128kb/s mp3s - good enough for portable

Sep 3, 2006 at 9:31 AM Thread Starter Post #1 of 75

astranovus

100+ Head-Fier
Joined
Apr 8, 2006
Posts
356
Likes
10
i just did some abxing with foobar, a bithead and my sennheiser hd650s.

i was also using lame's newest mp3 encoding release.

test file michael jackson's: "who is it"

as i know this track quite well.

i did manage to identify 128kb/s in all cases, it was by no means a walk in the park tough and i had to try hard at times, gone are the times where 128kb/s has wild distortions.

casual listening whilst you are distracted on the run nowadays definitely lends itself to using 128kb/s

i then tried 192kb/s, which i could not abx from cd quality.

as hdd space is a premium on my laptop, 192kb/s is the way to go, with cds as a backup medium.

for me i have decided that 128 kb/s is totally adequate for on the run , as it will save a lot of batterylife.

i would urge other headfiers do a test like this, in order to see what your needs are.
i do generally believe there are a lot of "i have golden ears" head-fiers who will kid themselves into the belief that they need extreme bit-rates

320kb/s on the run is total overkill,

a 5g ipod rockboxed at 128 or 192 will last a lot longer than one loaded with flac,
 
Sep 3, 2006 at 9:37 AM Post #2 of 75
I've always found that when ripping music, 192/Q6 Vorbis provides a very happy medium. I've noticed the drop in quality in 128kbps mp3s even when in the relatively loud listening environment that I'm often in.
 
Sep 3, 2006 at 9:48 AM Post #3 of 75
Quote:

Originally Posted by astranovus
320kb/s on the run is total overkill,

a 5g ipod rockboxed at 128 or 192 will last a lot longer than one loaded with flac



totally agree, not only do i find 320kbs+ indistinguishable on my portables but it certainly drains battery much faster then the simpler alternatives.
 
Sep 3, 2006 at 10:35 AM Post #5 of 75
I beg to differ somewhat. I've found that 128kbps for MP3's made most music I listen to virtually unlistenable. I definitely notice the garbling in the treble on most such files -- even on the run.
 
Sep 3, 2006 at 10:42 AM Post #6 of 75
Quote:

Originally Posted by Eagle_Driver
I beg to differ somewhat. I've found that 128kbps for MP3's made most music I listen to virtually unlistenable. I definitely notice the garbling in the treble on most such files -- even on the run.


mmhh. perhaps the new lame does well with certain kinds of music and worse on others
 
Sep 3, 2006 at 10:51 AM Post #7 of 75
Quote:

Originally Posted by astranovus
mmhh. perhaps the new lame does well with certain kinds of music and worse on others


And in my particular case, I definitely notice a big drop-off in sound quality between 192kbps and 128kbps. (The drop-off in SQ between 320kbps and 192kbps is much less noticeable than between 192kbps and 128kbps.)
 
Sep 3, 2006 at 10:54 AM Post #8 of 75
I too notice a big drop in sound quality when going from 192 or above down to 128, especially with certain kinds of music. I think it has a lot to do with what is going on in the music file at the time so generally I'll encode my MP3s at 192kbps CBR or 192-320kbps VBR0.
 
Sep 3, 2006 at 11:49 AM Post #9 of 75
i notice a difference on my Ipod unamped via my K81DJs from 128 to 192, but only if i listen very carefully. the difference between 192 and above is 99% of the time indistinguishable in my case.
 
Sep 3, 2006 at 12:34 PM Post #10 of 75
for me it's for portable & what's available. if i DL an album from artist sites they're often 192 or 320 mp3 anyway. from CD i want good sound, but also to dabble with rockbox on ipod, so 192 mp3 currently but later, better headphones & possibly an amp & i may goto lossless. sure with the improved lame you can do 128 mp3 but i have plenty of battery, storage & closed headphones, so no point, really
biggrin.gif
 
Sep 3, 2006 at 1:19 PM Post #11 of 75
128kbps might be enough for portable use, but it depends on some factors:
* The encoder. LAME is considered to encode at a higher quality than Fraunhofer, Xing, BladeEnc, etc...
* The music. Complex music like heavy metal need higher bit rates to be transparent, than "general" music.
* Your gear. A high quality portable rig (player, amp, custom mold IEM) show the audio files weakness easier than with a player and stock earphones rig.
* Your hearing. We don't have the same ability to hear higher tones, and the hearing can be trained.

I have performed some ABX tests myself, and my "sweet" point is about 160kbps. Meaning that with lower bit rates I can pick out the lossy one. But with higher bitrates (ex. LAME -V2) the music is transparent in most (90%+) cases.
For simplicity reasons, and too be 100% sure I get the exact bits, I am using Apple Lossless.
 
Sep 3, 2006 at 1:42 PM Post #12 of 75
I don't have a problem with carrying lossless/320kbps music around at the cost of a smaller on-the-go collection. The only problem is that high bitrate/lossless really limits my battery. In a real world situation, my 4G iPod only manages about 6-10 hours of playing time, which I find highly insufficient.
 
Sep 3, 2006 at 2:26 PM Post #13 of 75
Quote:

i did manage to identify 128kb/s in all cases, it was by no means a walk in the park tough and i had to try hard at times, gone are the times where 128kb/s has wild distortions.


Those who take the time to do an ABX test (rather than rely on what they 'know') will find that this is indeed the case more often than not these days when using a state-of-the-art encoder such as LAME and when listened to on a portable player. The truth is that 128 (whole not perfectly transparent in all instances) can be pretty damn good these days in spite of having a bad reputation from days gone by.

You also might want to try using a VBR mode that averages 128 or so vs. CBR at 128. This will allow the encoder to use more bits where it needs to and less where it doesn't, resulting in higher quality in the same filesize.
 
Sep 3, 2006 at 4:12 PM Post #14 of 75
@astranovus, I recommend that you try LAME's -V5 --vbr-new setting instead of 128kbps CBR. The files will average roughly the same size, but the quality of the -V5 files will be significantly higher.

@everyone, it is not at all helpful to comment on the quality of a bitrate without identifying the encoder. A 128kbps MP3 encoded with the xing encoder in 1999 sounds nothing at all like a 128kbps MP3 encoded with a recent version of the LAME encoder.
 

Users who are viewing this thread

Back
Top