SynthV Synthesizer V Discussion Thread
- Thread starter xuu
- Start date
-
- Tags
- synthesizer v
just woke up and downloading this right now! i dont have use for a rap voice for rap, but im curious to use this in other contexts for sure. releasing two new voices is cool, and i think its great they did because obviously, we dont have any voices that were recorded with the intent of using rap mode. im not convinced other vocals will sound great with it even if compatible- basically any new feature, even like the parameters having vocal modes simply dont work at all (yes, banks that didnt have that update in the 1.8 beta could not use the vocal mode parameter windows, even if they obvious had vocal modes). i feel like the update may just be to make sure they actually can be played properly and not exactly that every voice will even have the data to sound proper in that mode.
anyways. thats just speculation lol-- what i hope to imply is, i hope we will get some interesting voices on that rap mode, i hope others will work well with it and maybe have the added data to work even better for their specific voices too.
the other changes made are certainly nice quality of life choices too, stuff i have been quite pleased with every time.
Also, i assume it wasnt a mistake, but on the Feature Limited download page D-Lin is the only one with listed restrictions.
anyways. thats just speculation lol-- what i hope to imply is, i hope we will get some interesting voices on that rap mode, i hope others will work well with it and maybe have the added data to work even better for their specific voices too.
the other changes made are certainly nice quality of life choices too, stuff i have been quite pleased with every time.
Also, i assume it wasnt a mistake, but on the Feature Limited download page D-Lin is the only one with listed restrictions.
My desire to do a flos cover with Feng Yi just increased… please give her the rap update
Good observation about improvements to XLS in this update.
whoa, i didnt realize how much this affected all of it until just now. because i just went to use Yuma with his beta version and he.... sounds very noticeably different! i was shocked.
that strong "lisp" that he has in english isnt exactly gone, his unique tone is still there. but he does have a lot more clarity. his lisp is cute and i like it, but it would sometimes cause issues with some enunciation and it looks like thats been improved in that regard.
luckily, with being able to seamlessly change versions, you can always go back to that more slurry version of him, which is nice to be able to do tbh. (the idea of switch his versions between parts of a song seems cool tbh).
here is an example (first one is old, second is new). it hit me because i was literally listing to this file with yuma last night lol.
that strong "lisp" that he has in english isnt exactly gone, his unique tone is still there. but he does have a lot more clarity. his lisp is cute and i like it, but it would sometimes cause issues with some enunciation and it looks like thats been improved in that regard.
luckily, with being able to seamlessly change versions, you can always go back to that more slurry version of him, which is nice to be able to do tbh. (the idea of switch his versions between parts of a song seems cool tbh).
here is an example (first one is old, second is new). it hit me because i was literally listing to this file with yuma last night lol.
Ryo's english got a lot better with the update too!! ritchy has the potential to become one of my favorite new synthvs, he sounds nothing like anything else on the platform
excited to actually try out some of my non-english vocals now tbh. i had some gripes with a few of them (esp Ryo) and I'm more confident in actually buying some of the other non-english voices too if this really improved them this much.
Not super relevant to anything, but the new song/manual/rap color split is very satisfying to look at.
thoughts about the engine 
Have been working on my first 'full spec' SVP. Lots to do.
I bought SynthV Studio Pro yesterday and have been having the time of my life. It really hit me just how much this software has changed since we first tried it when R1 was around and GENBU was the coolest guy on the block. It's been about a year since I last used SynthV, and this is my first time using the paid version of the engine, and the progress is noticeable. As far as engine development goes, Dreamtonics has really exceeded expectations with SV, I feel. I don't know if we've ever seen such persistent, large (and free?!) updates to a software product in this space before, and that is worthy of praise. CeVIO AI has had some good stuff spurred on from 3rd parties too, but you can tell the research is still very much ongoing with this, which is good. I wish that the roadmap for how 3rd party vocals are integrated was more clear-- history shows that the communication likely isn't the best-- but at least versions are being tested in BETA now before being pushed to stable, which is an acceptable improvement.
The sidebar pop-out scroll up-down UI is still a little weird for my taste, but it works well enough and it feels easier to find things in general than it did at the beginning.
The workflow is decent and I haven't noticed particular annoyances at first brush with the exception of two things.
My #1 hope for the future of SynthesizerV has less to do with feature updates and more to do with people making more fun, quirky, unique (and not necessarily realistic) stuff with it.
TL;DR: fun engine, would recommend.
