Synthesizer V AI POPY & ROSE: New Japanese voice databases from Bushiroad Inc. available December 21, 2023

Lounge / 大厅 / ラウンジ
,
1

Two voice databases have been announced for Synthesizer V Studio, POPY and ROSE, releasing on December 21, 2023.

These characters made their vocal synth debut on CeVIO AI last year and are based on the vocalists of Poppin’ Party and Roselia from the multimedia project BanG Dream.

A note for international buyers:

As characters that are part of a large multimedia brand, these voice databases will likely have much more strict terms of use than we are accustomed to.

For users in Japan this will be less of an obstacle, but getting permission to release commercial works may be difficult for people overseas.

If the guidelines match those of their CeVIO counterparts, you will only be allowed to use TuneCore Japan and BIG UP! as distribution services. I do not know if these services are accessible by people outside Japan, so unless the terms are different than anticipated or someone can tell us whether that’s possible, international users may not be able to post their works to streaming services.

Please take appropriate caution before buying if you intend to use these voices commercially and do not live in Japan.

https://x.com/yumenokessho/status/1717538096483541044

(DeepL translation)

:gem: Yume no Yusho latest information :gem:

:sparkles: SynthesizerV AI Yume no Yusho POPY/ROSE release decision :sparkles:

Release date: Thursday, December 21, 2023

Limited made-to-order packages will also be released at the Bushiroad Online Store, and a distribution event will be held in which all purchasers can participate! !

Reservations are only available for a limited time from today to November 15th (Wednesday), so hurry up :blush:

#バンドリ #バンドリTVLIVE #夢ノ結唱

image
image

You can find more information about the original CeVIO voices, including information about the voice providers and character concepts, on the CeVIO wiki:

POPY | CeVIO Wiki | Fandom
ROSE | CeVIO Wiki | Fandom

Additionally, you can find the music of Poppin’ Party and Roselia on most major streaming services:

A demo will be released in the near future. The song is produced by Masayuki Nakano from THE SPELLBOUND (formerly BOOM BOOM SATELLITES).

https://x.com/yumenokessho/status/1717538685531652298

(DeepL translation)

The most recent information about :gem:Yume no Yusha​:gem:.

Masayuki Nakano (BOOM BOOM SATELLITES/THE SPELLBOUND) @BBS_nakano

Produced music will be provided!

Enjoy the collaboration with the singing voice of SynthesizerV AI Yume no Yusha​:blush:.

#BANDLI #BANDLI TVLIVE #Yume no Yusha

Pre-orders are available on the Bushiroad store for the physical editions. The Bushiroad store only ships within Japan, so purchasing the physical editions internationally will require use of a proxy or shipping redirection service.

Digital editions of the CeVIO versions of POPY and ROSE are available from DLsite and Amazon Japan, so we can likely expect their SynthV Studio counterparts to be sold there as well.

POPY’s vocal modes are Soft, Powerful, Mellow, and Joyful.
ROSE’s vocal modes are Natural, Powerful, Mellow, and Ballad.

Full press release:

「いいね！」 5
2

The first demonstration is now available, with our first look at POPY for Synthesizer V Studio!

「いいね！」 3
4

ROSE’s simple demonstration is posted.
https://x.com/ryutarofujinaga/status/1723279400748122427

「いいね！」 4
5

Looking good!

6

ROSE’s demonstration is now also available!

This is a comment from Masayuki Nakano of BOOM BOOM SATELLITES/THE SPELLBOUND:

(translated with DeepL, original text in the video description)

I used up all the supple expressiveness that ROSE’s voice possesses. I created the song with the image of a powerful diva, maximising both pitch and dynamics. Yusuke Kobayashi’s lyrics seem to depict the process of attaining freedom, where one can continue to flap one’s wings to any extent. It is at the same time emo, with physical expressions that transcend human limitations and the illusion that you are viewing an elaborate transhumanist science fiction work.

And because I missed it the first time, this is the comment from Masayuki Nakano regarding POPY’s demo (linked higher up in this topic):

Can we be emotionally involved in the generated singing voice? Can ghosts exist?
Like a living thing, like a non-living thing,
I want everyone to feel and confirm such sensations by touching this software.

It was a good opportunity for me to think objectively about the primitive act of people singing songs.
I felt like I was taking part in an intellectual game, to the extent that I was philosophising about the extent to which the synthesised voice and music have a soul.

