Synthesizer V - Singing voice synth for Win/Linux/Mac (Free & Payware)

VST, AU, AAX, CLAP, etc. Plugin Virtual Instruments Discussion
Post Reply New Topic
RELATED
PRODUCTS

Post

Pelicanomicon wrote: ↑Sun Oct 14, 2018 7:44 am Downloaded the latest Windows version and copied the notes/lyrics according to your screenshot. Rendered at 80 & 180BPM and on my machine, SynthV clearly pronounces the "v" in "leave."
What happens if you use phonemes instead of phonetic approximations? (Highlight the problem note, go to Edit in the menu bar, then Convert to Phonemes.)
In Arpabet, "leave" should be ".l iy v"
However, Eleanor seems to pronounce this as ".l ih v" (rhymes with "give").
".l iy iy y v" sounds better.
Splitting the word into two notes also works: "[.l iy iy y][.v]"
Ah, what was happening with me was that I was previewing the vocal along with my instrumental track.
The "v" was getting lost because of the track. When I unloaded my instrumental it was more clear.
But, this is because the v sound is not strong, its more like a breath out. If the breath frequency somehow finds a match with the instrumental then it gets lost, and the v gets lost too.
I still might append "voh" and glue it together to bring out the v more.
Top

Post

Bansaw wrote: ↑Sun Oct 14, 2018 1:24 pm Ah, what was happening with me was that I was previewing the vocal along with my instrumental track.
The "v" was getting lost because of the track. When I unloaded my instrumental it was more clear.
But, this is because the v sound is not strong, its more like a breath out. If the breath frequency somehow finds a match with the instrumental then it gets lost, and the v gets lost too.
I still might append "voh" and glue it together to bring out the v more.
Oh good to know it wasn't a bug and it got resolved :)

Some news: SynthV developer Kanru Hua has tweeted hints of features like multitrack vocals, "growl" and "scream." Stressed voices are something Vocaloid doesn't do well so it'll be interesting to hear how convincing the synthesis is.

ETA demonstration of Growl
https://twitter.com/khuasw/status/1051747173929377793
Top

Post

Pelicanomicon wrote: ↑Sun Oct 14, 2018 7:44 am
Bansaw wrote: ↑Sat Oct 13, 2018 2:47 pm I got hold of this recently, enjoying trying to make it sound the way I want.
I can't get it to say "leave" - the full sentence is "to make you leave" but its coming out "to make you leeee"
(pic attached)
I even tried to record "vuh" elsewhere and I'l paste it on the end of leee but I can't get it to just give me a "v" without it giving me a "veeee" sound.

Any ideas?
Downloaded the latest Windows version and copied the notes/lyrics according to your screenshot. Rendered at 80 & 180BPM and on my machine, SynthV clearly pronounces the "v" in "leave."

What happens if you use phonemes instead of phonetic approximations? (Highlight the problem note, go to Edit in the menu bar, then Convert to Phonemes.)

In Arpabet, "leave" should be ".l iy v"
However, Eleanor seems to pronounce this as ".l ih v" (rhymes with "give").
".l iy iy y v" sounds better.

Splitting the word into two notes also works: "[.l iy iy y][.v]"
Casually I have been working on a cover that includes both words in the lyrics: Believe (by Cher)

https://drive.google.com/open?id=1XF7I5 ... 3spVBHPny8

The only bug here is wagtunes, speaking without knowledge
Top

Post

Bansaw wrote: ↑Sun Oct 14, 2018 1:24 pm
Pelicanomicon wrote: ↑Sun Oct 14, 2018 7:44 am Downloaded the latest Windows version and copied the notes/lyrics according to your screenshot. Rendered at 80 & 180BPM and on my machine, SynthV clearly pronounces the "v" in "leave."
What happens if you use phonemes instead of phonetic approximations? (Highlight the problem note, go to Edit in the menu bar, then Convert to Phonemes.)
In Arpabet, "leave" should be ".l iy v"
However, Eleanor seems to pronounce this as ".l ih v" (rhymes with "give").
".l iy iy y v" sounds better.
Splitting the word into two notes also works: "[.l iy iy y][.v]"
Ah, what was happening with me was that I was previewing the vocal along with my instrumental track.
The "v" was getting lost because of the track. When I unloaded my instrumental it was more clear.
But, this is because the v sound is not strong, its more like a breath out. If the breath frequency somehow finds a match with the instrumental then it gets lost, and the v gets lost too.
I still might append "voh" and glue it together to bring out the v more.
Add breathiness to make unvoiced phonemes louder (also works as deesser with low values).

