I am not talking about FORMANT GET or SEE or GETSEE - those I understand.
My problem is trying to wrap my head around the exact differences between REPLACING, IMPOSING and VOCODING.
I have provided illustration for my experiment - through five different sounds. The 'source' sound, which is a strange drone produced by the BLUR process. The 'formant' sound, which is a clip of me speaking clearly. And then, the outputs for imposing, replacing, and vocoding the formant sound onto the source.
The formants were grabbed at the highest linear frequency-wise resolution.
The frequency domain is log scaled.
Source:
(https://i.imgur.com/mL4xVDs.png)
Formant:
(https://i.imgur.com/qU6NxIp.png)
Impose:
(https://i.imgur.com/dbwylvm.png)
Replace:
(https://i.imgur.com/cRQNboo.png)
Vocode:
(https://i.imgur.com/JxJj2k9.png)
Could anyone try explaining these to me? I'm probably just way overthinking it, but I'm quite obsessed with trying to develop a comprehensive, rigorous understanding of these processes.
went over the source code with a programmer friend. made a couple guesses but didn't really get anywhere. I don't really think I'll be able to get any definitive answers unless it was coming straight from the designers/programmers. oh well!