I know you hate it - They hate us too

[sl23]

Just for the hell of it, I went to the sales page. What does thins thing even do? Why would I download something that doesn't even tell me what it is?

Someone mentioned something about samples. Does this create sound samples based on your description? How the hell do you even describe sound?

Does anybody here know what this thing does? I know the OP's not going to tell me.

Funk de fino! No one knows!

Well, it can go either of two ways, we end up in a "Star Trek-esque" future, or an "Orwellian 1984" one! Either way, we wont be the ones that dictate that outcome!

[Innermost]

Just for the hell of it, I went to the sales page. What does thins thing even do? Why would I download something that doesn't even tell me what it is?

Someone mentioned something about samples. Does this create sound samples based on your description? How the hell do you even describe sound?

Does anybody here know what this thing does? I know the OP's not going to tell me.

Funk de fino! No one knows!

Well, it can go either of two ways, we end up in a "Star Trek-esque" future, or an "Orwellian 1984" one! Either way, we wont be the ones that dictate that outcome!

Since I dropped this post 36 persons downloaded my beta version. I imagine that they are starting to understand now, coz' I don't see them coming back here asking questions. They are too much busy to play with than answering your questions. I am here for that.

[wagtunes]

Is there a video that at least shows you using it? Because I'm old and stupid and I'll never figure it out on my own. I admit it sounds interesting.

[Innermost]

Is there a video that at least shows you using it? Because I'm old and stupid and I'll never figure it out on my own. I admit it sounds interesting.

Thanks. Yep, 85 videos on that playlist (Mmm I see that it only shows one video there, I invite you to take a look on my chanel in the Obsidian Playlist, there are all my videos accumulated in one year of work, you will see the evolution of the plugin):

[wagtunes]

Is there a video that at least shows you using it? Because I'm old and stupid and I'll never figure it out on my own. I admit it sounds interesting.

Thanks. Yep, 85 videos on that playlist (Mmm I see that it only shows one video there, I invite you to take a look on my chanel in the Obsidian Playlist, there are all my videos accumulated in one year of work, you will see the evolution of the plugin):

Okay. I'm downloading it now

2 hours? Seriously.

How many gig is this thing?

[Innermost]

Thanks a lot man. That means a lot to me.
Yep I know and the download seems to fluctuate on the VST information windows. Honestly that's something I have to correct coz' you're not the first who tell me that.
What happens at this point:
It's downloading the models from here: https://huggingface.co/innermost47/stab ... edium-onnx
In the app folder.
On windows it's in C:\Program Files\OBSIDIAN-Neural\stable-audio
The models are something like 7gb at the total

[wagtunes]

Thanks a lot man. That means a lot to me.
Yep I know and the download seems to fluctuate on the VST information windows. Honestly that's something I have to correct coz' you're not the first who tell me that.
What happens at this point:
It's downloading the models from here: https://huggingface.co/innermost47/stab ... edium-onnx
In the app folder.
On windows it's in C:\Program Files\OBSIDIAN-Neural\stable-audio
The models are something like 7gb at the total

Okay, cool. Anyway, I watched the one video (You look like you're having a lot of fun) and I already know, as a composer, how I would use this.

Step 1 - Create a main verse track.

Step 2 - Create a chorus track.

Step 3 - Create a bridge track.

Place each track where I want it to go in the song. This will be the song backing track

Step 4 - Write Lyrics.

Step 5 - Sing lyrics

Step 6 - Using my manual tools, create additional instrumental tracks to compliment backing track.

Leads
Pads
FX
Background Vocals

Essentially, this tool becomes a semi arranged drum machine plus. It's up to me to fill in the rest and actually make this a song that most people would not mistake for AI simply because only about 25% of it would be AI generated.

That's how I would use this. And I can see this being a great tool for generating new ideas to work around.

And now I'm going to watch some more videos because like, what the hell else am I going to do while this is downloading?

