Feature Request: Loudness Meta Normalizer - Solutions for short files (samples)

WaveLab
1

I’m trying to get Loudness Meta Normalizer to work well for shorter files - Samples, Video-Game Assets, GUI SFX, etc in batch processing.

I haven’t had much success.

One trick I use when I’m doing this work manually, is I’ll loop a sample and playback to get a loudness reading

Could such a strategy be implemented here, where the user enters # of passes over # of time of each file to get a proper loudness reading?

So far for me, it appears short one shot samples are the exact same volume as they were before processing.

2

R128 Loudness requires several seconds of audio. This was originally conceived to measure full songs, not short samples. You can use tricks as you mentioned, but that remains tricks and nothing “official”.

1 Like
3

I do like my tricks…

… But I’ll always take automation when it comes to menial tasks.

4

Perhaps a different measurement method could be utilized rather than loudness?

Like a combination of fletcher-Munson and spectral analysis of frequency loudness?

It would be a dream come true if I could level out my samples enmasse especially after doing other post-processing on them.

5

Mentioning Fletcher-Munson curves. Are there any EQs, which have this feature?

6

In my humble opinion, WL might benefit from a ‘Loudness Normalize to RMS’ function. Although not as accurate as LUFS loudness measurements, RMS can be used to process shorter audio files, which would probably have suited your purpose.

In the meantime, the Loudness Normalization plug-in in Audacity might be worth a try. This features an RMS option:

image
image449×320 13.2 KB

This can also be set up for batch processing.

7

I do recall one or two, but I think it was n frequency spectrum EQ matcher (source → Fletcher-Munson) which isn’t really what I’m looking for.

RMS could work,

interesting thanks for the Audacity hint, I’ll consider that.

1 Like