Page 1 of 1

Post process audio levels?

PostPosted: Tue Feb 23, 2021 4:03 am
by _RichG_
I have a large number of mp4 files with varying levels of audio. Is there a way to batch process all the files to a normalized level using DaVinci Resolve?

I'm hoping to do so without having to import the video, normalize the audio, then export the video.

If it can't be done with Resolve, is there something else that can do it?

Re: Post process audio levels?

PostPosted: Tue Feb 23, 2021 4:33 pm
by Jim Simon
Select all the clips in the timeline that you wish affected, right click>Normalize Audio Levels...

Re: Post process audio levels?

PostPosted: Wed Feb 24, 2021 12:05 am
by _RichG_
Jim Simon wrote:Select all the clips in the timeline that you wish affected, right click>Normalize Audio Levels...


To get to that state, I'm assuming I have to create a project, import the media, drag it to the timeline, then do the Normalize Audio... then export the project.

With hundreds of videos to process, this becomes a very time consuming task, especially if the video has to be re-rendered. I'd prefer to avoid any video processing. I'm looking to modify the audio levels only.

Re: Post process audio levels?

PostPosted: Wed Feb 24, 2021 3:03 pm
by Jim Simon
Well, you can't do anything in Resolve without starting a project, importing the media and adding it to a timeline.

That said, Resolve can't do what you want here.

I don't know of any software that can.

Re: Post process audio levels?

PostPosted: Wed Feb 24, 2021 5:34 pm
by ttakala
ffmpeg is a very powerful open source command line utility which seems to be able to do what you want. I got promising hits with a google search "ffmpeg normalize audio level".

Re: Post process audio levels?

PostPosted: Wed Feb 24, 2021 9:07 pm
by Andrew Kolakowski
It still has to rewrite new files, even if video part is just a copy of original data.
With ffmpeg all of it can be batch processed. You need to know some scripting language though or use one of the many GUIs. You just need to find one which works for your case.

Re: Post process audio levels?

PostPosted: Wed Feb 24, 2021 11:40 pm
by Marc Wielage
My opinion: this is what a sound mixer does. In other words, this is a skill and a process, not a plug-in or a button. You have to listen -- preferably on great speakers in a good room with good acoustics -- and monitor at a set reference level. That's how to achieve the right balance between different voices, music, and sound effects. Cheap speakers won't do it, and headphones won't do it.

There is no automatic way. I can refer you many books and courses on sound mixing, but be warned it's not a simple subject. There are good discussions about final mixing on the Gearslutz Forum's post group:

https://www.gearslutz.com/board/post-production-forum/

The discussions on reference level are the ones I'd start with. They have a ton of detail about this, and the principles of sound mixing apply to any software.

Re: Post process audio levels?

PostPosted: Thu Feb 25, 2021 12:23 am
by Andrew Kolakowski
What you said has about nothing to do with original problem.
It's about normalising many files to the same loudness and you don't need any audio guy who charges 1K$ a day for it. It can be normalised to eg. R128 automatically. How will it sound is a different matter - all depends on the original recording. If it was poor it still will be poor and if it was good it still will be good.
It's probably crap recording, so even audio guy with experience and gear won't make it much better.

Re: Post process audio levels?

PostPosted: Thu Feb 25, 2021 3:44 am
by _RichG_
Andrew Kolakowski wrote:It's about normalising many files to the same loudness and you don't need any audio guy who charges 1K$ a day for it. It can be normalised to eg. R128 automatically. How will it sound is a different matter - all depends on the original recording. If it was poor it still will be poor and if it was good it still will be good.

Exactly this. Without going into detail, I have a good understanding audio quality. In this case, the quality is what it is. The goal here is to make the relative audio levels between videos equal enough so the viewer doesn't have to adjust the volume level on their remote for each video.

It's probably crap recording, so even audio guy with experience and gear won't make it much better.
Yes, we're talking home video recorded to VHS then converted to digital format. Those old camcorders didn't handle high SPL very well.

I wasn't aware of ffmpeg. That might be just the thing I need.

I also found a utility called MP4 Gain which may also work.

Re: Post process audio levels?

PostPosted: Thu Feb 25, 2021 10:28 am
by Andrew Kolakowski
Anything with batch processing and eg. R128 normaliser will do. Adobe Media Encoder also has it and it may even pass video bit as copy ( not sure if it does it for h264). Ffmpeg based solution can pass video as is for sure and just adjust audio.
You have this as well
http://r128gain.sourceforge.net/