[Innermost]

Thanks for your feedback Very happy you see the potential.
Haha that makes me think that I should develop a mini video game in this downloading screen

[wagtunes]

Thanks for your feedback Very happy you see the potential.
Haha that makes me think that I should develop a mini video game in this downloading screen

I'm watching more videos. I definitely can't wait to start using this thing.

What was it I said? I wouldn't download this thing if it was the greatest thing since sliced bread? Guess I have to eat a lot of crow tonight. You still came off as quite a douche but I can't argue with the quality of this thing. In the right hands, it can be an amazing tool. It can also be as monotonous and boring as hell when used by somebody who knows nothing about music.

That's why AI will never really take over the music industry.

[Innermost]

I am very happy to read your message.
And I share your point of view a hundread percent:
"AI will never really take over the music industry."
By essence AI is conservative, because trained on past datas
Music is not only music, it's culture. Culture is emotions shared. Mental representations. Culture is Human only. It's not reproductible and worst : it's not predictable.

[wagtunes]

I am very happy to read your message.
And I share your point of view a hundread percent:
"AI will never really take over the music industry."
By essence AI is conservative, because trained on past datas
Music is not only music, it's culture. Culture is emotions shared. Mental representations. Culture is Human only. It's not reproductible and worst : it's not predictable.

Hey, something we agree on. Who'd a thunk it?

[audiojunkie]

Thanks a lot man. That means a lot to me.
Yep I know and the download seems to fluctuate on the VST information windows. Honestly that's something I have to correct coz' you're not the first who tell me that.
What happens at this point:
It's downloading the models from here: https://huggingface.co/innermost47/stab ... edium-onnx
In the app folder.
On windows it's in C:\Program Files\OBSIDIAN-Neural\stable-audio
The models are something like 7gb at the total

Oh! This contains a local model AI. That’s why the download is so huge! That makes sense.

[Innermost]

Hey, something we agree on. Who'd a thunk it?

Nobody would have bet on that
Furthermore on an internet forum

[Innermost]

Oh! This contains a local model AI. That’s why the download is so huge! That makes sense.

Yep. Everything runs locally. No Internet connection needed after the download (After a week downloading from huggingface seems very long)

[Innermost]

24 hours latter no mods came to move my post.
Then, I conclude the following facts:
- Yes a generarive AI based plugin can be considered as a musical instrument if it's designed to give to the user the power to manipulate sounds on the fly, playing and jaming with, mixing live with others instruments (vst, hardware, acoustic, etc), that there's a user Interface usable by anybody who don't know coding.

Then, until proof of the contrary:
- Obsidian Neural is the first to inherit of its definition.
- Innermost is the first artist who use a generative ai Sampler designed for Live and Jamming sessions
- Obsidian Neural, as long this project is an ai generative sampler designed for Live, it will be accepted in the instruments category on KVR forum.

Welcome in the new music era.
Let's all have fun. We are making History again. Isn't that exciting? When you were a child, would you imagine that you would live a new era of something in the music field in your life (well that depends how old you are)?

This thread is now saved in the wayback machine.

Anthony Charretier
France

Here is the first live ever recorded on YouTube with an ai generative sampler designed for Live:


Here is the first known project on Github:
https://github.com/innermost47/ai-dj

You can now refute everything I claimed here, with evidences, and I will update this message if necessary and save the update in the wayback machine.

You can contact me at https://obsidian-neural.com/contact.php or create a discussion on the Github I shared just above.

Wayback urls:
https://web.archive.org/web/20260711203 ... um.php?f=1
https://web.archive.org/web/20260711203 ... p?t=631583
https://web.archive.org/web/20260711204 ... 3&start=15
https://web.archive.org/web/20260711204 ... 3&start=30
https://web.archive.org/web/20260711204 ... 3&start=45
https://web.archive.org/web/20260711204 ... 3&start=60
https://web.archive.org/web/20260711204 ... 3&start=75
https://web.archive.org/web/20260711204 ... 3&start=90
https://web.archive.org/web/20260711204 ... &start=105
https://web.archive.org/web/20260711205 ... &start=120