What’s up guys! My last update was in February and I gave a one week ETA for his completion, so I just wanted to give a quick update on what’s going on with the voicebank.
IT IS NOT CANCELLED!! I am just finishing up labeling now. I didn’t mention it in the last article, but I had went to the ER like two days prior because I thought I had appendicitis, and to make a long story short, this was just the start of like a month of awful gut pain leading up to me realizing one of my headache medications was causing awful side effects, starting to develop horrible anxiety to the point where I had a panic attack for the first time in half a decade, and starting an anti-anxiety med. I fully spent like 2 whole weeks of March sleeping during the day and then trying to sleep at night to fix my schedule to no avail.
Anyways, I’m better now and am truly sorry for the delay, most of the data was recorded and I’m almost done labeling!! I thought I had finished yesterday only to realize Vlabeler actually just wasn’t showing all of the samples.
To make up for the delay here is a raw sample of me singing, I’m still not good at it, this is just to ground you into reality as to what this is gonna sound like.
I’ve also spruced up Dunder’s page a bit. At the time of writing it’s just the demo section but I’m gonna see what I can do for the voicebank one too.
I’ve finally made a Twitter account for the Dunder character as well, @DunderAI
It’s unlocked unlike Lain and Dawns, idk how much it’ll be used but it’s there!
I also decided to add these Early Alpha samples of Intercalaris AI that I chickened out of posting publicly to it. The GAME OVER sample was completely retuned in OpenUTAU, the Someone Else’s Man cover was converted from my SVP. They’re like a slightly updated bank from the one in the last newspost iirc that has some tone labels and correctly labeled breaths. Consonants still sound bad though.
They are still from February and have none of the power data I’ve been labeling the past week, please keep that in mind.
Thanks for Reading!
And that’s the article, told you it’d be quick! I don’t wanna jinx it but as long as he sounds good I may be able to release the voicebank soon!!!! After that I’ll work on Mercedonius AI is what I’ve decided, I was gonna do plus next since that’s not really any work but due to recent shenanigans they have DIFF-SVC locked up.