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: LAME 3.98b5 old and new method (Read 6250 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

LAME 3.98b5 old and new method

Just tried encoding using LAME 3.98b5 using the recommended -V2 setting, which is mapped at --vbr-new as default.

I tried to use to old --preset standard switch, but the q value is set at 3 as its default.

Just wondering why the deault -V2 (--preset fast standard) has a q value of 0 (ok for higher quality) while the old method is still at 3.  Definitely there would have been a quality disadvantage encoding using just --preset standard.  Why not setting both at -q0 as default?

It seems that the development of this version favors the new algorithm, but why leave the old one with an unfair quality default setting? Does applying the -q0 switch to the old method not improve the sound?

Ok someone might say, "have you tried to ABX?" but I would like to ask the devs why is it so.  Are there any members concerned with this switch, or is it just me?

So far, based on previous experiences, using -q0 can shave off bits, which can compensate for the bloating from using the new method (assuming same quality, but lower bitrate in effect).  Still, why not -q0 also in old...

Just asking...
"Listen to me...
Never take unsolicited advice..."

LAME 3.98b5 old and new method

Reply #1 –
Q0 ~ Q3 are same in new-vbr mode.

LAME 3.98b5 old and new method

Reply #2 –
if Q0 ~ Q3 are same in vbr new, how about in vbr old? is it the same? is Q3 in the old same as Q0 in the new (which i don't think so)
"Listen to me...
Never take unsolicited advice..."

LAME 3.98b5 old and new method

Reply #3 –
If I remember correctly , q0~q3 are not used in vbr-new they are just cosmetic. Gabriel said something about vbr-new doesn't work like old vbr in relation to noise shaping. It has been mentioned in an older thread. in my own tests -q0 / 1 were never worth the encode time - vbr-new speed / quality is too good. BTW- You can always use the old vbr by using Vx --vbr-old

LAME 3.98b5 old and new method

Reply #4 –
Quote:

Gabriel
Feb 24 2006, 20:43

-q3 or -q0 are producing the SAME file with vbr-new.
Under vbr-old or cbr/abr, the difference is the noise shaping amplification mode (ie how do you gradually increase scalefactors to match the masking values). As vbr-new is using a more direct method, this is irrelevant.

http://www.hydrogenaudio.org/forums/lofive...php/t41776.html

LAME 3.98b5 old and new method

Reply #5 –
Would it still be worth it to add -q0 in --preset standard using LAME 3.98b5?

Or they should have displayed -q3 in -V2 since it's also the same as -q0...

Whatever...

EDIT: Thanks to shadowking for the explanation.
"Listen to me...
Never take unsolicited advice..."

LAME 3.98b5 old and new method

Reply #6 –
preset standard in 3.98 is the old vbr. -V2 activates new vbr by default so q3 does nothing and you see q0.

Its like this:

--preset standard = --vbr-old = q3
--preset fast standard = --vbr-new = q0
-V2 = --vbr-new = q0

LAME 3.98b5 old and new method

Reply #7 –
I never cared much about -q x, but as a very high bitrate abr user I now wonder whether I can improve quality as a tendency by using -q 0.
lame3995o -Q1.7 --lowpass 17

LAME 3.98b5 old and new method

Reply #8 –
@halb27,

If my memory serves me right, I added -h (-q2) to abr 320 (not preset 320), and on that particular file it shaved off ~20kbps compared to a file encoded with plain abr 320 setting only.

If ABXing them (which I haven't done) would prove that they would sound identical, then the -q switch (at least -q2 or -h in this case) has a great impact.  Imagine, two files sounding identical, but one is smaller than the other, then it is more efficient.  Conversely, if it is a smaller file and still sounds identical, then it improves the quality by way of quality vs. size ratio.

@shadowking,

Would -q0 be of benefit to --preset standard than its default -q setting? Or there is no need at all?

Like what twostar had posted in his thread regarding "Should future LAME versions only have the -V switch enabled?"

http://www.hydrogenaudio.org/forums/index....=58161&st=0

I would suggest that the -q switch be disabled at least even when using VBR, if the said switch would not improve the sound of VBR files altogether, while maintaining its purpose on ABR and CBR encoding.

EDIT: Added "Conversely..."
"Listen to me...
Never take unsolicited advice..."

LAME 3.98b5 old and new method

Reply #9 –
-q x will do absolutely nothing with the new VBR method. As that is default on 3.98, it will do nothing for --preset standard.

LAME 3.98b5 old and new method

Reply #10 –
Quote from: Firon on
-q x will do absolutely nothing with the new VBR method. As that is default on 3.98, it will do nothing for --preset standard.


Though the default method for 3.98b5 is vbr new, --preset standard uses the old vbr method.  Would -qx have no effect on the old method, in this case --preset standard?

You might have been confused that since vbr new is default, using --preset standard would use the new method.  That would have been --preset fast standard.
"Listen to me...
Never take unsolicited advice..."
 

LAME 3.98b5 old and new method

Reply #11 –
Quote from: Rio on
Quote from: Firon on

-q x will do absolutely nothing with the new VBR method. As that is default on 3.98, it will do nothing for --preset standard.


Though the default method for 3.98b5 is vbr new, --preset standard uses the old vbr method.  Would -qx have no effect on the old method, in this case --preset standard?

You might have been confused that since vbr new is default, using --preset standard would use the new method.  That would have been --preset fast standard.


--preset standard is old vbr, therefore it is affected by -q 0~3