KVR Audio

Bojo Software · Post by **Bojo Software** » Thu Mar 24, 2022 9:26 pm

I used NI Picked acoustic, EZKeys Dreammachine epiano, a pad sound from Falcon and Chris Hein solo Cello.

Btw. I first heard the song in the soundtrack to Kojimas game Death Stranding.

Spitfire31 · Post by **Spitfire31** » Thu Mar 24, 2022 9:35 pm

@Bojo: That's a lovely ballad! I didn't know the song either, but it's very enjoyable, wothout reservations.

Would it be fair to say that, in competent hands, SynthV has removed the uncanny valley?

Very good work!

Pelicanomicon · Post by **Pelicanomicon** » Fri Mar 25, 2022 9:16 am

Synthesizer V Studio has been updated to v1.6.0! Run your updater to get two new features and the latest versions of compatible voice databases:

Vocal Mode: *SVS Pro only* A timbral preset/customization mode available in the Voice panel. Each database has a dropdown menu and sliders that alter the character of the voice. Vocal Mode currently only works with Dreamtonics' first-party AI voices and Quadimension Stardust Infinity AI, but Dreamtonics is working with their partners to bring it to more voices later on.
The available modes appear to be unique to the DBs. Ryo's Vocal Mode parameters are Open, Soft, Airy, Clear, Nasal and Resonant; Kevin's are Belt, Clear, Soft and Solid; Saki AI's are Chest, Airy, Open and Soft. If anyone has Qing Su or Stardust, please post what their parameters are?
Instant Mode: Allows the editor to execute Auto Pitch Tuning as soon as the user modifies any note or lyrics. Behaviors of the pitch generation AI can even be customized for each note, for example increasing/reducing the pitch dynamics only for some selected notes.

Changelog:
https://dreamtonics.com/en/synthesizer- ... -0-update/
Dreamtonics' Twitter thread:
https://twitter.com/dreamtonics_en/stat ... 9510470663

The Dreamtonics' website has been updated with a link to the store AND a new user-friendly EULA clarification page for first-party voices:
https://dreamtonics.com/en/terms

In other synthetic schoolgirl news, AHS will be releasing Natsuki Karin AI (Synthesizer V AI 夏色花梨) on 13 April. This character is part of the same crowdfunded series as AHS Koharu Rikka and is voiced by actress Takagi Miyu.

Preorders are open at the AHS digital download store (international checkout via Cleverbridge)
https://www.ah-soft.com/product/download.html

Pelicanomicon · Post by **Pelicanomicon** » Fri Mar 25, 2022 10:39 am

Over at BiliBili, Dreamtonics has posted a preview of their new male Chinese AI database codenamed MAN-M2. There's a naming contest open until 1 April, winner will be announced on the 3rd. Like all AI voices, it will be Cross-Lingual

Video Link:
https://www.bilibili.com/video/BV1xY4y1s7Ge

Hope they add it to the Dreamtonics store right away!

ZincT · Post by **ZincT** » Fri Mar 25, 2022 11:02 am

Thanks @Pelicanomicon

While updating SV Pro I also noticed that Kevin has now been updated to v1.02 to support these new features.

Bojo Software · Post by **Bojo Software** » Fri Mar 25, 2022 11:32 am

Pelicanomicon wrote: Fri Mar 25, 2022 9:16 am Synthesizer V Studio has been updated to v1.6.0! Run your updater to get two new features and the latest versions of compatible voice databases:

Nice update, it seems much faster to get good results with Kevin now. Those modes are fun to play with.

Pelicanomicon · Post by **Pelicanomicon** » Sat Mar 26, 2022 6:38 am

Bojo Software wrote: Fri Mar 25, 2022 11:32 am Nice update, it seems much faster to get good results with Kevin now. Those modes are fun to play with.

The Soft and Airy types are a welcome addition, they do sound different compared to tweaking Tension/Breathiness/Voicing. I just wish Vocal Mode had an automation lane like VOCALOID4's Cross-Synthesis feature.

But on the other hand I didn't have to pay for what are basically expansion packs, so

Pelicanomicon wrote: Fri Mar 25, 2022 9:16 am If anyone has Qing Su or Stardust, please post what their parameters are?

From HAYES, Qing Su's Vocal Modes:
https://soundcloud.com/hayes-prod/synth ... comparison

And this tweet w/ examples from @dududanyon says Stardust has 9! modes (including Default? I only count 8 in their video)
https://twitter.com/dududanyon/status/1 ... 4659536898

Why is it so hard to buy Quadimension voices

Pelicanomicon · Post by **Pelicanomicon** » Sun Mar 27, 2022 3:57 am

Eclipsed Sounds has released SOLARIA Lite, a free, quality-reduced version of SOLARIA. It is compatible with both the freeware Synthesizer V Studio Basic and paid Synthesizer V Studio Pro.

PLEASE NOTE: As with all Synthesizer V Lite databases, the lite version of SOLARIA may not be used commercially in any way, and any usages of the lite database must indicate in the upload title or prominently in the description or associated text that they used "SOLARIA Lite"

They have also announced a song contest with physical/digital copies of the full version of SOLARIA and merch as prizes.

