zebra spectromorph bug /highest frequencies set to normalize equals full saw wa

[gentleclockdivider]

When in spectromorph mode , when just selecting the highest portion of the spectrum ( far right corner ) , these are around 15 khz ..so just draw a spike waveform to select the highest frequencies ...
Make sure the lower harmonics are silenced ( draw flat line )
Now turn the normalize knob fully right , result is a full saw wave (thus all harmonics ) when only the highest harmincs are selected

[Sendy]

I'm trying it right now. Completely zeroed spectroblend waveform apart from the last (highest) partial which is set to full, Normalize set to full. I'm not getting a sawtooth, but if any of the lower partials are just shy of zero (i.e. they look zeroed but actually have a tiny nonzero value) the Normalize algorythm will make them massive, creating something like a sawtooth in some cases.

The problem I do get with this setup, though, is quite bad aliasing, which disappears if you add lower partials (thank god!). However, such a setup is pretty useless since the result can be easily achieved by transposing a lower partial, so I don't consider this a major bug at all.

[gentleclockdivider]

You have to use 'spectromorph' not 'spectroblend' ...spectroblend is drawing individual partials , spectromorph is isolating the frequencies( by drawing a curve .or whatever shape )
Trie again you 'll see what I mean.
At 2:11 in the video just select as mall portion in the upper region and normalize

thanks
Urs any insight on this ?

[Sendy]

Oh, yeah. I see what you mean. I think what's going in is that if you select such a high region of harmonics, they're too high for the normalize algorythm - perhaps the normalize algo doesn't act on or recognize the higest partials (to avoid ear-splitting noises maybe?) and hence it disregards them and normalizes the remainder of the spectrum - which is of course nothing so it raises the levels of everything creating a plain saw tone.

That's my hypothesis. Be interesting to see what the real reason is but I'm fairly sure it pivots on the fact that you're telling it to amplify very high frequencies and it's like "no".