Use note property "Syllabe coda" to change last syllabe phoneme duration. High values, longer duration.
Top

Post

This is really good Cher cover.
I completed the song a few weeks ago. One line I'm not too happy with is "Why you left me", the "left" didn't come out too well. Fiddled with it, couldn't make it better.
(Don't like what Soundcloud does to my high frequencies, but here goes):
https://soundcloud.com/bansaw-843690132 ... ps-avenger
Top

Post

Bansaw wrote: ↑Tue Nov 20, 2018 6:14 pm
This is really good Cher cover.
I completed the song a few weeks ago. One line I'm not too happy with is "Why you left me", the "left" didn't come out too well. Fiddled with it, couldn't make it better.
(Don't like what Soundcloud does to my high frequencies, but here goes):
https://soundcloud.com/bansaw-843690132 ... ps-avenger
Cool! I like the song so much, it makes me dance!

I think the song is too high for the voice and would fit better transposing the song 2 or 3 semitones down.

I you agree, I would like to take a look to your .json file and try to improve the pronunciation. Basically, there are some work to do with phoneme durations.
Top

Post

Thanks Jaleo!
I have linked the json file.
The lines I could not get right were, "to make you leave" and "why you left me"
for "leave" I had to append a "voh" and then in my DAW chop the waveform at the end a bit. Thats why "v" ends a bit abruptly.

json:
https://drive.google.com/file/d/11dAhMe ... sp=sharing
Top

Post

The full editor along with new voices has been released!

https://synthesizerv.com/en/

A license is not yet available for purchase, but there is an evaluation version available.

The license for Synthesizer V is 79 USD
Top

Post

Bansaw wrote: ↑Thu Nov 22, 2018 12:16 am Thanks Jaleo!
I have linked the json file.
The lines I could not get right were, "to make you leave" and "why you left me"
for "leave" I had to append a "voh" and then in my DAW chop the waveform at the end a bit. Thats why "v" ends a bit abruptly.

json:
https://drive.google.com/file/d/11dAhMe ... sp=sharing
Hiii!
I don't know what hapenned but I missed to answer. I'll check your file along this week I guess, thank you for sharing!

I love the song, I'm dancing again haha!

Meanwhile, enjoy more synthesizerV demos I did lately:

https://soundcloud.com/blancanegra/sets ... er-v-demos
Top

Post

Jaleo wrote: ↑Mon Apr 01, 2019 10:23 am
Bansaw wrote: ↑Thu Nov 22, 2018 12:16 am Thanks Jaleo!
I have linked the json file.
The lines I could not get right were, "to make you leave" and "why you left me"
for "leave" I had to append a "voh" and then in my DAW chop the waveform at the end a bit. Thats why "v" ends a bit abruptly.

json:
https://drive.google.com/file/d/11dAhMe ... sp=sharing
Hiii!
I don't know what hapenned but I missed to answer. I'll check your file along this week I guess, thank you for sharing!

I love the song, I'm dancing again haha!

Meanwhile, enjoy more synthesizerV demos I did lately:

https://soundcloud.com/blancanegra/sets ... er-v-demos
Actually sounds pretty good. But then I remembered why I didn't get it. Difficult to work with. Well, at least not as easy as V4 in Cubase. But the sound is definitely there.
Top

Post

The paid next-gen Synthesizer V and two free versions with feature limits have been released :party:

Official Synthesizer V website
https://www.dreamtonics.com/en/

Standalone for now, aiming for VST in v1.0.5.

ETA: AU also planned, testing to start on 8/4 with a small pool of applicants
https://twitter.com/dreamtonics_en/stat ... 9668747271

Compatible with Microsoft Windows 8.1 or later, Ubuntu 16.04 or later, macOS 10.11 (El Capitan) or later

There are no commercial English voicebanks right now as the pandemic has stalled the development of next-gen Eleanor Forte but you can use her Lite version for free with Basic and Pro

btw when you see "tuning" mentioned re voice synths, it means the editing of parameters (tension, pitch bend, formant shaping, breath, volume, etc) to create a vocal performance. as opposed to just entering notes/lyrics and smashing export.

