Hello everyone, today marks the 7th anniversary of Z Vocal Project. Hard to believe so much time has gone by since I started this silly little site but we’re here! In this article I will go over some recent releases and what’s being worked on.
Introducing the Club Room
The club room is a new part of the site that displays short samples of every publicly available voicebanks (and then some!). They’re all tested with the same song data file whenever possible, and it should hopefully give you a better idea of how each voicebank will sound raw then the demos I have up.
If you’ve been keeping up with out releases, you may notice some voicebanks that weren’t there before 🤔, let’s get into them! I will spare showing demos as you can view them in the Club Room!
Dunder Talk Voicebanks Aplenty!
There are now 4/5 Dunder Talk voicebanks available for you to use. These voicebanks come in the form of the “Dunder from Z Vocal Project” series for CoeFont and Dunder_EON for UniCoe.
“Dunder from Z Vocal Project”
“Dunder from Z Vocal Project” as of current is a series of 3 AI talk voicebanks we’ve made available on CoeFont. They come in the form of an English voicebank, a Japanese voicebank, and a Sorrowful append for the Japanese voicebank. If you have a premium CoeFont subscription, you should now be able to access and use these 3 voicebanks, however I’d highly recommend testing them in the clubroom first as there are some issues.
The English voicebank I recorded with a slight/strange accent, however it doesn’t come across at all due to just how low quality the 500 Yen plan ended up sounding. The normal Japanese voicebank also has this same 500 yen plan quality issue, made worse by the fact I struggled to read a lot of the Japanese. In the future I will rerecord these two on a higher data plan, the only reason I haven’t yet is you can’t just pay for more data on CoeFont, it has to be FULLY rerecorded.
The sorrowful voicebank has 3x the amount of data and it’s the maximum amount of data allowed for these appends. The Japanese was a lot easier to read for that script and I think it’s a lot more usable than the others.
Dunder_EON is a new talk voicebank for the engine UniCoe. This engine uses VCV voicebanks at it’s core, and the data for Dunder_EON is completely new recordings that were auto OTO’d. The engine let’s you use the voicebank in both talk modes and song modes, however I cannot stress enough the song mode is extremely buggy right now and crashes when talk doesn’t.
When in the second tab you can see the translated names for his 4 parameters, I recommend cranking them all up to 1.5 when using to make him sound less slow.
Dunder_EON’s current design is a placeholder, as I wanted to get him out ahead of time because the CoeFont site was acting weird and then my drawing tablet charger broke. Expect a design and OTO update within the next couple months.
Dunder_EON can be used commercially with the purchase of the Dunder UTAU Commercial License.
Dunder Song Voicebanks
When last left off I said I was going to be updating Dunder_CV and creating a sample pack for License Holders. This is still happening, but it’s taking a bit longer than expected due to a lot of stuff coming up in very quick succession (I’m going back to school for my master’s and my cat died earlier this month). I was however able to get some stuff out for you guys. All these voicebanks can be used commercially with the Dunder UTAU Commercial License.
Dunder_Fuyu Act II
Dunder_Fuyu Act II saw his release in July. Although his original counterpart was a whisper voicebank, Fuyu Act II is instead a bipitchFalsetto voicebank recorded in B4 and F4. His new design features him in purple winter gear, with a mask change to match. You can use him in the same ways you’d use a voicebank like Po-Uta, but I cannot stress enough he’s not pitched up, I suffered for him to exist so please use him.
Dunder_Genesis VCV TalkBeta
Dunder_Genesis VCV TalkBeta is the core voicebank of Dunder_EON. Due to EON not having a design yet, Genesis VCV TalkBeta currently uses an image of Dunder_Intercalaris AI as a placeholder.
As mentioned prior, Dunder_EON has not been manually OTO’d, so I would expect some bugs when using this one, but he is currently the smoothest of Dunder voicebanks released. If a proper Dunder_VCV set is worked on this voicebank will probably go under a different name, but for now he’s Dunder_Genesis VCV TalkBeta
Ascenia.EXE Song Voicebanks + Commercial License
It took a bit longer than it should’ve to get them out, but Ascenia.EXE CVVC, Ascenia.EXE CVVC ERROR, and Ascenia.NDS have finally been released! You can also get their commercial license for $10 on their download page. If you already own Dunder you will get 20% off if I’m remembering correctly, however this does not work in reverse so make sure you buy Dunder first if you’re planning on getting both.
I’ve talked about them before so I won’t go into individual detail, but I will address now Ascenia’s RVC voicebank is not coming anymore and I honestly might discontinue Dunder’s due to impersonation concerns. Do not try calling these lost media, preserving them, or creating your own based on their samples if I do go this route, even for private use.
An Ascenia.EXE talk voicebank may be recorded in the future. I tried just pitching up the Dunder_EON samples but it flat out didn’t work for some reason. It will honestly depend if I can do the voice for a long enough period of time to actually record the VCV needed.
Fuuka Shigi S!ckBetas
Fuuka Shigi (S!ckBeta)& Fuuka Shigi – 💚 Append (S!ckBeta) are two upcoming CV voicebanks that represent the voice I want to eventually find for Fuuka. They are a mix of voice acting and formant shifting, and though I originally wanted them out by today, it’s become clear they’ll need some more work to sound right. I hope to release them by end of year.
💚 is meant to be a cute append, and the original was recorded with a forced accent as opposed to unintentional like Ascenia and Dunder’s. The original is the only one I’ve done the OTO for and she’s still sound a bit off. She will probably be receiving new art for the S!ckBetas, but not a new design unless I feel like it.
Plans for the Future
Last year I posted a roadmap that I did not follow at all and I will not do that again. To clarify something based on it, I have not started the search for VPs or a professional engine at all unless you count CoeFont.
What I would like to get done by the end of the year is the following:
- Complete Dunder_CV Act II
- Complete Dunder Sample Pack
- Complete Fuuka Shigi (S!ck Beta)
- Fix the OTOs for old Dunder and Ascenia Voicebanks
- Work on Zealous Villain Project designs for vx_beta
- Find something to replace Shopify cause I ain’t using that anymore.
I need to shade or redo this art, but the priority right now is Dunder_Genesis Act II. he at the very least is going to need one more pitch to match Fuyu Act II, but I’m aiming for more since I don’t have to strain my voice or VA at all for him.
Thank you for your continued support of Z Vocal Project!
I think that’s all I’m ready to post right now, I haven’t worked on the diffsinger or ENUNU banks at all in the past couple months. If I get around to it tomorrow when this post goes up I might draw and then add some 7th Anniversary Key Art to this post. I’ll spend more time on it then last year’s, I have somewhat of an actual art style now so it should look better.