AC3 encoding: dialog normalization and preprocessing.

[arnie.d]

Hi,
I use soft encode to encode my waves into ac3.

For dialog normalization I choose -27 or -31. Does this actually affect the audio or is it a "flag" that tells the decoder how to play it? Can it be changed afterwards in the ac3?

To make the ac3 sound as the original wave (normalized in sound forge at
"-16 db speech"), in preprocessing I deselect all the input filters (digital deemphasis, high-pass, low-pass) and for compression characteristics I choose "none". And for dialog normalization I choose -31. Now I was wondering if this is a good method or am I totally screwing up my audio? I can't hear the difference between the wave and the ac3 on my THX 5.1 set.

Oh, I read the guide by SomeJoe in the sticky thread but if I use his method the audio sounds waaaaaaay to soft.

[mic]

There's some good threads & excellent sticky on why to use normalization & what to set it at. How ac3 flags work, particularly in downmix are also covered.

That said, if your stereo source is from a DVD &/or recorded at good levels, lot of folks (myself included) don't go to the bother of normalization.

For input filters some defaults are listed here: http://forum.doom9.org/showthread.php?t=98686

"I can't hear the difference between the wave and the ac3 on my THX 5.1 set."
As ac3 is compression, not enhancement, I would think that's good.