is independent of any timebase, it's just a string of 1s and 0s.

Jitter is only a factor at the point of digital to analog conversion.

I've never noticed CDRs to sound worse allthough I've never made a direct comparison. I'd find it hard to understand how they could sound worse though unless maybe because of the different nature of the pits in a CDR the laser servo had to work harder and somehow introduced more noise or jitter when reading them :confused:

Michael.
 

Latest posts

Back
Top