Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Experience with non standard LAME VBR parameters? (Read 960 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Experience with non standard LAME VBR parameters?

Have you guys ever used these non standard VBR LAME 3.100 parameters?
-f -V 0 or -f -V 0 -Y
 
What do you guys think of using the -f switch? Is better than using -q 0 or does it make it sound worst?
I have found using -f switch for -V4-V6 makes it sound a bit better than using the default -q 0 . It also help fix the artifact issue for certain tracks

 




LAME: -f -V 0 -Y
 Xing: -V150 -X2 -U2 -HF-1 -TX0

Re: Experience with non standard LAME VBR parameters?

Reply #1
I've tested and use -f for a while. Generally it improves quality in most samples with tonal artifacts. Impulse heavy sample like EIG are also handled well. Its a bit more pronounced in cbr / abr mode although vbr does also benefit.

I like -V0 -f -Y and --preset cd -f (abr mode 192k) .
wavpack -b2.9cs.5
lame --preset cd -f


Re: Experience with non standard LAME VBR parameters?

Reply #3
Do you know that -f equals -q 7?
Reference:
https://svn.code.sf.net/p/lame/svn/trunk/lame/USAGE
Yep,  f = q7 = q8 = q9.   Parts of the that guide is maybe outdated.  -q9 was the bad one (no psymodel)  but since lame 3.99 its mapped to q7.  -f / q7 uses the psymodel for pre echo and mid-side stereo but different methods for NS.  For cbr/abr noise shaping isnt used. For VBR-new -f,  NS / noise calculation is as a quick guess method from what i understand. For vbr-old its same as cbr/abr except q7-9 are mapped to -q6.
wavpack -b2.9cs.5
lame --preset cd -f

Re: Experience with non standard LAME VBR parameters?

Reply #4
-q n           Internal algorithm quality setting 0..9.
               0 = slowest algorithms, but potentially highest quality
               9 = faster algorithms, very poor quality
            Default is 3. Read the differences between VBR and CBR below.
-h             same as -q2
-f             same as -q7

=======================================================================
Algorithm quality selection
=======================================================================
-q n

Bitrate is of course the main influence on quality. The higher the bitrate,
the higher the quality. But for a given bitrate, we have a choice of algorithms
to determine the best scalefactors and Huffman coding (noise shaping).

For CBR, ABR and --vbr-old modes, the following table applies

-q 0            Use the best algorithms (Best Huffman coding search, full outer
                loop, and the highest precision of several parameters).
-q 1 to -q 4    Similar to -q 0 without the full outer loop and decreasing
                precision of parameters the further from q0. -q 3 is the default
-q 5 and -q 6   Same as -q 7, but enables noise shaping and increases subblock
                gain
-q 7 to -q 9    Same as -f. Very fast, OK quality. Psychoacoustics are used for
                pre-echo and mid/side stereo, but no noise-shaping is done.

For the default VBR mode since LAME 3.98, the following table applies

-q 0 to -q 4    include all features of the other modes and additionally use
                the best search when applying Huffman coding.
-q 5 and -q 6   include all features of -q7, calculate and consider actual
                quantisation noise, and additionally enable subblock gain.
-q 7 to -q 9    This level uses a psymodel but does not calculate quantisation
                noise when encoding: it takes a quick guess.


=======================================================================
Ignore scalefactor band 21
=======================================================================
-Y Ignore noise in sbf21, like CBR mode does

Allows -V2, -V1 and -V0 to not encode the highest frequencies accurately,
if doing so causes disproportional increases in bitrate.
This is the same that CBR and ABR modes do.

Due to the design of the MP3 format, to keep precision in the last scalefactor
band, an encoder needs to increase the precision in all the bands (not only in
this one).
The consequence is an increase of bitrate (+60kbps in some cases) compared to
not keeping that precision. Generally, this band should allow for distortions,
so using this switch shouldn't cause harm.
wavpack -b2.9cs.5
lame --preset cd -f

Re: Experience with non standard LAME VBR parameters?

Reply #5
It seems to me that using the -f switch makes the -Y switch more effective. Like I've notice the file sizes for -f -Y is much smaller and a tad better than the default -q 0 .

-f -V 0 -Y: ~220 kbps vs -q 0 -V 0 -Y: ~245 kbps

LAME: -f -V 0 -Y
 Xing: -V150 -X2 -U2 -HF-1 -TX0

Re: Experience with non standard LAME VBR parameters?

Reply #6
It seems to me that using the -f switch makes the -Y switch more effective. Like I've notice the file sizes for -f -Y is much smaller and a tad better than the default -q 0 .

-f -V 0 -Y: ~220 kbps vs -q 0 -V 0 -Y: ~245 kbps



Yes, It seems that -f behaviour with vbr is that quite part use much less data, while louder use more (than without -f). Good news is that  it doesn't seem to affect quality for me and I didn't manage to abx them on -YV0 or even less.
wavpack -b2.9cs.5
lame --preset cd -f