[MEncoder-users] Problems with 'volnorm' filter when the 'method' parameter is set to '2'

[Jonathan Busby]

RC wrote:
> On Wed, 01 Apr 2009 18:38:56 -0500
> Jonathan Busby <jonathanbusby at gmail.com> wrote:
>
>   
>> after initiating playback it sounds as though the 'volnorm' filter has
>> not kicked in -- ie. the volume sounds as if the 'volnorm=2' filter
>> has not been used -- after which the volume level will suddenly
>> increase and usually remain at said increased level throughout the
>> remainder of the playback. 
>>     
>
> That's not a bug, that's how volnorm=2 works, and it's precisely the
> difference between 1 and 2.  What part of the documentation lead you to
> believe otherwise?
>
>   
The man page :

"
volnorm[=method:target]
Maximizes the volume without distorting the sound.
<method>
Sets the used method.
1: Use a single sample to smooth the variations via
the standard weighted mean over past samples (de-
fault).
2: Use several samples to smooth the variations via
the standard weighted mean over past samples.
"

I the above I take the usage of "samples" to mean PCM values. If instead it
means "audio packets" then what you said might hold water, but the delay
between initiating playback and the volume increase would still be at
most a second. When I play a certain MPEG-2/AC3 source video
*it takes __15-20__ seconds* before the volume increases with 'volnorm=2'
and it happens while someone is speaking -- their voice suddenly
increases in a poppy, stuttering manner. So you're telling me that this
is "not a bug" and that it "Maximizes the volume without **distorting**
the sound"? If what I just described isn't an instance of the sound being
distorted then I don't know what is.



Jonathan