Any usable 2-pass audio encoders?

[junglemike]

Hello All.
I know this question had been rised couple of times. But I still want to ask.
Is there anything that encodes audio in 2 passes?
It seems very logical to me. Especially for movies sound tracks. Thre are many silence places in the movies and other parts that have very simle sound and can be encoded at much lower bitrate. So those bits could be saved and redestributed to other more complicated parts of the sound track. Something like how Xvid does with video.
Anybody has any info on this?

[Sunhillow]

Hello Mike
you only need 2-pass if you have a particular overall target bitrate, e.g. if you encode a movie and want to burn it on a CDR.
With music I think there is no need for 2-pass encoding. When using LAME or OGG you set a quality parameter and the encoder decides how many bits are needed. Less complex parts need less bits, more complex parts need more bits.

[kotrtim]

Just to answer your question,
Currently only WMA have this kinda 2-pass audio encoding options

I've encoded 6 ch he-aac once, at dialog only = 100 kbps
parts with music = 220+ kbps
total silence = 4 kbps
just use quality mode, the codec will automatically choose the bitrate

[sterlina]

Hello Mike
you only need 2-pass if you have a particular overall target bitrate, e.g. if you encode a movie and want to burn it on a CDR.
With music I think there is no need for 2-pass encoding. When using LAME or OGG you set a quality parameter and the encoder decides how many bits are needed. Less complex parts need less bits, more complex parts need more bits.

that's not a point, and I could reply that in video encoding too you can put a "quality based encoding" and let the codec choose how many bits to use

I think an audio encoder would have a gain if it first explore the source then compress... of course, there are two main problems: programmers should spend time to make it, and size gain is not so much

[Mug Funky]

i think 2-pass would be kinda cool as well. that way a zealous type could shoot for 5000 songs on their iPod and not be disappointed when they can only fit 4863 on there.

it could also aid in finding problem parts of files, like filtered sawtooth waves, artificial formants, applause, etc. they can be given a high bitrate without making the file larger.

however, quantization in audio is a tad more complicated than the current line of video codecs. writing good ratecontrol for such a thing might be difficult (i wouldn't consider WMA as an example because it clearly fails - WMA is unsafe at any bitrate, and adding 2 passes to it is like putting flame decals on a car with no wheels - not too useful).

[Dams]

Hello All.
I know this question had been rised couple of times. But I still want to ask.
Is there anything that encodes audio in 2 passes?
It seems very logical to me. Especially for movies sound tracks. Thre are many silence places in the movies and other parts that have very simle sound and can be encoded at much lower bitrate. So those bits could be saved and redestributed to other more complicated parts of the sound track. Something like how Xvid does with video.
Anybody has any info on this?

Azid do this :

--maximize
----------

Default: omitted

This option will enable a two-pass maximize function of azid. BeSweet
will in the first pass scan the entire file to find the maximum
level. In the second pass the audio will be properly decoded, gaining
it up to 0dB FS.

http://besweet.notrace.dk/azid.txt

[Kurtnoise]

azid is not an encoder...

[Dams]

azid is not an encoder...

Excuse me ... in addition with BeLigth or Besweet of cours' :stupid:

[kotrtim]

maximize is normalize to 100% not 2-pass audio encoding

Nero AAC encoder has 2-pass mode, it is disabled in the release version, only devs have it... I remember reading it somewhere