Normalize an AC3 File!

[milOtis]

I have been pulling out my hair trying to figure this one out.

I have two files...a demuxed video (one .264 video file, and one .ac3 audio file). I know how to mux them back together, but before I do that, I need to normalize the AC3 file (the volume is too low).

I want to keep the video stream as it is. All I need to do is normalize the 6-ch AC3 audio file. I have tried every AC3 Normalize tool on this site, and none do this! I do not want to convert to a different format....just AC3 --> to --> Normalized AC3 (keeping same number of channels, bitrate, frequency, etc).

Is there a tool to do this easily, so I can mux these two files back together and be done with this finally!?

[WWM_UK]

Do you still have the original file? If so, you could use something like "eac3to" to normalise the audio while it demuxes.
So an example for an MKV would be something like:

eac3to 1: video.h264 2: audio.ac3 -normalize

This will demux both tracks and do a second pass normalising the ac3.

[milOtis]

Do you still have the original file? If so, you could use something like "eac3to" to normalise the audio while it demuxes.
So an example for an MKV would be something like:

eac3to 1: video.h264 2: audio.ac3 -normalize

This will demux both tracks and do a second pass normalising the ac3.

The original file I have is a TS file. My ultimate goal is to put it into an MKV container with normalized audio, and no re-encoding of the video stream. What would the command line be if I used the original file?

-edit-
I figured it out, and used this as my command line...
eac3to original_file.ts 1: video.h264 2: audio.ac3 -normalize

It is working on it now. But I notice it is creating a WAV file...? I want an AC3 file, like it was originally, so I can remux with MKVMerge after. Will it remake it into an AC3 file on pass2? The WAV file is already up to 8GB!

[WWM_UK]

Yes, the final file will be AC3.

EDIT: I just realised, I forgot to put the 'input file' in my earlier command line, doh, well it was 8:20, hadn't had enough coffee by that time........, glad you 'sussed' it out.

[milOtis]

Yes, the final file will be AC3.

EDIT: I just realised, I forgot to put the 'input file' in my earlier command line, doh, well it was 8:20, hadn't had enough coffee by that time........, glad you 'sussed' it out.

Thanks! You've been a big help! I knew that there had to be some tool to do this, and it worked perfectly. I even used GDSMux , a small tool in the Haali (Matroska) Media Splitter folder. I used the original TS file for the video, and the new AC3 file for the audio! It's perfect. The video that eac3to put out is not usable actually, so glad I still have the original.

[PuzZLeR]

I have a question if I may. Can this process (normalization), or what eac3to does, be done losslessly on the AC3?

[WWM_UK]

I have a question if I may. Can this process (normalization), or what eac3to does, be done losslessly on the AC3?

What makes you think it isn't?

Not that I know the 'inner workings' exactly of eac3to but the 'wav' file it produces is just a 'pass1' file, on pass2 it re-writes the 'ac3'. Apart from the obvious changes brought about by 'normalisation' the file is the same.... channels, bitrate, samplerate, e.t.c.

[PuzZLeR]

What makes you think it isn't?

Truly, I personally didn't think it was lossless but I asked because apps like MP3Gain promise no loss to the MP3 it normalizes, or raises the volume on.

I was wondering if this "promise" can be fulfilled by an app on AC3 since I'd love to know of one.

[jagabo]

Are you sure your AC3 audio is lower than normal? AC3 is normally recorded at about -20 dB to leave room for big peaks (explosions, thunder, gunshots, etc.). All you usually have to do is turn up the gain on your AC3 decoder.

[PuzZLeR]

Hi Jagabo. Yes, AC3 is recorded more "movie friendly" unlike MP3.

The reason is that I have many different sources for TV captures, and many are music videos, and many that I've captured with NeroVision (for a specific firewire tool) are in a lower-than-normal AC3 volume IMO. There is no setting in NeroVision to adjust this for capture either AFAIK.

What I'd like to do is have all my content at a similar volume when feeding into a playlist. I can easily adjust the volume manually for each during playback to get a desired sound, but it's a bit of a pain. I was hoping there's a lossless way to do this for each file.