French vocal/voices for synthesizer v

Synthesizer V Studio
1

Hi everyone,
New here (first post) and a recent user of Synthesizer V (either as a standalone application or as a plug-in in Cubase Pro 13).

I hope you’re all doing well!

I think the topic title speaks for itself :slight_smile: English is great, Chinese, Spanish… also, but I’m desperate to find a (female) voice capable of singing properly in French.

I spend hours working with Felicia (with the additional help of ChatGPT-4o to find/transcribe the best phonemes [by the way, for your information, AI in this context provides valuable and truly effective help]), but… nothing works, an accent is an accent, a language is a language :slight_smile:

So here are two questions:

  1. Are there any of you who use the wonderful Synthesizer V to produce songs in French? And if so, is there an existing voice that handles French better than others?
  2. Why isn’t there a French voice for Synthesizer V yet? I do understand that developing a voice is a huge undertaking and that market demand can influence developers, but French is still ranked quite high in terms of global usage. Moreover, French, when it comes to artificial voices, is already available at a very high level of quality with generative AIs and tools like Suno (for example). They offer perfect pronunciation from simple written text.

That’s it. I’m starting this topic in the hope that I’m not duplicating another conversation on the same subject (apparently not). If I’m not alone in this situation (wishing to use Synthesizer V with one or more French-speaking voices) and if you too are struggling to achieve satisfying results, let’s discuss it! And if Dreamtonics developers happen to be reading this… maybe it will help!

Best regards to all!

2 Likes
2

I have been experimenting with producing French songs using Solaria, Natalie and Kevin voice banks. I combine Spanish and English phonemes to approximate 16th Century French pronunciation:

Here is a sample:

Claude Janequin: Le chant des oyseaulx (c.1528) with Synthesizer V (c.1528)

1 Like
3

Here is another sample from the 17th Century:

1 Like
4

Hi htokuno,
Thanks for your reply.
Yes, I had already discovered and listened to your work. Impressive!!!
I also wanted to give an example of what I’m working on but I discover that it’s not possible to post an audio file here. And I don’t use a Youtube channel. I’ll see how else to do it via a link.
Thanks a lot !

5

This looks really interesting.

Because I like Patricia Kass, I have chosen the phrase “Mon mec à moi…” and transcribed it into English phonemes … mo(p) me(n) ca(p) mwa. The letters in brackets I remove so the phoneme become … m aa m eh k ae m w aa

It sound like this:

To add this audio file I used Dropbox and changed the ending from dl=0 to raw=1 and just posted the link without [url] tags - it seems to work.

6

As to why no french VBs… the developer and most of the associated companies that develope the SW and VBs are Japanese/Asian. Only one that I am aware of is based elsewhere, Eclipsed Sounds, from the US. So finding a native speaking voice provider might be a bit more difficult for them than with Japanese/Mandarin/Cantonese/English/Spanish.
Then, the French-Speakimg market is also smaller, so there is not the pressing need for them to develope French as it was English.
Hopefully with V2.0 more languages will be made available. I am sure French and others will arrive eventually.

7

Thanks Perrygoround.

I still think that the French language market is not negligible.

FYI :
According to the 2022 report by the International Organization of La Francophonie (OIF), French is the fifth most spoken language in the world, with approximately 321 million speakers. The languages with more speakers than French are Chinese, English, Hindi, and Spanish.

In terms of geographical spread, French, along with English, is the only language spoken on all five continents.

Additionally, French is the fourth most used language on the Internet.

Thus, French holds a significant position both in terms of the number of its speakers and its global presence.

Moreover, okay, in absolute numbers, Chinese and Indian languages are highly represented, but that doesn’t necessarily mean a significant number of potential buyers :wink:

1 Like
8

I have chosen this second phrase: “Il parle d’amour comme il parle des voitures.”

:arrow_right: (p)eel parl(iment) dam (t)our com(ma) (p)eel parl(iment) day vwa chew rou(gh)

And it sounds like this:

It’s true to say I’m having fun experimenting !

PS: I forgot to add the actual phonemes…

iy l p aa r l d ae m uh r k aa m eh l p aa r l d ey v w ax ch uw r ah

9

That’s why there is a realistic chance French will be the next language to be released.
If you look at the current languages supported, 4 of them are 4 of the 5 with more speakers than French (English, Spanish, Manadarin and Cantonese). And the fith is the developers native language.
So it will make sense Hindi and French will be next in line. However, total amount of speakers does not equal to market size. Many French speakers are in Africa, where I guess the market for vocal synthesizers is not so big. Similar for Hindi and India. When looking at the market for bocal synths, I guess Korean and German maybe could be on par with French?
K-pop is huge and a natural fit for this product. Germany and Austria are very wealthy places and combined have a big population.
I hope all of them will arrive soon.

1 Like
10

Hello Perrygoround,

Thank you for your pertinent feedback. You’re absolutely right about the importance of the French language on the African continent and this market, which is undoubtedly still small in terms of equipment. And as you say, it must be similar for China and India.

We can also discuss the subject from another angle: music production, and particularly song production (music + vocals). These days, there’s hardly any creative work that doesn’t involve a studio equipped with PCs, Daw, etc., which is why it’s so important to have the right equipment. In 2021, world music production was dominated by the following countries: 1- USA / 2- UK / 3- France / 4- Germany / 5- Canada (including a French-speaking part) / 6- Japan.

