I’m incredibly happy to see how much care is put in every update, and I appreciate the fact that updates are provided 100% free of charge! So please don’t get me wrong on what I’m about to say, but I think there may be a potential issue with the way phrasing is handled in the new version of the already existing voicebanks, like Solaria, Saki, Kevin etc.
First of all: I think the DPM implementation is amazing and has a lot of potential, some improvements are pretty evident already, for example in the way SynthV now handles falsetto, vibrato in staccato notes and fast melodies.
The problem is that in order to make this possible, to me it seems the engine changed the way it handles phrasing: up until 1.7.1 voicebanks known for their expressiveness, like Solaria, played every single line in a very lyrical legato fashion, but after the 1.8b update the engine always opts for a portato/staccato melodic style.
If, on one hand, this can be incredibly useful in some genres, on the other it makes the voicebanks nearly unusable in some others, unfortunately. I’m saying this because I’ve tested the issue myself on some old projects made before the update, and the difference in performance is evident. Not necessairly better or worse, just… incredibly different.
Please note: I’m appreciating the improvements in the update overall, and I don’t have any problems at all with the performance style of new voicebanks, like Mai (btw, free as well - thank you so much!). Quite the opposite, I think providing voicebanks singing in different styles is a great decision and it will be surely beneficial for SynthV as a whole on the long run.
I’m aware the update is still in beta, in addition to that I know it’s possible to switch voicebanks to their previous version if desired, but I guess it’s safe to assume the old versions will be progressively less and less supported in the future, so I was wondering: would it be possible to implement a new function (like tension, gender, breathiness etc.) where the user can decide whether to play “legato/smooth” or “staccato/defined”? I don’t know it the programming behind SynthV would allow such a thing, but it could be a possible solution…
Take this for what it is, an attempt at constructive criticism, and please keep in mind that I would have no problem whatsoever in further elaborating this request or in providing audio examples if something is unclear. And English is not my first language, so apologies in case of mistakes.
Thank you for your attention!