Edit: The bug might be the issue then....

Nice.. but Input level should be so precize.. hope You'll do something for this.

Yes, please try again with BETA020

I can hear a difference in that wav file now, but ONLY if I move the assume tilted values to be clipped slider all the way down below 3100. The other slider really doesn't have much impact for me. Even with that, it has its' own distortion sound too, like perhaps sample quantization errors (might be using a totally wrong term, dunno).

Still not very impressive, to be honest...especially in light of the CPU consumption.

That is strange. It should sound horribly clipped, and after restoration, nearly perfectly restored (except for 'S' sounds) - as in the 2nd .WAV file in the post I linked.

How are you playing it?

Winamp, same way I am playing everything else. Same settings. Although I do have a few changes to the preset that I had uploaded, there's no significant difference with what I had uploaded with respect to this new change.

Bear in mind you're running on a newer processor, and specifically an Intel processor.

All-in-all, unless you make some significant improvement, this would not be a selling point for your product for me.

Ah, you're using other filters as well. Specifically, this filter does NOT (yet!) work properly if you combine it with a non-phase linear highpass filter.

What happens if you turn all other filters off (press RESET)?

(And, just to be sure: You DO get horrible distortion without this filter, right? Otherwise you might be using the wrong file.) The output should be just as different as the 2 files on that other page:
Original (BAD): http://www.stereotool.com/temp/original ... PPEDx4.wav
After declipping (GOOD): http://www.stereotool.com/temp/01_origi ... PPEDx4.wav

Also, your nag message kicking in every 60 seconds pushes the dB value through the roof, so the filter disables when that kicks in,

I did RTFM Hans... I specifically said that I completely disabled bandpass to test this. There is no highpass or lowpass involved at all.

Yes, the distortion is there due to overdriving, and I have the correct file. The only way I get an appreciable improvement is NOT with the first slider, but with the tilted values slider. I've tested with everything disabled too. It's the same way. The only way it works for me is with the titled values. It's still set for 0 degrees on the tilt.

Also, your nag message kicking in every 60 seconds pushes the dB value through the roof, so the filter disables when that kicks in, and I think it takes a few moments to stabilize under the significant CPU load, so between the lack of it doing much of anything good and the high cpu load, it's not impressive in the slightest.