Anyway, all this is not to polemicize, just because I love looking at subjects from all angles, it’s interesting! (and talking about it here might get back to the developers and motivate them :wink: Come on! V2 will be accompanied by other languages for sure! :slight_smile:

Sincerely

1 Like
11

I really hope all the languages of the users here will be covered soon.
My own songs are 40% in English and 60% in Spanish. And for me, the release of Spanish was a massive improvement over mixing English and Japanese phonemes to try an approximation.
Since I have been in both situations I know first hand how much needed is to obtain your own language in the software.
I really whish the developers will take note and kerp releasing new languages.
The users pressure demanding Spanish did work, so I agree with you, leaving these messages in the forum (and other relevant social media) asking for French, Italian, German, Korean, Vietnamese, etc seems like the polite way to go.
Good luck getting French!!

(I was a student of French language for a couple of years when I was younger and it is a beautiful and very musical language, indeed).

1 Like
12

I see Korean has been included now as the 6th language supported, starting with the release ov 2.0. I remember I mentioned it here, as one of the languages I was predicting it could be next in line, due to prevalence of K-pop and its suitability for a product like Synth V, and the developers being asian and probably being easier/more profitable for them to add Korean before other languages (for different reasons).
Good news are various: (1) one more language supported, which is great news by itself (2) the confirmation that they continue working in more languages and are not just limiting to the current ones (3) this means the next supported language should be closer once the previous one is released.

1 Like
13

Hello !
Moi j’essaie de faire du français avec SynthV. Ce n’est pas toujours hyper concluant, mais on parvient parfois à des choses qui s’en approchent. J’utilise des voix paramétrée en anglais, et la plus grosse difficulté reste le “r” si typique au français.

Quelques trucs que j’ai pu constater : “j’aime le français”, je l’écrirais 'zh eh m l ax f rr" et là un espace vide “r aa n s eh”. Parfois, il faut doubler les consonnes, car elle sont souvent davantage accentuées en français. Par exemple, en reprenant la phrase précédente, avec une respiration avant, on aurait (en supposant une note par syllabe) : “br zh”,“zh eh m”, “m l ax f f rr” un espace vide, “r aa n s”, “s eh”.

La difficulté, à part le “r”, c’est les sons presque uniques au français “on” (poisson) : “ow n” est ce qui me semble le moins mauvais
“in” (lapin) :“ih n” est ce qui me semble le moins mauvais
“en” ou “an” (“venant”) : “aa n” est ce qui me semble le moins mauvais

En revanche, un mot comme “lié” se fait assez facilement, avec “l iy eh”. Il faut juste en fait couper en deux notes collées (“l iy” d’abord, “eh” de l’autre) et trouver la bonne durée pour le “l iy” en tâtonnant.
C’est vrai de la même manière pour “moi” qui donne de manière convaincante “m ow” puis “aa”.

Le “u” (urluberlu) est délicat à faire; on peut utiliser (note de passage) “uw”. Mais aussi “y”, avec un soucis pour ce dernier : il y a comme une latence avant l’établissement du son.

Le “ou” (loup) se tente, à défaut de mieux avec “uh”…

Avec tout cela, on pourrait faire “casser la voix” : “k aa s”,“s eh l”,“l aa v”,“v ow”,“aa”. ou “Mon mec à moi” : “m ow m”, “m eh k”,“k aa m”,“m ow”,“aa”.

Voilà à l’usage ce que mes expériences m’ont apportées. Je suis preneur de compléments !

Et aussi de toute info, rumeur, etc. concernant la prise en charge du français par SynthV officiellement.

1 Like
14

Hello!

I’m trying to make French with SynthV. It’s not always very successful, but sometimes we manage to get something close. I use voices set to English, and the biggest difficulty remains the “r” so typical of French.

Some tricks I’ve noticed: “j’aime le français” (I love French), I would write it as ‘zh eh m l ax f rr’ and then a blank space “r aa n s eh”. Sometimes, you have to double the consonants because they are often more emphasized in French. For example, taking the previous sentence, with a breath before, we would have (assuming one note per syllable): “br zh”, “zh eh m”, “m l ax f f rr” a blank space, “r aa n s”, “s eh”.

The difficulty, apart from the “r”, is the sounds almost unique to French:

  • “on” (poisson): “ow n” seems the least bad to me
  • “in” (lapin): “ih n” seems the least bad to me
  • “en” or “an” (venant): “aa n” seems the least bad to me

On the other hand, a word like “lié” is quite easy to do, with “l iy eh”. You just have to split it into two glued notes (“l iy” first, “eh” the other) and find the right duration for “l iy” by trial and error. The same goes for “moi” which convincingly gives “m ow” then “aa”.

The “u” (urluberlu) is tricky to do; you can use (passing note) “uw”. But also “y”, with a problem for the latter: there is like a latency before the sound is established.

The “ou” (loup) can be attempted, for lack of better, with “uh”…

With all this, we could do “casser la voix” (break the voice): “k aa s”, “s eh l”, “l aa v”, “v ow”, “aa”. or “Mon mec à moi” (My guy): “m ow m”, “m eh k”, “k aa m”, “m ow”, “aa”.

That’s what my experiences have brought me. I’m open to any additions!

And also any info, rumors, etc. regarding the official support of French by SynthV.

1 Like
15

Veuillez partager des échantillons audio?

16

Thank you so much for sharing what you experiment :slight_smile: I am a French native and came across the same difficulties

17

Hello, my workaround for the french “r” is: write the word in Spanish and the in the phoneme section adjust the “r” with [ x rr], then put the x length all way down to 20% and tweak the rr length to reach the nearest plausible result.

1 Like