As you can see. Drag and drop a synth v svp on the vocaloid1.bat, load the midi, and fix anything that maybe be random stuff the file creator left the in the file than press play :) Side, not sounds like now will replace the n aU with n A: . I think it's just a engine thing but, isn't a big deal.

Here's the convert .bat file for vocaloid 1 users : https://www.mediafire.com/file/aehfto82ukq05j6/Vocaloid1_and_OpenUtau_Converter+(2).zip/file

The .bat does convert data utamatix fails to convert so this should help make vocaloid 1 users life so much easier :)

Essentially, this .bat converts synthv/ vocaloid projects. Readapts english phonetics and changed it to how my utau can sing those same words without needing a phonemizer thing. As you can see, the plugin words after a few fixes. The .bat can convert the synth projects into multiple singing styles spicificly tailered based on my voicebank flexibility and vocal range.

This was more of a proof off concept to see if I can use a .bat file to insert actual vocal rock tuning without needing a utau plugin or importing a synthv gen 2 ai tuning exported to utau. Because, this only works with my utau. This will not work with any other utau .

As you can see this simple .bat file converts synthv and procloid projects to usts while retaining all edits done in those software. This was just a dumb test. when you drag and drop a vsqx on the .bat you get a ustx wherever the vsqx is. You can than import the ustx to open utau. This is just a synth 2 utau thing. I had claude do other than that enjoy :) https://www.mediafire.com/file/0zrttby14dk79ok/OpenUtau_Converter+(2).zip/file

Unlike, moresampler this was trained on my utau voicebank with many samples not found in vccv voicebanks. so, this auto ini now has a much larger list of samples it can auto oto for you. As you can tell this will generally do a pretty good job of auto otoing.

When you drag and drop your utau voice onto the .bat the cmd window will pop up , read the wavs, and finish up in about the same time as moresampler if not faster. You will automaticlly get - intro and ending - parts of otos pre done for you. I think if you record in strings like dadedadi the .bat will do a better job of auto oting for you.

Obviously you will need to check your oto but otherwise here's the .bat file if anyone want's it: https://www.mediafire.com/file/acq4ncy3jq32y97/MakeOtoV2.zip/file

I did use claude ai to train the .bat features off my voice how I oto and comonly accepted oto methoids. So this should drasticlly help with making oto's ^.^

This is heavilly trained based on my utau Zannon Koitsu voicebanks. The auto oto thing was created using claude using a mix of what is commonly used when otoing voicebanks, how I oto, how to auto add samples like -a a a- and so on. I'm gonna try to see if claude can do a .bat to make svp and vocaloid projects convert to ustx while adapting the project files to since naturally with how my voicebanks are recorded. I'm hopeful claude could add vocal rock support as a plugin to my voice but one thing at a time XD

KasaneTeto RVC Act 2

Have fun ^.^

I think this should sound more like the utau teto. If not, I'll try a different approach with the rvc training.

Yes after, spending 11 hours training a rvc voice based on my utau voice samples. This might be the best test I done so far. I did add a second track to sound more robotic. But, overall I think the vibrato was correctly trained. So, the voice will have vibrato when there normally wouldnt be to match how the gen 1 pro voice banks are created.

I wanted to created a vocaloid 3 voicebank but that seems unlikely right now even tho I have the devkit. I would try to create a modded voice over Miriam's tho I dont know how to use sms =.= But, I think this sounds close enough to what the voice could sound like if he was in vocaloid or at best synthv.

Here's the rvc files if anyone want's them: Zannon Koitsu improved

I wonder which voicebank people would use more XD

You can deffiantly hear vocaloid 1 really struggling to keep up with all the engine noise lol

But, it's still interesting how vocaloid 1 can even manage to keep up with what I would assume be more like vocaloid 3 edits. Because, vocaloid 1 can not do sounds like consanents on there own in the same way you can do s or z samples in the phonetic tabs. You basiclicly have to treat vocaloid 1 as if it was a cv utau so to get extended sounds you do have to make mini ze ze like samples if you want to extend words like cries or does .

