Gday..

i have two Q`s.. related to lame.

1. implementing of --twopass
as: "--twopass
Perform two pass processing so that clipping is avoided.
At the first pass, the program converts sampling rate of input file and
write to a temporary file in float numbers while scanning clippings. At
the second pass, the program attenuate the volume so that clipping is
prevented, and write to the output file."

2. in a quick comparrison of two aps..
a. aps
b. aps -k -ms

results: a. 5.4Mb b. 4.5Mb
with no freq. cut. and stereo.. the file should be
bigger.. not smaller.. wazzup..

Maybe the both topics below are related to your first question:

Wavegain vs. MP3Gain

Multipass encoding

Yeh, it's probably just a really compressed track to begin with, so the second pass brings it down maybe 3dB which will result is some size decrease.

Course, that file size change is pretty significant for just a 3dB change... might be something else at work too.

Gday..

well.. the track itself is perfect. gainwise..
no adjustment in anyway needed.
in wavgain form: 0.00 89dB..
Gain | Peak | Scale | New Peak | Track
---------------------------------------------
+0.00 dB | 12204 | 1.00 | 12204 | C:\tools\wavegain\2.wav

in mp3gain: both have the same value : 0.00 89.1dB (assuming 89dB is perfect)

foobar replaygain wav: replaygain_track_gain = -0,000000 dB
replaygain_track_peak = 0,372437

foobar replaygain EX: A: replaygain_track_gain = -0,060000 dB
replaygain_track_peak = 0,430422

foobar replaygain EX: B: replaygain_track_gain = -0,090000 dB
replaygain_track_peak = 0,439272
and yes.. the track is using rather heavy compression.

but again.. whatever value. the aps -k-ms
should have higher bitrate..

a. aps. 5.04Mb

Bitrates:
----------------------------------------------------
32 0.2%
128 0.9%
160 ||||||||||| 14.3%
192 |||||||||||||||||||||||||||||||||||||||| 48.0%
224 |||||||||||||||| 20.1%
256 ||||||| 9.1%
320 |||||| 7.4%
----------------------------------------------------

Type : mpeg 1 layer III
Bitrate : 208
Mode : joint stereo
Frequency : 44100 Hz
Frames : 7779
ID3v2 Size : 0
First Frame Pos : 0
Length : 00:03:23
Max. Reservoir : 511
Av. Reservoir : 66
Emphasis : none
Scalefac : 1.4%
Bad Last Frame : no
Encoder : Lame 3.90

Lame Header:

Quality : 78
Version String : Lame 3.90
Tag Revision : 0
VBR Method : vbr-old / vbr-rh
Lowpass Filter : Unknown
Psycho-acoustic Model : nspsytune
Safe Joint Stereo : yes
nogap (continued) : no
nogap (continuation) : no
ATH Type : 4
ABR Bitrate : Unknown
Noise Shaping : 1
Stereo Mode : joint Stereo
Unwise Settings Used : yes
Input Frequency : 44.1kHz
-------------------------------------------------------
b. aps -k -ms 4.05Mb

Bitrates:
----------------------------------------------------
32 0.2%
40 0.0%
128 ||||||||||||||||||||||||||||||||| 30.7%
160 |||||||||||||||||||||||||||||||||||||||| 36.6%
192 |||||||||||||||||||| 19.2%
224 ||||||| 7.3%
256 |||| 4.2%
320 || 1.9%
----------------------------------------------------

Type : mpeg 1 layer III
Bitrate : 167
Mode : stereo
Frequency : 44100 Hz
Frames : 7779
ID3v2 Size : 0
First Frame Pos : 0
Length : 00:03:23
Max. Reservoir : 511
Av. Reservoir : 68
Emphasis : none
Scalefac : 0.8%
Bad Last Frame : no
Encoder : Lame 3.90

Lame Header:

Quality : 78
Version String : Lame 3.90
Tag Revision : 0
VBR Method : vbr-old / vbr-rh
Lowpass Filter : 19000
Psycho-acoustic Model : nspsytune
Safe Joint Stereo : yes
nogap (continued) : no
nogap (continuation) : no
ATH Type : 4
ABR Bitrate : Unknown
Noise Shaping : 1
Stereo Mode : Stereo
Unwise Settings Used : no
Input Frequency : 44.1kHz

----------------------------------------------------------------------
bitrate: 167 versus 208
am using 3.90.3 dibrom

@n68:

The aps -k -m s version is smaller. Might have any number of causes, but perhaps it happens not to have the same bitrate bloat caused by the workaround for the lack of sfb21 scalefactor for some reason to do with either the lack of filtering or the forced LR stereo mode.

Could you try using:
lame --alt-preset standard -Y
and report the filesize/bitrate?

I'd certainly expect aps -Y to be smaller than aps -k -m s if it's related to sfb21 bloat.

Comparing "aps -k" to "aps -k -m s" might also narrow things down, as might trying "aps -k -m j".

Ciao..

and here`s the results:
(all)

aps: size= 5.04 bitrate= 208

aps -k -ms: size= 4.05 bitrate= 167

aps -Y: size= 3.97 bitrate= 164

aps -k: size= 4.07 bitrate= 168

aps -k -mj: size= 4.07 bitrate= 168

good evening from oslo..

Results as expected - still surprised at the lack of bloat with

OT: Haiku Master

That is so cute! Its the sweetest thing I've seen in yonks

HIHAA..


BTW:
just playing with the lame command`s
is more a mpc guy..
now.... shure i can read me to it..
but if i wanna the most identical mp3..
regardless of bitrate..

stereo and no freq. cutoff.
(i wan`t full range..)
as it states about -k..
"Can cause ringing and twinkling"
is there anything i can do..
to guarantee.. it`s not happening.

what will the line look like..

@n68: Lame -APS isn't tuned for -k, so there's not a lot you can do. I don't know exactly what problems it creates though.

I guess you could try increasing the lowpass to the levels used in insane, which I think are 20.5 kHz. That might work and might only cause bitrate increase. Can't see why you'd need it that high though for CD music, which explains why few people will know the answer.

It does look like the -k just happened to overcome a bitrate bloat (sfb21) issue, possibly by luck on that particular file.

I wonder if the same file, might happen to encode with a lower bitrate bloat, closer to 168 kbps, if you used, say,

lame --alt-preset standard --scale 1.4142
(a constant gain of -3.00 dB in mp3gain would restore it to normal)

or
lame --alt-preset standard --scale 1.2957
(a gain of -2.25 dB would be required to bring it back to 89 dB SPL volume, which is between the 1.5 dB steps of mp3gain)

You might have stumbled on a file that's actually made worse (regarding sfb21 bloat) by encoding it at 89 dB ReplayGain, contrary to the file in the Wavgain vs mp3Gain thread.