There is a lot of inaccurate information in this thread.
1. "Joint stereo is a no-no"
FALSE. This is a persistent myth that has been debunked many times.
There are two different things that are sometimes referred to as "joint stereo." The "bad" joint stereo is referred to as "intensity stereo," and it is implemented (at least by LAME) only in very low bitrate files. Intensity stereo joins the two channels to mono when they present similar information.
The type of joint stereo that is used in higher bitrate MP3s is called "mid/side stereo." Mid/side stereo-coding calculates a "mid"-channel by addition of left and right channel (l+r)/2 and a "side"-channel (l-r)/2. When the information in the left and right channels is the same or similar, mid/side stereo can use a lower bitrate to encode the side channel, so that the overall bitrate will be less than it would be if encoding the left and right channel separately. Mid/side joint stereo does not join the two channels to mono. Instead, it uses a more efficient encoding method to capture stereo information.
Intensity stereo destroys phase information. Mid/side stereo, on the other hand, keeps the phase information pretty much intact. Correctly implemented, mid/side stereo does very little or no damage to the stereo-image and increases compression efficiency either by reducing size or increasing overall quality.
It is also worth noting that the MP3 codec handles LR stereo or MS stereo on a frame by frame basis, so LAME can uses LR stereo on those frames that require it, but use MS stereo on those frames where MS stereo can save bits.
2. CBR is better than VBR.
False. Except for 320kbps CBR, a well-implemented VBR is always going to be of higher quality than a similarly-sized CBR file. Take as an example a file that is encoded with LAME using the -V2 preset (which used to be known as --alt -preset standard). Files encoded with this setting will average somewhere in the neighborhood of 192kbps. There is no constraint on the upper or lower encoding limits with this setting, so MP3s encoded at -V2 can contain frames that are anywhere from 32kbps to 320kbps.
It is true that a 192kbps CBR file will never go below 192kbps, whereas a -V2 VBR file may go down to 32kbps (actually, I think that 32kbps is used only for silence). But it is also true that
a 192kbps CBR file will never go above 192kbps, while a file encoded with the -V2 preset will use frames of up to 320kbps if that is necessary for a particular part of the music.
The Hydrogen Audio knowledge base has this to say about CBR:
Quote:
Note: The rule of thumb when considering encoding options: at a given bitrate, VBR is higher quality than ABR, which is higher quality than CBR (VBR > ABR > CBR in terms of quality). The exception to this is when you choose the highest possible CBR bitrate, which is 320 kbps (-b 320 = --alt-preset insane), but this produces very large filesizes for very little audible benefit.
CBR: constant bitrate mode. CBR encoding is not efficient. Whereas VBR and ABR modes can supply more bits to complex music passages and save bits on simpler ones, CBR encodes every frame at the same bitrate. CBR is only recommended for usage in streaming situations where the upper bitrate must be strictly enforced. |
Source: http://wiki.hydrogenaudio.org/index....commended_LAME
That H.A. page is a good read for anyone trying to decide what settings to use for MP3.
3. Transcoding from a lower bitrate to a higher bitrate to improve quality. This thread demonstrates that transcoding from a lower bitrate to a higher bitrate actually makes files sound worse, not better.