I believe that this new voice synthesis software will give creators a great deal of freedom and at the same time provide them with a new venue to showcase their talents and abilities.

POPY is a song that is not in this world, but is a memory of a singer and her views on life and death.

I hope that listeners will enjoy whether these can guide the listener, or whether the listener can relive the experience of a singer who does not exist.

I am excited that when the day eventually comes when we cover this song with THE SPELLBOUND, we will be able to reach the essence of the singer, her voice and her music.

「いいね！」 2
7

POPY and ROSE are now released and available for purchase. The Bushiroad store still only shows purchase options for the limited physical editions (no longer for sale), however digital download versions are available from DLsite and Amazon Japan.

POPY - DLsite | Amazon Japan
ROSE - DLsite | Amazon Japan

Starter packs including Synthesizer V Studio Pro are also available:

POPY + Pro - DLsite | Amazon Japan
ROSE + Pro - DLsite | Amazon Japan

Additionally, the usage guidelines have been published. As part of a multimedia brand, POPY and ROSE are under more strict terms of use than we might be accustomed to from prior Synthesizer V Studio releases.

In general, getting permission to use the products commercially is likely not a significant obstacle for independent musicians (based on the fact that there are plenty of examples of songs on Spotify using their CeVIO AI counterparts), however there is a requirement to use an approved distribution service when publishing songs to streaming services. Users outside Japan may not be able to use the pre-approved services (TuneCore Japan and BIG UP!) and should inquire about alternatives.

If you intend to use POPY or ROSE for commercial works, please review the usage guidelines before purchasing, and consider reaching out to Bushiroad Music if you have any concerns.

If you purchase from DLsite, you must download the installer from the Dreamtonics website: Dreamtonics Download

「いいね！」 2
8

"

how to effectively kills sales outside japan! :frowning:

「いいね！」 2
9

Agreed. I think BIG UP! allows international users to sign up and use the service, but the website is only in Japanese (with very rough machine-translated English…) so it would be difficult for a foreign user to navigate and use the service while understanding the various terms, fees, etc.

Beyond just using the service, I do not know if you would encounter issues cashing out any revenue.

Basically, until someone outside Japan has successfully used BIG UP! and cashed out their earnings (and told us about it), or unless someone has received approval for a different distribution service, there are no guarantees.

I hope that Bushiroad Music sees the international demand due to SynthV Studio’s cross-lingual capabilities and considers more feasible options.

「いいね！」 3
10

I came across this Reddit post, where one of the commenters confirms that BIG UP! does allow overseas users to cash out their earnings via PayPal, while TuneCore Japan requires a Japanese bank account.

ROSE AI for commercial use on Spotify? : SynthesizerV (reddit.com)

It also seems the permission from Bushiroad is granted on a per-song basis, so you cannot get approval in advance before purchasing the product.

Of course BIG UP! is still not very usable for most English-speaking users, since understanding terms of use and fees is not easy when relying on DeepL or Google Translate.

I also don’t know exactly what sort of revenue split is included in the requirement to add POPY/ROSE as a collaborating artist on your song. The reduced earnings may understandably be enough to put people off the product entirely.

This does seem like a standard method of publishing vocal synth music in Japan though, since the list of vocal synths on BIG UP! includes others like Koharu Rikka, Megpoid, and KAFU, but it’s possible that the revenue split is not a constant value and instead based on the agreement between the specific vocal synth developer and BIG UP!. Since the AHS vocal synths listed like Tsurumaki Maki don’t have any restrictions on distribution method, I have to wonder why anyone would use this service if it involves an unnecessary fee or revenue split, but at this point I’m speculating quite a lot.

I’ll keep updating this topic as I come across more information. Naturally it would be very helpful if someone has personal experience using the service and can share their findings.

「いいね！」 4
11

@claire Completely agree that this would be enough of a (potential) headache to just avoid these banks altogether… Especially when there are others around that do not present the same issues…

「いいね！」 1
12

Thank you Claire, your contribute to this forum is amazing. I will never thank you enough.
In this case, thank you for making it clear for me that I’d better not even listen to Rose and Popy, so that I don’t even want to buy them :wink:

I’m writing from Europe, and from my perspective this policy completely kills my interest in these voicebanks.
Whether these distribution services can, or cannot, be used from here, is irrelevant. In any case, those who already publish on streaming services (like me) already have a distributor, and adding a second one, in a language I cannot even type on my keyboard, is surely not viable.
Moreover, there are limitations that are not even predetermined, and which might involve a negotiation with AHS, and eventually signing contracts that will surely be written in Japanese.

