freetochoose
28th October 2002, 02:47
I am not sure where to put this thread. But I suppose it will be moved to its rightful place if this is not.
Here are a few notions I have about the Lame encoder. I would like you guys to point out whether I am right or wrong.
(1) Variable bitrate is better than Constant
Obviously, the variable bitrate method gives more bitrate to where is needed. But what is the difference between the new method and the old method. I am using the new method because I think there is not much point in giving out a newer method which is worse than the older one. The old method, however, is the default one. This makes me wonder.
(2) A lower q value
A lower q value means better sound quality. From the FAQ of Lame, I am not sure whether this is achieved exclusively through a better and slower algorithm or the use of bit reservoir. Since q is valid for both constant bitrate and variable bitrate, it seems to me that setting a lower q would not increase the file size.
(3) A lower V value in variable bitrate method
A lower V value implies a higher sound quality. It seems to me that setting a lower V value just makes the encoder stuff more sound (music) into higher bitrate, thus increasing the sound quality at the expense of increasing file size.
(4) Minimum bitrate in variable bitrate method
Lame's FAQ advised to set a minimum bitrate so that it would not compress too aggressively. But if I set V=0, does it make any difference to the sound quality whether the minimum bitrate is 128 or 192 or 256? At V=0, I suppose the encoder will try to put sound into the highest bitrate possible if it is necessary to retain the sound quality. So a higher minimum bitrate is just forcing it to put these sounds into a higher bitrate for no good reason.
(5) Stereo vs. Joint Stereo
If in the joint stereo, both the L+R and L-R are encoded.
Then 0.5 * (L+R + L-R) = L and 0.5 * (L+R - (L-R)) = R
and there should be no loss in the joint stereo mode.
(6) High pass
The default is no high pass, does that mean all the lowest frequencies get encoded? Even those below audible range?
(7) Low pass
I did not set any. But the encoder (Lame) uses a range of
19383 - 19916. Does it make any difference in file size or sound quality if I set it to use 20000 where the cutoff of audible range is?
(8) What is lowpass-width or highpass-width? How to set it meaningfully?
I appreciate any comment.
Here are a few notions I have about the Lame encoder. I would like you guys to point out whether I am right or wrong.
(1) Variable bitrate is better than Constant
Obviously, the variable bitrate method gives more bitrate to where is needed. But what is the difference between the new method and the old method. I am using the new method because I think there is not much point in giving out a newer method which is worse than the older one. The old method, however, is the default one. This makes me wonder.
(2) A lower q value
A lower q value means better sound quality. From the FAQ of Lame, I am not sure whether this is achieved exclusively through a better and slower algorithm or the use of bit reservoir. Since q is valid for both constant bitrate and variable bitrate, it seems to me that setting a lower q would not increase the file size.
(3) A lower V value in variable bitrate method
A lower V value implies a higher sound quality. It seems to me that setting a lower V value just makes the encoder stuff more sound (music) into higher bitrate, thus increasing the sound quality at the expense of increasing file size.
(4) Minimum bitrate in variable bitrate method
Lame's FAQ advised to set a minimum bitrate so that it would not compress too aggressively. But if I set V=0, does it make any difference to the sound quality whether the minimum bitrate is 128 or 192 or 256? At V=0, I suppose the encoder will try to put sound into the highest bitrate possible if it is necessary to retain the sound quality. So a higher minimum bitrate is just forcing it to put these sounds into a higher bitrate for no good reason.
(5) Stereo vs. Joint Stereo
If in the joint stereo, both the L+R and L-R are encoded.
Then 0.5 * (L+R + L-R) = L and 0.5 * (L+R - (L-R)) = R
and there should be no loss in the joint stereo mode.
(6) High pass
The default is no high pass, does that mean all the lowest frequencies get encoded? Even those below audible range?
(7) Low pass
I did not set any. But the encoder (Lame) uses a range of
19383 - 19916. Does it make any difference in file size or sound quality if I set it to use 20000 where the cutoff of audible range is?
(8) What is lowpass-width or highpass-width? How to set it meaningfully?
I appreciate any comment.