SynthV Synthesizer V Discussion Thread

[___]

So uh...hate to double post and hate to be that person, but I feel like this still needs to be hammered in unfortunately.

Has anyone out there tried to get a refund for Eleanor AI's box and been successful in receiving it?

 

[peaches2217]

So uh...hate to double post and hate to be that person, but I feel like this still needs to be hammered in unfortunately.

Has anyone out there tried to get a refund for Eleanor AI's box and been successful in receiving it?

Honestly I was just wondering this a while back too. I would guess no, given they’ve said exactly one thing since April and it was about Vocal Modes, but I feel like it’s time to start addressing it in earnest.

 

[___]

ok an update! I could find one instance of fellow member @Krathia getting a refund:


would be really interested in hearing other people's experiences!

 

[PearlStarLight5]

AH-Software has Yuma and Natalie too
HOLY SHIT ARE THOSE THE CHINESE VOCALS WAS NOT EXPECTING THIS
Don't mind Kevin, also I noticed no Weina
They're not on DLsite but this might mean Yuma and Natalie will be added there soon.
And perhaps we can expect physicals for, ALL of them?

 

[Cluemily]

The video for the 1.8.0 release just got uploaded! It seems to be a beta, so you'll have to manually update voices if you want to test things out.
Alongside MIDI keyboard and metronome support, they've also added note level cross lingual synthesis, so you no longer need to create new groups/tracks for switching languages. They've also added Diffusion Probabilistic Models (DPM)! Their release notes explains it all in details, along with lists of VBs being updated to support new features.

Mai is also free for all SV Pro users; the installer is included in the release notes.

If you've registered your VBs to AHS' MyPage, you can get the beta versions for those listed in this Tweet.
Eleanor's vocal modes are also included in this update; all eight of them. Rikka's are also added!

From a quick test this is like a massive improvement. Any issues I personally had are basically fixed, and pronunciations are sounding much better overall. I appreciate them listening to the feedback.

 

[morrysillusion]

they've also added note level cross lingual synthesis, so you no longer need to create new groups/tracks for switching languages

oh man, how nice. i know myself and others were thrilled to see thats how it worked in V6 when it released, so i guess its no surprise Synth V followed suit. definitely curious about this DPM thing and its good theyre considering this a beta first after, well, the hdvm stuff. really cool to hear Mai is available for free too, and curious to see what Ele's vocal modes will be like. this sounds like a real good update overall.

also it looks like in this update hdvm is set to 0 by default now. the UI changed in general to where the slider is now at a middle point marked 0.

EDIT: oh... erm, wheres eleneaor's beta installer? i know people have used the vocal modes. saw someone on twitter say it was removed from the dreamtonics page??? and i definitely dont see it there.

 

[lIlI]

Having a free voicebank available with Pro is a very good move, I'm looking forward to trying out Mai! I'm seeing a few bugs reported, so I'll probably wait for the non-beta release to try it out. It's great to see that Dreamtonics is continuing to listen to feedback.
Volor, in classic Volor style, has pulled the Eleanor update due to a miscommunication. It's still being tested, and they have demo songs on the way.

 

[blackout]

I really am liking Mai and she might be my favourite of the alphabet so I’m glad to see that she’s free as a Pro owner haha, this seems like a good update for SV this time around!

 

[___]

1st the wrong version of Chifuyu lite, with vocal modes and crosslingual gets released and then this rtyu

 

[PearlStarLight5]

Finally Volor tweets publicly again and it's a fuck-up (and still ignoring the boxes), what a world!
Anyone have any data on what Eleanor vocal modes were?

 

[lIlI]

She had 8:

Melancholic
Solid
Warm
Powerful
Dark
Clear
Tender
Bold

Here's a good comparison of Solaria pre and post update (Feng Yi is also in the same thread!)

 

[PearlStarLight5]

Dreamtonics updated their website, now EVERYONE IS HERE

Synthesizer V | Dreamtonics株式会社

dreamtonics.com

About time, this is seriously really good for their marketing

 

[Maxi]

I'm really enjoying the new update so far tbh
Also the fact HDVM seems to be defaulted at 0 and you can enter negative numbers for it now!

 

[lIlI]

The overal reception to the 1.8 update has been very positive, with voicebanks generally considered to be at their best. However, there's some downsides, which I thought I'd compile here for everyone's awareness and convenience:

1. Phonetic strength no longer functional on AI voicebanks
2. HDVM retakes no longer functional
3. Chinese to English crosslang quality reduced (engine based, cannot be fixed by downgrading to an older version of the voicebank)

 

[___]

The overal reception to the 1.8 update has been very positive, with voicebanks generally considered to be at their best. However, there's some downsides, which I thought I'd compile here for everyone's awareness and convenience:

1. Phonetic strength no longer functional on AI voicebanks
2. HDVM retakes no longer functional
3. Chinese to English crosslang quality reduced (engine based, cannot be fixed by downgrading to an older version of the voicebank)

Some updates and changes are worthy of praise, some are worthy of critique, but overall the sentiment I share is that I wish buying SynthVs felt more like getting professional consistent product and less like sign up to be it's perpetual "beta" tester.

 

[DefiantKitsune]

Also heard reports some banks like Natalie have audible room noise now

 

[lIlI]

Worth noting that these issues are related an opt-in beta release of 1.8, and the update hasn't been pushed to all users yet. I hope the bugs get ironed out before the new version is rolled out to everyone officially, but I wouldn't be surprised if it took until the next major release (depending on how technically complex they are to solve).

 

[peaches2217]

It’s fair when these issues are present in beta, but nearly every new release in the past year has done something to throw the engine and VBs’ quality all over the place. It’s so painfully unstable, and with each new update you never know what’s gonna be affected next. I do hope the open betas help mitigate those issues, but that’s part of the reason I just haven’t used it much in so long: it’s just so unpredictable.

 

[___]

Nevermind we wouldn't even have those beta releases if Tokyo6 wouldn't have given DT slap on the wirst like a little kid.

edit: should specify that my use of the phrase "beta tester" was used figuratively to say that the products are constantly changing in their nature, on the whole scale of negative to neutral to positive.

 

[morrysillusion]

i have to imagine that issues like HDVM retakes no longer functioning (tho it seems to work for me) would definitely not be there final release, it is a beta after all. much better we're opting in for it than not tho. i know myself and others are always way too eager to see these new things even if theyre broken, so i dont care all too much as long as theyre getting reported and that the final update pushed fixes those issues- which we will have to wait and see how that goes. hopefully dreamtonics can leave a good impression w this attempt at a beta release