Who is able to create a simple VST plugin or nq?

Thank you trebor for the links. I often visit the site for sfx’s in my audio dramas.
After all, the simulation is not as bad as you may think. When I am not mistaken, the clicks shall only be removed within the silent parts of the transcription file. With a original sample I would have been able to set valid defaults in my plug-in since recordings differ widely from one another. Here’s a sample unprocessed/processed radio recording from the 1920’s.

Threshold: -54 dB
Click-duration: 85 ms
Distance: 43 smps

It would be possible to put a bandstop filter before the sound detection, somewhere between 200 and 4000 Hz, where speech “abides” habitually. Has any one tested the plug-in yet?

Yes, I’ve tested it. It does what it is intended to do.

In practice, as shown by tour “about Holmes” sample, the sudden muting can be as distracting as the problem that it is fixing. The main problem is the suddenness of the muting. For the “about Homes” the audio can probably be cleaned up more effectively with “traditional” filtering and noise removal as in this sample:

It will be interesting to hear how it performs on jankobenko’s audio.

Time for some input from you jankobenko.

The final t of “thought” has been lost in the left channel, (see attached excerpt).
The amount of silence which must elapse before the filter is activated needs to be increased to avoid losing click-like speech sounds like t,k,b,d at the end of words.

the sudden muting can be as distracting as the problem that it is fixing.

Sudden, complete muting is a theatrical plot conceit. It get’s the audience’s attention. It’s what you do just before the real murderer is revealed…

In Engineering, it’s what happens just before the phone rings from people wondering what happened to the show.

Have we got the PAF yet?

Koz

@Trebor
Now you can see how sensitive the effect reacts to small differences. With a half percent higher threshold, the “-ught” would have been removed in both channels and its entirety. The listed values are not those I would use to remove mouse clicks. But that wasn’t a problem in those long gone times anyway…
It’s really easier to work with mono sounds since you are occupied enough with the finding of the settings for one channel solely. However, with a really sophisticated algorithm (like CEPSTRAL or LIFTER)we could isolate the speech and a lot of noise would have been removed before the actual click removal. But that’s fruitless thinking unless Jankobenko gives his feed back.
Maybe he had to deliver his translation before yesterday and the plug-in will accordingly disappear in a virtual drawer.

@Koz
Jankobenko wrote that he removes the clicks with Ctrl-L and produces thus perfect silence. For the first version I’ve stuck to this principle but I don’t like it myself. The next version will include a short fade to a certain level (maybe a additional control for testing purposes).
Thank you guys for the replies.