Hello everyone,
is it possible to train your own voice in synth v and then save it as a fixed voice and then select it?
Or are only the unfortunately few voice models available to use?
It would be cool if you could create your own models.
This has been asked a couple of times, most recently here:-
DIY Voice banks
You can already “use” your voice as an input to Vocoflex by Dreamtronics, I have never used the program but there is a whole category for its discussion (left side of screen). There are a few songs in the Showroom that use Vocoflex such as …
vocoflex-synthesizerv-ayame
1 Like
Sorry, I guess I didn’t look properly.
Thanks for the links.
I am lucky that my singing and Asterian’s singing are interchangeable with Vocoflex. If I want to use Asterian’s singing in Synthesizer V Pro, I will first have to sing the song and transfer it to the Synthesizer, where Asterian will take over my voice along with my Czech. Then it is easier to work with phonemes, so as to achieve better quality Czech. The output of Asterian’s version can then be converted with Vocoflex to my voice color. With other AI voices, it is not so convincing.
1 Like
I definitely have to buy vocoflex. But at the moment I’m still waiting for an offer, it’s a bit too expensive for me. I also have to buy the voices little by little. At the moment I only have May. I’ll try to experiment with German then.
I don’t know if you will have seen this but I just got an email from:-
… and within that was the following invitation (edited to just show snippets)
.
Happy New Year! December - January 2, 2025
.
NAMM, New Development, and More!
.
Eclipsed Sounds This Month
.
This past December, ES has been making steady progress on our new projects. Alongside development, we will also be making an appearance at the 2025 NAMM Show at booth #15324. We will be heavily scouting voice providers for new projects alongside showcasing upcoming ES products at the Show, so I highly encourage you to come visit us if you’re attending! I can’t say more beyond that, so you’ll have to just see for yourself; I guarantee you won’t be disappointed with what we’re bringing to the show.
(edit…)
In case you haven’t seen Dreamtonics’ announcement on social media, they also teased something new that should have more information at the NAMM Show - Synthesizer V Studio 2! Our booth is right nearby, so be sure to stop by both Dreamtonics and ourselves later this month!
.
.
.
Calling all Singers!
.
As we move into the next phase of ES’s development cycle, we are putting more focus on specific genre styles of singing. We are looking for expressive, genre specific voices that can fill out our upcoming roster. If you’ve ever wanted to work with Eclipsed Sounds as a vocalist, now is the perfect time to let us know! We’ll begin reaching out to new singers in Spring of 2025 , thank you to everyone who reached out so far - please understand as we’ve stated previously we can’t begin making new arrangements until that time due to our already busy development schedule!
.
Learn more
1 Like
There’s currently no option to train your own voice using SynthV.
Don’t expect Vocoflex to be able to clone your voice. If you read the description carefully, you’ll see that they don’t make that claim.
The style of the underlying provider gives a lot of cues to the voice, so even if Vocoflex closely matched your voice timbre, the result would still have a strong flavor of the original vocal.
I’ve experimented a bit with Ace Studio custom voices, but although the result was in the general ballpark, it wasn’t close enough to the provider to be useful. However, I’ll continue working with it and see if I can get better results.
1 Like
I jumped in on Vocoflex a few weeks ago, and using your own voice sample will give a subtle hint of your vocals over Synth V, but that is all. If you are expecting to be able to run a sample into Vocoflex and it come out sounding like Bieber or Ariana (or yourself!) then you will be disappointed.
I’ve been experimenting with the So-vits project converter ( GitHub - svc-develop-team/so-vits-svc: SoftVC VITS Singing Voice Conversion ) for a while, and you need huge amounts of clean vocal samples to even get near to producing a clone of a vocal style. Vocoflex only needs 10-15 seconds to work, so it is never going to be able to do a full voice conversion.
(As a side note, I am really mystified why owners are required to upload all manner of identification before using Vocoflex, given that the output is clearly not an AI clone. Unless of course they have ideas for future development that will bring us nearer genuine AI cloning?)
1 Like