While, Vocaloid 1 is tedious. I still love the old engine quirks ^.^

Obviously this is just a joke XD But, after trying to make an ai Adachi Rei voicebank for diff singer and I forget what else since, my gtx 1050to bit the dust T.T So, this is just a left over mp3 file I had -.-

The fact the ai voice pretty much can only say E e e E ~ cracks me up to no end XD

I left a download link for the mp3 file https://www.mediafire.com/file/r2un8apjg144271/Miriam+butchers+vanity.mp3/file

The cover can be on soundcloud now: Zannon Koitsu Butchers VANITY English cover by Dionn Koitsu

Idk if anyone would want the vocal only stuff. But , here's the low and normal pitch vocals https://www.mediafire.com/file/v86x4netecwi2j9/Miriam+BUTCHER+VANITY.zip/file

Here's the demo of the voice: Zannon Koitsu Butchers VANITY English cover by Dionn Koitsu

If you just wan't the rvc voice files. you can get it here: this used samples of me pretending to be prof snape lol

The vocal style used a second copy of track 1 moved behind the main track by 0.5 seconds.

The song is called life kinda sucks by Kasane Teto

the voice is a bass tenor so the vocal range usage should be pretty flexible. If anyone want's to get the rvc/ index voice files. You can download them here: https://www.mediafire.com/file/q91kmdg7oxndsb9/Calm+male+voice+demo.zip/file

If you wan't a rvc voice model created feel free to hit me up :)

No joke after, all this ai ad garbage started. Youtube keeps pushing bestiality into my ads. I dont watch ai, I barely watch furry content, I hate ad generated videos/ ai generated music.

I've gotten to the point the ai ads are seriously starting to p.ss me off. To the point i just block and report every ai ad I see on YouTube.

I mainly watch gaming , Friday night funkin/ vocaloid music covers. I cant be the only one getting annoyed that youtube is allowing very obvious bestiality on the YouTube app.

I tried reporting this to team youtube on x. An zip nadda. It appears team youtube is ok with sharing illegal content. With the age verification stuff going on. How the h,ll is this even allowed on a service kids can have access to. :/​​

First, you will need applio rvc Release Applio 3.6.3 · IAHispano/Applio · GitHub

scrowl down till you see the following text:

Installation Steps

• Download Compiled Version: Get the zip from this link. Click the blue text on the page to start the download.

Now get stem roller: StemRoller - Make stems, instrumental, or acapella version of any song!

This program will isolate vocals from background audio. This tool is very useful if you wanna clean your vocal recordings or a video game voice recording.

Now, after you have both files. Extract both files at the same time with something like winrar so both files are put into there own sub folders. Now, install stem roller. It doesnt matter where you put it since the installer is a extraction exe tool.

Now, you must have at minimum 5 minutes or records as a wav file. Put the recording into a file. Launch applio. A cmd window will open. Leave it be. Applio will open in your web browser. Once, this happens go to the training tab. You will see a tab with a model name. You will insert your name or a character name here. Copy the folder path to the wav file you recorded. Scowl down on the applio page and look for the dataset tab. Click on it and press control+v this will auto past the copied file directory you copied from earlier. Click Preproccess dataset and wait till the proccess is down.

Scowl down and click extract proccess and wait for that proccess to finish. now go down to the save every epoch and set it to 100 now change total epoch to 300 I recommend this since I doubt everyone has 8+ gb of vram. Click the left arrow to make it a down arrow. click fresh training and checkpointing. This will free up gpu vram. Now go to the bottom and click Generate index. Wait for it to finish which wont take long. Now click start training. This will start the long training proccess of cloning the voice file from the wav database you created earlier. If you lack 4gb of vram your computer will not beable to progress. If you lack a Nvidia driver to use cuda features. You will not beable to continue any further.

All model data will be in you applio/logs/ character name. You will have a index and .pth files. You will later use these for rvc voice changing programs such as rvc gui or real time voice mods. When you copy the two files. In a rvc program look for a model folder. Create your character folder there and past the two files into that sub folder. This way the rvc program will now find your rvc voices to act as a voice changer or voice pitching model.

Now, when you have your rvc model done. When you open applio the next time your rvc model will show up and the voice can use isolated vocals to convert into the voice you want to use :)

Now, load stem roller and select the audio you wish to seperate. When the proccess is done. The off vocals and stem audio will now be found at c:/music/stemroller/ audio name which that folder will have multiple isolated instrumentals and vocal stems that you can use :)

Sorry that this was a bit lengthy but, I hope it helps people out ^.^

When you heard miriam start saying Meeeee~ you'll understand what I mean by I deep fried miriam in the vocaloid 1 editor XD

I know I'm probbly one of few who still uses vocaloid 1 in 2026 with no issues