De-Clipper False Readings(?)

[Lord Theo]

Hi everyone,

After restoring and gain-scaling a 13-minute raw 24/96 file to -1.7 db from approx. -20db, de-clipping resulted in 5603 clips fixed.

True or False?....I can't hear any difference between the two.

[Craig Maier]

Hi,

Statistically speaking, 5603 "clips" in a 13 minute file is insignificant and probably not audible. If you fixed that number of clips in a 10 or 20 second file, then I would expect you to be able to hear the difference. Just my $0.02.

Craig

[Dan McDonald]

I de-clipped a file recently that was about an hour long. over 400,000 clips; reversed the file and got 100,000 more. I think I could hear a difference, but not sure.

[Craig Maier]

The De-Clipper algorithm in Non-Adjustable Strength mode is more sophisticed (adaptive) than the variable strength mode that a user can set. I find myself always using the Non-Adjustable Strength mode lately. (Of course, in either mode, you have to provide some headroom for the interpolator - - - I usually use -6 dB) As for 500,000 clips in one hour (8,333 / min), I suspect that would be audible (depending on the source material). It might be less noticable on rock 'n roll music than on classical or accoustical music. I have even heard of some folks intentionall clipping their audio by normalizing to something greater than 0 dB (like +3 or +6 dB). This feature is found under the CD prep menu (Normalize Gain Scaling). Some folks like that sound on certain types of music.

Craig