FR: Auto Gain Batch Process

[Hexspa]

With the rise of 32-bit float field recorders and interfaces along with podcasts, desk feeds and other 'unmixed' live captures, a plugin that can analyze windows of time, say 20 seconds for a long window and maybe a smaller transient-sized window, to normalize the audio to a certain loudness would be very useful.

Obviously, the current way to do it is to chop up the wav, apply clip gain, and process from there. I don't see why a plugin couldn't write a new file with these essential, broad-strokes, processing built in. This is not something for which I'm urgently clamoring but merely an idea I had whilst eating a vegetarian burrito. If any of you can profit from it then be my guest.

Thanks.

[sirmonkey]

Maybe MLoudnessAnalyzer could fit your needs. It's in the free bundle, and it has a batch auto-gain feature. As described in the manual (page 6).. . " auto-gain button lets you batch process files and make the integrated loudness match the specified Target. You can either click the button and select a file to process, or drag & drop any set of files onto the button. Note that despite the plugin can read several audio file formats, it will always output WAV files."

[DarkStar]

Also, from the Info:

WAV button lets you process a file using the plugin with current settings. You can either click the button and select a file, or drag & drop the file (or multiple files) onto the button. If you let the plugin process WAV files, these will be saved with the original settings. If you use a different file type (such as MP3), the plugin will create WAV files with 32-bit bits-per-sample floating point.

Please note that the files will be overwritten, so make a copy first if you want to keep the original.

[Hexspa]

Thanks for your responses, my friends. While a loudness normalization is possible with that plugin, you still have to chop up the raw capture into segments prior to get an even result.

I guess it's not a big deal but I thought it'd be cool if a plugin could do the same kind of thing. You know, machine learning or whatever. Thanks.

[steve2KVR]

@Hexspa, without your system details it's had to recommend a solution. The various DAWs I'm using have "native" solutions of various sorts, maybe ask around the your DAW's forum to see what you get...hth.../s~

[Hexspa]

There are native solutions for chopping up and 'auto-gaining' long segments of audio? I know reaper has dynamic split... Either way, this was just an idea I thought was unique and maybe fun to implement - I'll just manually chop for now Cheers.

[MeldaProduction]

Actually , MAutoVolume

https://www.meldaproduction.com/MAutoVolume

[Hexspa]

That's true. I actually love MAutoVolume and use it all the darn time. Did a shootout with Waves Vocal Rider and, while the results were different - particularly in the transients - the quality is equivalent which is great because I really liked Vocal Rider.

Cheers.

[vectorwarrior]

That's true. I actually love MAutoVolume and use it all the darn time. Did a shootout with Waves Vocal Rider and, while the results were different - particularly in the transients - the quality is equivalent which is great because I really liked Vocal Rider.

If you're wanting to adjust how autovolume reacts to transients AND have MXXX... autovolume can use the sidechain input, so just send the input through a transient module and the send the output of that to the AutoVolume sidechain input. That way you can increase/reduce the importance of transients to the AutoVolume signal.

This is what I love about MXXX, all the tools and modules are useful for various unusual tasks like this.

[Hexspa]

Wow that sounds good. I need to dive more into MXXX and I'll be revisiting this thread anyway for the other tips as well. Thank you