[Nocturnal310]

Quote:

Originally Posted by poo /img/forum/go_quote.gif The main difference to me is significant distortion and sound quality that does no justice to the original music... no hint required.

the people who were behind the MP3 project said that.

 

[LostMoogle]

I used to be not able to tell. Then Ifound a song where it made a huge difference. I suppose it also depends on the song too. Sometimes when I rip into lossless and compare it to a 320kbps, I can't tell the difference and other times I can tell immediately. It really depends on the music.

 

[scompton]

While I can't hear the difference, I believe everyone should do their own test. Supposedly, where the difference is most noticeable is cymbal crashes.

 

[jordanross]

I'd have to say I hear the difference as well. As time goes on an my ears "mature" I can hear substantial differences. I ripped a few songs in three different formats: mp3 128K, 320K and I used Apple Lossless. I choose the 320 over the 128 every time, as for the lossless I can hear differences but I think My set up needs some improvement before i'll really hear the difference.

 

[71877]

Looking at the place where this topic is posted, shouldn't we discussing what (as measured and 'showable' findings preferably) the differences are rather than if we can hear the difference, which is rather subjective?

Wasn't there some program that 'visualised' the forms of compression? If so, then we should 'see' the differences, if there are any (which could depend on the record as alexpea said).

 

[scompton]

The developers of lossy formats use psycoacoustics to make the difference as hard to hear as possible, even when there is a noticeable difference in a visualization. There are a lot of things that are measurable but not audible. Blind listening tests are a very acceptable way to tell if something is audible.

 

[uraflit]

depends on music, encoder, audio gear, and individual ears...

all in all, if you use -v 0 LAME MP3, there will only be slight nuances (which are enough for some!) that will be heard using more expensive equipment and with scrutinizing ears.

which is why my minimalistic setup is fine for me =]

 

[Luminette]

your hearing may be flawed, or your ear training may just be very adept and you don't know what to look for / you don't listen to your music that closely

or your system may just not be resolving enough

or some combination of things..




but the differences are there, and many people hear them

Lossless for me.

 

[LnxPrgr3]

Quote:

Originally Posted by BigTony /img/forum/go_quote.gif Remember tha MP3 doesn't just chop the music up, it also 'shapes' it - i.e. to make it more listenable and sound nice (i.e. it adds some eq.).

I can't speak for other encoders, but LAME's only "eq" is a very high frequency rolloff -- at 192kbps, it's set to 19kHz. It is not going to change the overall tonal balance.

Are you thinking of noise shaping? MP3 does shape the distortion it causes, but the goal is to avoid any audible change, not to try to create a pleasing change.

 

[krmathis]

Why not git it a try yourself?
Only you can tell if there are an audible difference, using your audio files, gear, ears, ...

 

[progo]

Haha, on some other forum this would actually be a good topic, but since we're here at the SS troops...

Code:

[left] $ flac -cd ORIGINAL.flac | lame -b 192 - 192.mp3 $ flac -cd ORIGINAL.flac | lame -b 320 - 320.mp3 $ ls -lh total 23M -rw-r--r-- 1 progo progo 3,2M 2008-09-04 15:47 192.mp3 -rw-r--r-- 1 progo progo 5,3M 2008-09-04 15:47 320.mp3 -rw-r--r-- 1 progo progo 14M 2008-09-04 15:46 ORIGINAL.flac $ md5sum * 72a771ad99a4e19b8ba931541b540d7b 192.mp3 03a525704ff07688ad2787fe49f94b31 320.mp3 aacdc64bbc923959c9d73ffe1f97f221 ORIGINAL.flac[/left]

Conclusion: since their hashes differ, the contents of files differ as well. It is thus different.

 

[krmathis]

Quote:

Originally Posted by progo /img/forum/go_quote.gif Conclusion: since their hashes differ, the contents of files differ as well. It is thus different.

True!
But the audio data may be 100% identical, and the MD5 hash may still be different. So MD5 is not useful at all when comparing audio files.

 

[progo]

Quote:

Originally Posted by krmathis /img/forum/go_quote.gif True! But the audio data may be 100% identical, and the MD5 hash may still be different. So MD5 is not useful at all when comparing audio files.

So true. If I'm not mistaken, GoldWave (windows app) can export audio files to text files (having one sample, or two, on a row) .. that could do the job. Too bad Audacity cannot do that.

 

[scompton]

Quote:

Originally Posted by progo /img/forum/go_quote.gif Conclusion: since their hashes differ, the contents of files differ as well. It is thus different.

Of course they're different with different bit rates. It only matters though if you can hear a difference. Doing ABX tests, I can't.

 

[LnxPrgr3]

Quote:

Originally Posted by progo /img/forum/go_quote.gif Haha, on some other forum this would actually be a good topic, but since we're here at the SS troops...

Ah, but you forget the science of psychoacoustics...