Best shit shifter!

 

I guess a whole octave without artefacts is beyond the capability of any pitch shifter. Maybe when using multiple passes of a few notes the result improves. Your source should preferably be recorded with a high sample rate - at least 96 Khz.

 

I'm not offering any suggestions here. I'm only asking a question to see what I might learn.

Consider an analogy - transforming the resolution of a visual image?
Sequence of Transform steps...
[1] from original -> then lowering-resolution -> then increasing resolution back to original
versus
[2] from original -> then increasing resolution -> lowering resolution back to original

[1] is (by definition) destructively discarding information on 1st step
whereas
[2] might generate new information and then discard it again on 2nd step

So [1] will fail to restore to original, whereas [2] might succeed in retrieving something close to the original

From the analogy

Your experiment with the order of steps being -1 octave shift and a +1 octave shift in series is similar to [1]

So, a question is...
Would your tests be less destructive (closer to passing null test) if the series is flipped to +1 octave shift first then -1 octave shift second.

@Sinus Well,
Sorry if this is just an irrelevant distraction; I'm just curious to know the answer and you've already got it all set up ready to conveniently find out.

 

"absurd shifts with minimal artifacts."

that like wanting to eat pizza without cheese

 

I logged only to give this post a rating!
Nice one!