mp3: 128 bitrate vs. 192 bitrate - plz help!

[nehalmistry]

Hi,

i have been playing with encoders and i have tried converting some of my 192's to 128 (off course keeping the original 192) and I could NOT notice a single difference, using my awe32 and sony in-ear headphones... I also tried at my friends house with sb live and yamaha speakers... i couldnt notice anything....

i also tried converting to ogg vorbis format and couldnt notice a difference (although i just tried it out recently)...

does it really matter if it's 192 or 128? or i am i just partially deaf? (both possibilities btw)

I would really appreciate it if someone could post/send me a small sample of the same mp3 in both 192 and 128 where it makes a difference... if i see a difference, i will convert the 192 to 128 bitrate using dBpoweramp(LAME) and see if i *still* hear a difference...

[Greebe]

Are you compression limiting it?

Are you recording both from the original source? (I take it no, but I still need to know)

[Indiana]

Are you really saying that you're taking a 192 KBps mp3 and REcompressing this into a 128KBps one and still don't hear a difference?!? Then you're partially deaf - or maybe this can be attributed to your Creative soundcard or the sony headphones (they have some nice ones but also some VERY crappy).

Taking an original WAV ripped from CD, there are some songs where you might not hear much of a difference between 128 and 192. But then there are other songs where you hear the difference even between 128 and 160 KBps. And decompressing an mp3 and doing another mp3-compression on this will results in degraded quality, no matter if you're recompressing a 192 KBps file into a 128KBps one or even a 256KBps. It's like doing jpeg compression on a file multiple times.

And some encoders (esp. some Fraunhofer revisions) seem to be optimized for 128KBps and not optimally use higher bitrates so with a crappy encoder you maybe won't hear a difference but with a good one you should.

P.S. for the "partially deaf" comment: Sometimes people that are in some way hearing impaired do hear the mp3 comression artifacts more than those with normal hearing. This is because mp3's psychoacoustic models try to hide those artifacts under other sound parts, but they were of course made for the normal human ear. If you have some deficiencies at certain frequencies this might not work and you'll clearly hear the aliasing / ringing artifacts that are meant to be hidden.

[nehalmistry]

greebe:
i have used both downloaded from internet and cd one's... off couse u cant make a comparison to CD since the AWE 32 has a treble/bass amplifier which actually make the mp3's sound better (which off course they're not)... btw, i have played with both bass/treble controls.... i usually keep them both to max still i hear no difference..

and no im not doing any compression/limiting at all.... that just completely changes the sound and is no good for comparing audio quality

Indiana:
woah, i guess you haven't heard the AWE sound cards... they were the first (or at least one of the first) sound cards to be *true* cd quality....... if you compare it to a CMI or Yamaha sound card (which i also have, but not installed currently) there is so much difference it's not even funny....

and the sony phones i have, they are the excellent kind

PS, LAME is an excellent encoder, if you recommend me to another encoder, i will try converting one of my CD tracks using both bitrates and then compare


well, since comverting a mp3 to mp3 might not be the best idea, i have done this: i took one of my CD's out, and converted a song straight from cd into both 192 and 128.... i STILL hear no difference,

[Liquid Snake]

It depends on the sample you're encoding. Some samples MP3 can handle easily, so 128 doesn't sound any different from 192. Check http://www.r3mix.net, you can find some samples posted there or in the forums there.

[nehalmistry]

liquid snake: i cant find, can you point me to a more specific URL thx appreciate it

[Indiana]

Nehalmistry look at the "quality" and "analysis" chapters on the link LS provided (excellent read, one of the best sites for mp3-compression by the way).

And yes, Lame is the BEST mp3 encoder. Still you get aliasing in the trebles with 128 KBps that are hearable on standard equipment in SOME songs (while others are near to perfect), going to 192 KBps should really give CD-quality on all songs for nearly everyone that doesn't have highest end hifi equipment and is trained to search for such small differences.

Beware that in some cases mp3-encodings can actually sound "better" at the first hearings because some "harshness" and other imperfect things in the orinial are filtered out. Again a bit like (M)JPEG, where the filtering can smooth some sharp edges and also squishes the luminancy range - thus giving a "better" / more contrast look.

[nehalmistry]

i dont care about analysis and stuff like that unless the graphs can actually be representative of what I actually hear, which it doesn't.

I will not believe there is a *noticeable* difference (at least for me) unless i actually hear a sample.

What i am trying to do is prove that there is an audible difference between the two, and the only way to do that is to use mp3 samples, not graphs. If I cannot prove that, then i will remain at the conclusion that 128 and 192 are the same (Both CD quality) and maybe i will convert all my 192 mp3's, (probably about 1/3 of my total 600) to 128 one day

[TnT]

What type of music are you listening too?

[nehalmistry]

haha, i had a feeling someone would ask that, i listen to rap, hiphop, and r&b (basically 'urban' music), some of the newer rap has a great mix of high treble and low bass frequencies (JD, Ludacris, Jay-z to name a few)...

i have tried classical and pop music too.... it still sounds the same...

[JLP]

Well take some CDs and encode them into 128 and 192. Then lsiten to the MP3s and you will se if you can hear the difference. I know I hear the difference and it's quite a big difference. also try to encode the CD with --r3mix setting which is very wll optimised. Also if you can use variable bitrate and not fixed.

[Wombat]

You should definitely hear the difference encoding classical music. It's the biggest victim of MP3 artifacting.

I do most of my rips at 160kbps. The others I do at 192kbps. The difference in sound between 128 and 160 is quite noticable for me. The difference between 160 & 192 is there for me too, but the whole point of MP3 is compression. If I really want quality sound, then I'll listen to the original CD.

[nehalmistry]

wombat: what sound card/speaker combo you got? and what kind of classical you listen to, name an artist... ill see if i have CD

i heard a difference in 96kbps, and at 112, i could only notice a difference in some songs, and i had to listen extremely carefully to notice even the slightest difference, but i wanna be safe, i will stick to 128...

[JLP]

Just one more thing before you decide to downgrade you MP3s to 128 bps. Think about the future. If you have a lot of MP3s and are planing to buy a new soundcard in a year or two, then you would be sorry you downgraded MP3s. Also if you share your MP3s to other users on the internet leave them at 192 as a lot of users have better soundcards and have them connected to hifi systems and there is a lot of difiference if you listen to MP3s on that.

[Maggi]

Static bitrates are crap ...

I recently switched over to use variable bitrates (VBR) at highest quality setting and the file size is comparable to 192kbit/s, but the sound is awesome !!!

If you're doing comparisisons between various bitrates, have a close hear to bass and treble signals, cos that's were they first start capping the quality, ie. a decent bass gets converted to annoying 'honks' or even 'mooohs' and filigrane trebles get indifferent, scratchy and hissy (imho).

My VBRs clearly draw a 'sound picture' while even at static 320kbit/s, my MP3s lack those differences, but rather 'reproduce' single instruments on their own, without 'showing' you the whole orchestra at once ... (pretty subjective view, I know)