Hi,

Since lame version 3.89 i´ve used the -V1 -q1 -mj switch, and i´d like
to hear if other people think it's still up-to-date. I've been very happy
to the results and it still gives (lame version 3.96) great sound quality in my hifi-system and to
my ears where for example --alt-preset standard fails.

I recently got a minidisc-recording of my favourite live-band, and i tried
--alt-preset standard to compressing it. It didn't give good results, the highest
frequencies (around 14khz) being totally different sounding than in the original flac-files.
However, the familiar settings gave me again good results, and the mp3s were
transparent to the original. This was weird to me, since the difference in file
sizes was only about 10% and the difference in sound was huge and existed in all minidisc files i encoded. Could this
have something to do with minidisc compression artifacts in original files
that completely fooled the --alt preset standard switch?

Do other peole think -V1 -q1 -mj is still generally the best possible in sound quality and in its bitrate class (around 220-250kbs?).

-bluewater

Can you ABX the --aps mp3 against the original? If it's as bad as you say, 8 trials should be enough to establish this.

Note that -V1 is a higher "quality" than --aps, which IIRC is -V2. However there are few if any known samples that are detectably different between -V2 and even -V0.

The canonical wisdom on the forums is that you should stick to the -V options and avoid tweaking with -q and -m, based on a number of ABX tests. However, not much testing is done on -V1.

I did ABX tests and they turned in favor of --aps as you suspected,
the "huge" difference was between the -v1 -qx -mx and -v2
encoded files, i thought that the settings familiar to me were the ones closer to original
but it is infact the other way round. So I got the answer to my own question and
it seems i´ll use --aps in the future. Thanks.

-bluewater

No problem. Thanks a lot for taking the time to look into this.

Can this thread be made 'must read' for all new members?

--Rigapada

Yup, everyone that suspects that they can hear a difference should be made to read this thread.

That´s really cool. I need transparent presets and I have nothing against making this a good example for beginners. Sometimes simple logic doesn´t apply to the connection between vbr-presets and human ears because someone unaware of the history of --aps and abx- tests can think that

" -q3 can´t be as good as -q2 or -q1"
" -V1 can be better than -V2 or at least it can´t be worse"

just by looking at the verbose - window of lame when it´s doing --aps. And i was
one of those.

-bluewater

@bluewater:

It seems to me that people willing to admit they were wrong are few and far between nowadays. Especially with the grace and poise that you showed.

You are a classy person, and I salute you.

Regards,
Madrigal