Stereo Tool

Van Zutphen's Eleven

There´s no frikandel in Las Vegas

You're absolutely right !

But try to imagine how many frikandellen Hans can buy in the Netherlands when he will return from Vegas.

michi => lol !!!!

Hey all, I'm back! (VERY tired, but inspired!)

(Tired because I visited the NAB every day, and took a helicopter flight over the Grand Canyon, and last night I have not slept at all because I hadn't had time yet to see the Strip, so I walked there from 1:00 AM to 5:30 AM, and then I had to get ready to go to the airport).

Anyway, after being in the Omnia booth listening to the Omnia 9 for 3 days, I think that it really sets a new bar for FM processing. The audio is insanely clear, dynamic and vibrant - it really sounds better than the original CD for many tracks (of course the declipper is an important part of that, but so is the dynamics increaser that Leif made).

Until now I thought that the sound was fine if it didn't HURT the quality (for modern, already compressed, CD's, my goal has always been to change them as little as possible in Stereo Tool; for older CD's I do make some improvements by increasing the clarity and bringing out details).

Until now only 2 processors more or less accomplished this goal: Stereo Tool and Breakaway (all others cause distortion, unless you set the output level very low). Currently, both Stereo Tool and the the Omnia 9 take things a step further - Stereo Tool by adding a declipper, and the Omnia 9 by adding the same declipper and what Leif calls 'undo': a dynamics increaser.

During my work on the Processing Freakday audio files, when comparing them to other processors (a.o. Omnia 11), I had already noticed it - but when comparing with the Omnia 9 it became really clear: There's still something wrong with my clipper (Loudness filter). Except for that, the filter is becoming more and more complex - which is usually bad (simple filters work better). So I'm planning a Loudness filter redesign...

Hey Hans,

Hope you had a great time in Vegas. I'm looking forward to test all the new inspirations in the beta versions.
Keep up the good work.

They will be great... Among others you can expect a big declipper improvement, if my idea works...

Do you mean adding a clipping level detector to automate the threshold like Leif is doing? That definitely would be a big improvement.

I'll definitely add that (of course), I just haven't had time for it yet - first thing was to check if my algorithm would work AT ALL, and for that automatic detection was not yet needed. (Actually this is one of the main reasons that version 6.20 is still beta).

But I think I've found something much better: What if I can detect clipping REGARDLESS of where it happens in the waveform? Even if - for example - one sound is clipped, and other instruments have been added later. Or if the audio has been processed afterwards.

For example in the Metallica track, clipping happens 'all over the place' - that is, anywhere above about -2.5 dB. So currently I need to throw away everything above -2.5 dB, which is the majority of the samples (!) - which makes it VERY difficult to properly reconstruct the audio. Many of the samples above -2.5 dB are actually ok - "all" I need is a way to find them.

Hans, superb that you had a wonderful time. Now go sleep and eat some healthy veggies after the hamburgers and all the meat.
I'll betcha the grand canyon tour was breath taking. Still remember my first time. Well after 3 times you get kind of used to it.

With what you wrote about the sound and processing changing sound but not only in the dynamics section we need to rethink about the term 'sound processing'. I would rather call it 'sound shaping'.