Have been working on my first 'full spec' SVP. Lots to do.I bought SynthV Studio Pro yesterday and have been having the time of my life. It really hit me just how much this software has changed since we first tried it when R1 was around and GENBU was the coolest guy on the block. It's been about a year since I last used SynthV, and this is my first time using the paid version of the engine, and the progress is noticeable. As far as engine development goes, Dreamtonics has really exceeded expectations with SV, I feel. I don't know if we've ever seen such persistent, large (and free?!) updates to a software product in this space before, and that is worthy of praise. CeVIO AI has had some good stuff spurred on from 3rd parties too, but you can tell the research is still very much ongoing with this, which is good. I wish that the roadmap for how 3rd party vocals are integrated was more clear-- history shows that the communication likely isn't the best-- but at least versions are being tested in BETA now before being pushed to stable, which is an acceptable improvement.
The sidebar pop-out scroll up-down UI is still a little weird for my taste, but it works well enough and it feels easier to find things in general than it did at the beginning.
The workflow is decent and I haven't noticed particular annoyances at first brush with the exception of two things.
- In an effort to get out of my comfort zone and really give the new features a fair shot I've been building this file so far in instant mode.
I think, in general, instant mode does what it's supposed to do-- and people that aren't invested in the concept of tuning are going to get incredible milage out of it.
The thing is, and I know for certain this has been discussed ad nauseum, but I gotta add my voice to the choir-- the AI pitch models are super biased towards huge, deep vibrato, and adjusting it while instant mode is on is near impossible. You have two options: use new seeds through AI retakes and hope that a new value brings it closer to what you want, which works sometimes, or try to adjust through the vibrato panel-- which will only work the way you'd expect in 10% of situations. You can't try to "draw over" the vibrato with pitch deviation-- it will offset the existing AI generated pattern in instant mode.
Vibrato works fine in the manual mode, but it feels like a large oversight for a feature that has very much been put in the spotlight, whether it's through "instant mode" or the new "song" note type.
I haven't tried the BETA yet, but if setting individual notes between manual and song is granular enough, it'd be an easy fix. Until then, the "best of both worlds" method is to go through and tune with AI support in the instant mode, then convert all the pitch edits to manual points in manual mode, delete the points on the vibrato you want to fix, and then rewrite with the vibrato note settings (which work perfectly fine in manual mode). I think the way to have the most fun with SynthV's features could indeed be combining your own tuning sensibilities and skill with sensitive AI pitchbends, so hopefully the new update brings that into balance.
I think I prefer the CeVIO angle, where vibrato is separately parameterized from pitch deviation and can be drawn or otherwise easily manipulated with a direct onscreen method. But it does take confidence to draw vibrato, so SynthV's options have good value in that respect. - There are no advanced options for keyboard binding. For example-- zoom control must be bound to one of three options, all of which include the scroll wheel (or OS touch gestures by extension). My pointing device doesn't have a scroll wheel, instead it has a button to activate a scrolling mode. Big accessibility issue.
- As an extra note, the most annoying characteristic of SynthesizerV's sound, in my opinion, the raspy, engine-y quality of soft voices, is still very much present through parameterized editing, but vocal modes make this infinitely easier to avoid.
My #1 hope for the future of SynthesizerV has less to do with feature updates and more to do with people making more fun, quirky, unique (and not necessarily realistic) stuff with it.
TL;DR: fun engine, would recommend.
I am very very excited, I know of Nene Hieda’s voice work and I have loved all of TOKYO6’s characters. They all have the quality of being “normal girls” that still have incredibly unique designs and silhouettes. They have the feeling of not being “one dimensional”, and of course, the voicebanks are also good. Here’s hoping.
I wonder when they are done with the seasons if they are going to make a male character. I'm curious what they have planned to expand the cast. Synth v seems ever updating I doubt we'll see a synth V 3 anytime soon that will cause them to have to update their existing characters like vocaloid.
AUDIOLOGIE has two new voicebanks planned for 2023 and 2024, according to the prize details for their song competition (also: they're doing a song competition)
wow, that is some golden prize results for first place... like, really lol. i feel like so many contests i see as of late havent had nearly as interesting prixes. so, props to them for giving so much, almost eager to do it just for that even though i havent made music in years lol.
AHS has teased Haruno Sora for SynthV in their latest livestream. She sounded extremely good on VOCALOID5, so I'm excited!
Here's the tweet announcement
(I LOVE THOSE BOOTS ALREADY)
Also, looking forward to hearing her because I used her V5 trial once and really liked her Cool voicebank
(I LOVE THOSE BOOTS ALREADY)
Also, looking forward to hearing her because I used her V5 trial once and really liked her Cool voicebank
Omg Sora update! Now im waiting for Poron to get a vb somewhere! also, she'll be shown at that one vsynth online live event on June 25th if i'm not mistaken?