Statistics panel and clipping.

[Jonesy]

I am curious to know how the Nuendo statistics panel calculates clipping. I assume that it registers a clip each time the input level of the audio clip exceeds 0dBFS, but does it take into account interpolation clips? What I mean is, when Nuendo or Cubase draws a curve between two samples, if the two sample points are below 0dBFS, but the interpolated curve between them rises above zero, is it logged as a clip?

[Vinark]

No but there is a free ssl plugin that will do that X-ISM, have a look at their site here (http://www.solid-state-logic.com/music/section_free_plug-ins.asp).

[Jonesy]

No but there is a free ssl plugin that will do that X-ISM, have a look at their site here (http://www.solid-state-logic.com/music/section_free_plug-ins.asp).

Thanks so much, that is exactly what I am looking for.

[neilwilkes]

If you have 3 consecutive samples at the limit point you have set, be it 0dBFS or -0.1/-0.3 whatever, the odds are very high that you have inter sample peaking at the reconstructed wave out of whatever player will be outputting the music. At -0.3dBFS limit you will (probably) need more than 3 consecutive samples, but not many more. It all depends on just how fast that limiter is - and a setting of 1ms is equal to 44 samples at 44100, 48 samples at 48000 and so on, so the odds of even an extreme setting like 1ms attack is just too slow to prevent inter sample peaking. You'd need to be in the microsecond attack range to get them all.
The best approach is to not go anywhere near 0dBFS, and peak at no higher than -3dB, preferably lower than that.