|
|
Post by mark on May 28, 2017 23:31:14 GMT
Just changed to MP3 Gain from the other volume leveler I had for so long.
This is better as it just replaces the files with the new leveled ones where they came from.
But if I set the db at 93 for example when done I wonder why they don't all show 93 db....some show 92.5 or 93.3 etc etc etc...
Why not all the same?
Also shows some tracks as clipping but this isn't heard when listening and some of the tracks with a Y in the clipping column are actually a lesser db than ones that are not indicating clipping. Only a few out of them all have the Y in red.
Anyone have this and what have you found?
Mark
|
|
|
|
Post by Boomer on Jun 1, 2017 5:45:54 GMT
I use two levelers, MPTrim and then MP3 Gain, a 1-2 punch. MPTrim does a rough normalization, I think it only looks at samples every so often in the file and averages those to come up with a gain level and set it.
It works well on modern music, pop or country etc., bur that can cause problems with peaky material, songs with lots of dynamic range, and poems or spoken word, forget about it, unless the material is compressed strongly already. MPTrim can boost dynamic material far too high, it's not unusual to see +4.5 db over the clip threshold.
The clip detector on MP3 Gain seems to be fully accurate; if I normalize to under clipping with Gain, then decode the mp3 file to a wave, it's up to the threshold, but I've never seen clipping.
The reason I run files through MPTrim first and then MP3 Gain, is MPTrim can find corruption and outputs a results file that tells where it is, and it can repair seek timing in files, plus trim files and add fades.
After that it's a trip through MP3 Gain, to see if there any files that clip, and reduce just those by selecting those clipped files, then reducing their gain in 1.5 db steps until they show no clipping.
Someone posted they used hardware processing to normalize on their station I guess. Well, I find that doing all of the non-destructive normalization ahead of time with Trim and Gain does a lot to keep the files in the 'sweet spot' for levels, so processing doesn't have to compress so much or have as much risk of being fooled by sudden loud sounds, so the sound quality on the air is much smoother.
My station can keep the AGC down and let limiting take off from the top where it needs to, and even do it with single band processing without too many audible gain changes on air.
The fact that mp3 files can have their volumes changed in a lossless way through a meta function like this is one of the great features of mp3 that I don't find in other audio file types.
Does anyone know of other files that can be "Gained" in such a way? As of now, I get other file types like Flac, Ogg, m4a, aac, Opus and more, and have to open them in Cool Edit, adjust volume or add fades and other things, then save them as Flac or wav to prevent further loss.
Boomer
|
|
