Auto Split - Audio is missing

I have to Auto-Split a lot of Audio-Files for audiobooks. But I don’t want to shorten the files. I tried every option (silence at the end, silence at start, no silence) but at the end it differs to the original files. 11 hours of audio, Im missing 4 minutes. How I can auto split audio without loosing silence/audio?
Ohne Titel.pdf (190.1 KB)


What kind of audio file wav, mp3 e.t.c ?

regards S-EH

1 Like

How do you split? there are so many possibilities.

I just did a quick test: When you use the method “Split at Silence”, Wavelab (I’m using Pro 11.2 on Windows) seems to ignore silence at the beginning of a file when set to “Keep Silence After” or silence at the end when set to “Keep Silence Before” - and “Exclude Silence Parts” should be obvious. I didn’t change the detector settings.

As a workaround, you can use “Split at Silence” with your preferred settings (even “Excude Silence Parts”), set the destination to “Create Markers at Split Points (No Split)” and use “Generic Marker” as marker type, so Wavelab just marks the cuts for now. (“Temporary”, “CD Track Index” and “Playback” markers will also work, but if you set any kind of region marker, Wavelab will not include the silence parts later).
After that, you execute Auto-Split again, but instead of “Split at Silence” select the method “Split According to Markers” and choose the type of marker from before, then you can save as separate files or create a montage, and any silence should be retained.


I just use “split at silences” set up my duration before a split & the minimum silences & the defined silence (changes per job and bg roomtone level), I don’t use the automatic level dectection,
use exclude silent parts & remove silent parts and it does the job, never typically see a differing duration

occasionally I’ll get it where the audio all bunches up and goes awry, normally a WL restart solves that

1 Like

one thing I do see however, is that if I render out final mp3’s & wavs you will see a differing duration on the mp3s, which has nothing to do with strip silencing at all,
I can’t exactly remember why something to do with mp3 headers, PG mentioned the reason somewhere, and its buried somewhere on this forum, its one of those quirks to be aware of for long form audiobook editing & podcasts but I workaround that issue by batch converting to mp3 by RX/reaper or whatever, depending on what’s being used by the QC checking and haven’t had an issue for years doing it that way