WEB SYNTHESIZER V (FREE)
  • Browser-based, requires Chrome 77+
  • Projects: 1 track, 300 notes
  • 2 background threads for live rendering
  • Can only use web-based voices
  • URL: https://synthesizerv.com/web/
SYNTHESIZER V STUDIO BASIC (FREE)
:!: Note the use/attribution restrictions on the Dreamtonics' download page. If you publish material using Lite voices, you must mention it

SYNTHESIZER V STUDIO PRO (PAID)
  • Projects: Unlimited tracks
  • No thread limiting for live rendering
  • Automatic pitch tuning
  • Swap alternate pronunciations of phonemes into notes
  • Aspiration export (splits vocals into de-essed and ess-only files for mixing)
  • Supports Lua and Javascript scripting for VOCALOID job-like functionality
DREAMTONICS' YOUTUBE TUTORIAL PLAYLIST

https://www.youtube.com/watch?v=5f50-Ir ... 0SGRJgRIPz
While UI in the vids is Japanese, Synthesizer V is multilanguage

WHERE TO BUY

Japan partner AHS, price ¥9,800 Taiwan partner Anicute, current price NT$ 2,350
  • Product page: https://www.anicute.com/pages/synthesizerV_studio_pro
  • Chinese and Japanese SynthV voicebanks from VOLOR, Quadimension and AHS on sale too
  • Pro+voice bundles available
  • Paypal checkout
  • Sale prices good at least through August, normal price of Pro will be NT$ 2,650
Amazon Japan (boxed) I'm led to believe that dev licensing for Synthesizer V is cheaper than VOCALOID so fingers crossed we get some good English voices out of this.
Last edited by Pelicanomicon on Wed May 12, 2021 9:11 am, edited 2 times in total.
Top

Post

Hi,

Synthesizer V Studio Basic crashes for me with illegal instruction (AMD FX 8120). Does it requires an AVX2 (or newer) CPU?
Top

Post

Kott wrote: ↑Sat Aug 01, 2020 1:34 pm Synthesizer V Studio Basic crashes for me with illegal instruction (AMD FX 8120). Does it requires an AVX2 (or newer) CPU?
Dreamtonics website only says Core i5 or AMD equivalent. For more specific hardware requirements, maybe you could email or tweet them?

Here is a post from them that encourages troubleshooting/bug reports and gives out their email
https://twitter.com/dreamtonics_en/status/1288643637933297664

They also have a forum
https://forum.synthesizerv.com/
Top

Post

Quick roundup if you haven't been following Synthesizer V developments:

Synthesizer V Studio Pro (the paid version of the editor) supports VST/AU

Synthesizer V voicebanks can now come in "Standard" and "AI" versions. AI VBs use deep learning models of the voicers for humanization and easier tuning. Here's a short comparison of Saki Standard and AI:
https://www.youtube.com/watch?v=T63KKcG1jBM

And at last! some encouraging news for forlorn Western users :party:

A new company called Eclipsed Sounds has announced a crowdfunded native English voicebank for Synthesizer V. This voicebank, "SOLARIS", will be performed by professional singer Emma Rowley.

IndieGoGo link:
https://igg.me/at/synth-v-solaris

Homepage: https://www.eclipsedsounds.com
Twitter: https://twitter.com/eclipsedsounds
Q&A page: https://eclipsedsounds.tumblr.com
Emma Rowley's Youtube channel: https://www.youtube.com/channel/UCsgVcP ... sqYUh0mAXw

As proof of intent, Eclipsed Sounds has released a test based on Synthesizer V's Standard recording method. Per their Twitter, please keep in mind that the test is monopitch, "extremely limited". The goal is to fund an AI voicebank that approaches the actual voice provider.

SOLARIS:
https://www.youtube.com/watch?v=oA6QZFEtKzU

The same song, performed by Emma Rowley:
https://www.youtube.com/watch?v=PJTUiUVviVA

Campaign goes live this Friday, May 14, 2021
Top

Post

ENV1 wrote: ↑Mon Aug 27, 2018 3:18 pm(Forum software insisted on killing my tabulators so i put periods. Still messed up but better.)
I know it's been nearly 3 years, but protip: if you want to retain spaces and tabs, use CODE tags :)


This is mighty interesting software, Vocaloid finally getting some proper competition!
Top
Post Reply

Return to “Instruments”