Thank you AHS, but I would not go down that route even if those voices were Grammy awarded.

And I would think twice before doing it, even if I lived in Japan…

my 2 cents

13

It’s important to mention that while AHS did the development for POPY and ROSE, they are still Bushiroad products; AHS’s own products are not subject to the same terms.

For most Japanese voices it tends to be that you have two options:

  1. Publish your song through BIG UP! with the vocal synth as a collaborating artist, losing some revenue, but gaining discoverability and causing the streaming service algorithm to put your song in mixes with other musicians who have used the same vocal
  2. Publish without mentioning the vocal synth at all, and keep all revenue

The big difference for POPY and ROSE is you cannot choose option #2.

I certainly have an appreciation for Eclipsed Sounds breaking the trend though. They offer a Spotify profile for SOLARIA and ASTERIAN (and presumably soon SAROS as well) which you can use without needing to pay any royalties or split revenue, while still benefiting from the added discoverability.

「いいね！」 4
14

There are probably bigger cultural differences than we expect. I usually do not mention the usage of Synth Voices (or any other VSTs), because when you talk to people, they think that if something is made using a PC “you did nothing, the PC did all the work”. It’s stupid, and if they only knew how much VSTs are used by the big artists they listen daily…But still it’s this way, so basically stating that you used VSTs disqualifies your work.

And consider that there is not a single “real” instrument in my music…but if I don’t say anything, people give it for granted that someone played guitars, bass, drums…and someone sang…and this makes the song “more valuable”.

I of course don’t subscribe to this point of view, but still, I can’t change the mind of all the people… :frowning:

BTW, out of curiosity I looked for some songs by Rose, and I found a Celine Dion cover…IT’S AMAZING! Now I’m even more sad for all these limitations :rofl:

「いいね！」 1
15

Picked up ROSE. The limitations don’t really bother me as I don’t really write anything original, I just do covers and rarely even publish anything. It’s a hobby for me alone.

She’s fun to work with, very expressive, with lots of power. She needs some work though. Sometimes her K’s turn into G’s, T’s sometimes turn into D’s. When she’s belting that’s less noticeable, and I’d guess the model picked up the ambiguity, but when you use the Mellow or Ballad modes it’s incredibly noticeable.

You can even tell from the waveform!
Screenshot 2024-01-30 211812

So far I’ve mostly noticed this in Japanese, as I’ve not really poked around in Chinese or English with her much.

16

Option 2 not being available seems pretty ridiculous though. I wonder if Image-line, SSL, Sennheiser, Gibson, Celestion, Yamaha, Microsoft and Intel also want a percentage of my money in the future :melting_face:

I was really looking forward to getting ROSE tho

17

I don’t think it’s difficult to understand why a vocal synthesizer would be treated differently in that regard.

A voice is much more unique than just about any other intrument, software, or hardware you might use, and vocals are often the most memorable and recognizable element of a song.

Listeners do not care which DAW or piano VST you use, but the vocals are crucial to whether they will continue listening.

On top of that, these two vocals in particular are just a small part of a large multimedia brand, and this interview suggests that the motivation for making them was largely to market the BanG Dream property, not just to create a virtual instrument.

To summarize, they mention these as the reasons the project was started (and specifically, the reason to expand to Synthesizer V Studio after already releasing the products for CeVIO AI):

  • For use in “the next generation of BanG Dream” (ie. as part of the official media)
  • For people unfamiliar with the property to discover BanG Dream
  • For fans of the voice actresses or series as a whole (ie. to make fan works and music inspired by the series)

Of course, I’m trying to piece it together with machine translations, so if I’m misinterpreting feel free to correct me, but this is my takeaway from the interview.

I would certainly prefer less restrictive licensing (my intention is to explain why it is this way, not defend it), but I definitely can see why some vocal synths are treated differently, especially in the Japanese market where these sort of brands are much more common.

「いいね！」 3
18

Oh no worries, I think everybody prefers Eclipsed Sounds approach in that regard :grin:

「いいね！」 1