Really over 300 dB? WOW! Sounds like the end of the world…
A lot depends on what the audio is how best to proceed. There are ways to do voices that are different from what you can do for music. Can you tell us what the audio is? Thanks!
Limiters are designed for this purpose—for example, WaveLab’s Peak Master or a brickwall limiter.
But a peak of 517 dB—what is its origin? When you record audio from an analog source, the A/D converter caps the level at 0 dB maximum, since the signal must fit, for example, within a 24-bit range.
Alternatively, could this be generated by a software instrument? In that case, the instrument may have a bug.
Or is the issue a corrupted file? If so, manually restoring manually these “peaks” would be the appropriate solution (WaveLab has tools for this).
recording of combustion with special equipment. recording 32 Bit, Mix-Pre II.
There are some electric bursts, that causes exactly 1 peak at a time.
So, is it possible to (bulk-)delete selected datapoints above a certain dB-level or not?
No interpolation, no FFT-stuff here, no nothing, just a plain delete of a datapoint (or several, i.e. marked for instance).
(Yes, the sound file will become shortened, of course.)
(Nope, nothing wrong with the equipment.)
So, it is not possible to simply remove (delete) simple selected (those above 0dB) datapoints in the dataset with WL 12?
I do not want to overcomplicate things here.
Again:
Select all datapoints, that are above 0dB.
Than delete those points from the dataset.
Is it possible in simple terms to do this in WL12 or is it not?
If it is, how?
We’re not talking about rocket science here, again… just a simple editing feature.
Sure, I could write a short programm for this, working direktly on the raw data.
Yes, this is going to be a little off from your actual question, but if you want to avoid doing this by hand and the limiter suggestions don’t work for you, the most elegant way I think of for doing this is with iZotope RX’s “find and replace” function. It will zap the offending instances automatically. I use it when I need it, and it can be very effective.
Here’s AI’s description of the function… would have written it myself, but am in the middle of a project! Hope this is helpful, if not specifically addressing the fine details of your question.
To address at least one of your issues specifically: You could try WL’s Declicker module (part of the Restore Rig, in the Steinberg “Restoration” folder). iZotope’s version of the same is my first line of defense against these sorts of annoying artifacts, and if used carefully can be very effective and absolutely transparent. For what it’s worth, I just tried WL’s, and it’s pretty good!
Chewy
Says AI:
In iZotope RX, you can find and replace sounds using the Spectral Repair module and its Replace tab, or through the Find Similar tool. The Spectral Repair module allows you to isolate and replace unwanted sounds by visually identifying and selecting problem areas, then replacing them with similar sounds from surrounding regions. The Find Similar tool can help locate and replace similar instances of a sound, allowing you to fix multiple issues simultaneously.