SOLARIA is an AI-modelled English-language female vocal based on the voice of Emma Rowley, a Broadway singer.

Download Lite from here:
https://www.eclipsedsounds.com/solaria-original-contest

Download Synthesizer V Studio Basic and other Lite voices here:
https://resource.dreamtonics.com/download/

ZincT · Post by **ZincT** » Sat Apr 09, 2022 6:38 pm

Synthesizer V Studio has been updated to v1.6.1, details on this link...

https://dreamtonics.com/en/synthesizer- ... -1-update/

Pelicanomicon · Post by **Pelicanomicon** » Mon Apr 11, 2022 12:43 am

Dreamtonics' new male Chinese voice, Mo Chen (f.k.a. MAN-M2), is now available!
https://store.dreamtonics.com/product/m ... -database/

Mo Chen is made to fit the characteristics of C-POPs. His mature and stable voice presents a unique tone that stands out in the mid-bass range. The clear and on-point pronunciation makes him ideal for testing out your original lyrics.

Vocal Modes. Mo Chen comes with three variations to his voice timbre: Open, Soft and Clear. These options can be freely combined and varied to provide great coverage for genres such as pop, folk and blues.

Gonna pick him up, I don't have any Chn dbs and I want to hear how the cross-lingual English fares compared to Jpn.

Pelicanomicon · Post by **Pelicanomicon** » Wed May 04, 2022 4:31 pm

Out of the blue Dreamtonics released An Xiao, a new Chinese AI male database!
https://store.dreamtonics.com/product/a ... -database/

The voice database has five built-in vocal modes – Airy, Chest, Open, Power, and Soft – to suit the needs of different musicians. With the Synthesizer V Studio Pro editor, creators can fully explore An Xiao’s vocal potential and make detailed adjustments to his vocal tone and technique to suit the genre and style of different musical works.

BiliBili demo: https://www.bilibili.com/video/BV1iY411w7vm

Like a brighter Mo Chen, hopefully more versatile since it has more vocal modes. I did buy Mo Chen and that is a very dark, soft voice.

Spitfire31 · Post by **Spitfire31** » Wed May 04, 2022 4:58 pm

Thank you for keeping tabs on the voices!

When I saw that you'd posted, I thought, yes! finally CangQion is out! Alas…

I do hope she'll appear this summer.

Pelicanomicon · Post by **Pelicanomicon** » Thu May 05, 2022 6:57 am

Spitfire31 wrote: Wed May 04, 2022 4:58 pm When I saw that you'd posted, I thought, yes! finally CangQion is out! Alas…

I do hope she'll appear this summer.

I check the news every day, hoping for the same! A bit disheartening that there was an initial flurry of demos and then, nothing... My understanding is the Quadimension characters are part of some videogame project that's been in the works for a long time, which is nearing release. So the company is probably focusing on that.

Pelicanomicon · Post by **Pelicanomicon** » Thu May 05, 2022 7:47 am

I've been remiss in taking so long to post about this, but ppl here might be interested in a technique that extracts pitch information from an isolated human vocal and applies it to a SynthV track? The technique uses the freeware speech analysis program Praat plus a SynthV script suite called Real Voice. Real Voice is developed by hataori.

Example:

Hataori's YouTube tutorial playlist:

Bucket's Simplified Praat Tutorial on the official Synthesizer V forum:
https://forum.synthesizerv.com/t/topic/5639

hataori's Github: https://github.com/hataori-p/real-voice
Praat: https://www.fon.hum.uva.nl/praat/

I have not tried it myself, but there's a very lively SynthV discussion on vi-control where hataori chimes in.

Pelicanomicon · Post by **Pelicanomicon** » Wed May 25, 2022 6:50 am

Eclipsed Sounds has put out a teaser for SOLARIA's Vocal Modes! Featured here are "Soft" and "Solid", but the wording of their PR suggests more:

SOLARIA's vocal modes update is nearly complete! For users of Synthesizer V Studio Pro, vocal modes will add a new level of expression to the SOLARIA vocal. You can hear a sample of 2 of her upcoming vocal modes below: Soft and Solid! In this sample, the only parameter adjusted is the pitch parameter, and all 3 samples come from the exact same set of notes - the only change is the vocal mode selected!

Link to Eclipsed Sounds' Indiegogo updates page:
https://www.indiegogo.com/projects/new- ... pdates/all

(In their latest YouTube livestream with Dreamtonics' Kanru Hua, AHS confirmed they are also working on/currently testing Vocal Mode updates. Not gonna embed since it's over 1.5 hours long and in Japanese, but the YouTube ID is k1RFs1_Ky0Y , video title AHS公式生放送第180回 , timestamp 1:29:10 if you reeeally want receipts)

BTW if you haven't checked Dreamtonics' free download page lately, they made Mo Chen's demo project files available. Useful for those of us improving our tuning!
https://resource.dreamtonics.com/download/

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