Automatically normalize audio?

Jun 15, 2015 at 1:39 AM
Edited Jun 15, 2015 at 1:40 AM
I realize that at present, HandBrake (and therefore VidCoder) doesn't provide for automatic normalization of audio tracks, but I did notice in their changelog that the Linux version does have an option for normalizing when downmixing audio.

Any idea whether this is indicative that maybe they are planning to implement normalization (or "volume leveling", or "automatic replay gain", or whatever else anyone may want to call it) in the near future, and that thereby it will be implemented into the VidCoder interface?

As it stands right now, I have basically only two options using the software I have:
  1. I can extract the audio from my MPEG-2 source file as a WAV file using AVIDemux, and then open and analyze it in CoolEdit to find the current peak level, so I know what value to use in the "Gain" box in VidCoder.
  2. I can open the source file using Vegas Movie Studio, and apply normalization to the audio track, and look at what value it applies.
Option 2 is probably the faster, as it doesn't need to convert the audio to WAV format first, and I can add as many video source files as I want, and normalize them each without needing to close one file before opening another (like with AVIDemux).

But neither option is practical for everyone, because both CoolEdit and Vegas Movie Studio are paid programs, so if someone doesn't already own them, they likely would not want to pay for them just for this. I tried Audacity (since it's free, and similar to CoolEdit), but it provides no analytics to show peak levels. Also, neither option would work at all for DTS audio.
Jun 16, 2015 at 10:04 PM
Normalization and other filters would be a welcome addition to Vidcoder/Hb. It would be an addition familiar to those of us who are moving from Auto Gordian Knot to Vidcoder/Hb. When an alternate audio stream is requested from the default audio stream
the ability to add audio filters such as normalization or an .exe program would add a lot to the program. In my case a use a leveling program for audio which is free called The Levelator. You can read about it here:
It was developed for podcasts which have a lot of different mics and voice levels. But it works with .wav files so I have to demux audio, convert to .wav run Levelator then reverse the process. Normalization and the ability to add filters would get on the
path to doing these more advanced techniques in a batch mode. I've found that live recordings of stage plays benefit most from The Levelator.
Jun 17, 2015 at 5:11 PM
The support doesn't exist in HandBrake. If you'd like them to add it, ask them; they don't necessarily read these forums.