VBR vary the bitrate given to each part of the song.
When the sound isn't complex, it give a low bitrate to that part of the song, and increase the bitrate where the sound is more complex.
Sometime VBR will not give enough bitrate to some frames and introduce bad sound at low bitrate.
At low bitrate, it's better to use CBR.
Higher bitrate, VBR is most of the time, better.
well, in this case I have 2 ask, what's the threshold:
at what bitrates is VBR better than cbr?
4 instance, 128 kbps is considered the minimum bitrate 4 cd quality music, is it better 2 use VBR in 128 kbps or in higher bitrates like 160 and 192?
VBR will be always better than CBR if and only if your encoder use the latest LAME codec (3.92).
It has been tested at 128 kbps and up.
LAME VBR is very good because it can give 320 kbps to a frame and 32 kbps to another. Some encoders have problems with that.
A real VBR file, the little problem with it, you can't predict exactly which filesize you'll get. If you encode classical music, you may end with a smaller filesize than predicted. Very complicated music with alot of noise like metal could end with a bigger than expected filesize.
At least, you'll get very good sound quality, higher than the equivalent in CBR.
If it's another encoder, the problem I mentionned in my previous post could happen.
It's strongly advised to at least encode over 112 kbps in VBR with poor encoders.
Please use LAME, other encoders mostly sucks.
It will be better for your ears :wink:
thanx 4 the help...
oh, and another question:
what if I take an mp3 file lets say 128 kbps, decode it 2 a wave file and then re-encode it in higer bitrate - will it improve sound quality?
I know that mp3 is a lossy compression, but mp3 also utilizes standard loseless compression methods...???
You can't increase the quality of an MP3 already poorly encoded. The only remedy is to re-encode it.
MP3 is lossy, if you decode in wav you'll get a wav with the information of the 128 bits MP3. Maybe if you find a decoder that can do a bit of dithering you could improve the sound a very little bit, but it doesn't worth it.
When re-encoding the wav, you'll maybe even lose some information instead of gaining better quality. When the guy encoded the wav in MP3 he lost information because the encoder tried to remove things hard to hear and this changed the waveform. Re-encoding the wav file will get you another waveform. Re-encoding in MP3 will change the waveform again, because the encoder will think something is hard to hear but this sound was already modified by the first encoding. At the end, you'll get something that isn't very close to the original. You can't get more than the 128 kbits MP3 had at start. The encoder will not "invent" what is missing.
That's a bad issue, this is why we always should encode in good quality at the beginning.
I think it's better for you to keep the 128 kbits MP3 like it is. You can make experiments if you want too! I don't think it's worth it, but you'll see by yourself!
There's a MP3 software player that do dithering when playing songs. It improves the sound a bit. Try it if you want to. It's Coolplayer <A HREF="http://coolplayer.sourceforge.net/" target="_new">http://coolplayer.sourceforge.net/</A>
This player will give you the best sound quality.
You only need to set it to Cooler Waver Mapper instead of DirectSound and that's it.