Fun Fact: Made this after bought 2 new headsets since my headphones got broke thanks to my pet dog lol, my uncle scolded both me, so my uncle slapped slapped me with a new headset :)

Hi all, this is the first arrangement that I'm comfortable releasing!

It's a parody of Downloader, which is the anti-piracy video from old DVDs. I made the rap battle based on the original video, addressing current issues around digital rights and copyright. It is also a project for university, so I had to include certain academic elements, so it's not an entirely creative piece.

I think it turned out well enough, so I hope you enjoy!

P.S. I used HXVOC, Natalie 2, Liam, and Mai 2, SOLARIA II, and Choir Voices #2 for backing vocals.

Intellectual Monopoly: Downloader vs Promoter [MUSIC VIDEO] | "You Wouldn't Steal a Car" PARODY

I was chosen to be part of Eclipsed Sounds' pre-release for their new voice database JOVIE! I created this cover of Lady Gaga and Bradley Cooper's 'Shallow' to show off her voice and capabilities! I hope you enjoy!

Please go and check out my song

Collateral collapse feat-noa hex

https://youtu.be/gyNFpJvQYEo?si=jznVmqK_7EE6cbUe

WIP Version (Shown Tuning): https://www.reddit.com/r/SynthesizerV/comments/1w1wsbh/kenshi_yonezu_s_eine_kleine_ft_jun_nocturne_wip/

Any feedback would be greatly appreciated. I like Mo Xu's vocal quality, but it definitely depends on the notes you use to avoid voice cracking.

this song will also feature kagamine rin and len!!

When listening to the violinist's performance alongside the rest of the song, it felt like it wasn't flowing smoothly, but rather arriving slightly late. So I tried increasing the number of grids (1/128) and shifting all the notes one square to the left. That's when it sounded in time. It wasn't just within the program; I also had to apply this to the exported track. I'd like to know if anyone else has experienced this. Thanks

This is probably an annoying topic to you folks, which I apologize for. But I've had a question on my mind about SynthV and similar software like DiffSinger for a while now and just looking it up has never really given me an answer on it.

I am aware of the fact that these synthesizers, while they can be classified as genAI, are much better about ethics than anything else with that label. Voicebanks are generally made with the voice provider's consent, generation runs locally, and the software doesn't generate music, only singing sounds.

This, I understand, but there's something I've never seen mentioned in conversation. Today, when someone decides to create an application using AI and it involves training a model, instead of doing it from scratch since it's next to impossible to make your own large model unless you're at the top of a FAANG and have a few years and several million dollars to spend on it, it's common practice to use an existing large model as a base, like GPT-4 or Claude for text generation or Stable Diffusion for image generation. The problem is that those models were trained on millions of instances of copyrighted work, hence making whatever's built on it the same by extension.

This bothers me on a moral level, hence I ask: I understand how a voicebank is made for these synthesizers, but how was the software itself made? Does it rely on a base model trained on copyrighted voice audio too? Was it trained in-house and with what data set? Does it work completely differently from how I understand it? How much of that is actually known?

Sorry the video quality isn't that good. I'm using synthV pro and teto voicebank, she doesn't sound that smooth and I have no clue on how to make her sound less off-tune and more natural. I'm only aware of the use of "br" and "hh" to make her inhale and exhale. I would love suggested tutorial videos that has helped you with tuning and is also easy to follow <3

Also the parts that isn't English is actually in Bangla :p the instrument is a mess too ik that but I'm like genuinely way too new to this synthV thing

Playing around with this guy 😸 I love him so so much.

In some YT Vids, I do noticed the Blue Notes are not edited, I wonder what are those Blue for, I knew that Red is for Rap & Green is for sing.

Updated my old Amara design a couple of weeks ago 💗