Vibe coded plugins

[VitaminD]

It's just another synthedit era. Most will flash in a pan. A handful will grow something meaningful. It's not that big of a deal in the current form.

[Funky40]

Request: somebody who actually can code wants to team up with me ?

I can`t even vibe code. But i can tell how to make -a in fact very simple- Delay that surpasses everything on the market in some specific area for some specific usecase. Exactly how we would patch in HW modulars.
But that Delay -if donne well- will even outperform the eurorack delays, the ones totally suited for that usecase. (say: Eurorack: everthing donne 4 years back and further back)

The idea would: vibe code it !.....only vibe coding ! say: 98%+. As a test. We could proof something vs. some nonsense that has been posted here.

The point here: it`s not even necessary to create "new functionality" to create GREAT plugins.
Sometimes the stumbling blocks lie somewnhere else. Small details.
....and nobody in that effin Delay market has "some specific 1-2 points" right. They all overlook it from what i can tell.

As i said, a simple Delay. No fancy tapps or anything: Buuut: it would require VERY accurate linearisation/delinearisation of some parameter values vs. how these would react vs. a "knob movement". It`d be just that ! The whole clue would lie just there.....with a little twist under the hood ( i`v e also never seen "that" incorporated anywhere in a delay)

I can patch / can you vibe code ....and then overlook that code ? Want to do a experiment ?
no financial goals.

( only one thing in this case: i am on Mac. It had to be Mac compatible.....AND: stable, good code !!)

anybody ?

i tell you the concept, you tell the computer the work she has to do

[whassup]

Request: somebody who actually can code wants to team up with me ?

That's not me.
But I like the idea.
This would prove (again) that good and well executed ideas always beat the "system".

So popcorn!

[audiojunkie]

I’ve noticed a lot of fly-by-night developers putting out sub par products of late. Seems anybody can make a plugin now… Just not very well.

Are we going to see an analog revival, and go back to more hardware, built with talent by humans. Personally I’d love to see it.

Hardware these days is largely plugin software run on top of Linux. It may be a solid foundation (Linux), but you never know what the plugin software on top is.....even if it's built into hardware--could be vibe coded.

[IvyBirds]

I’ve noticed a lot of fly-by-night developers putting out sub par products of late. Seems anybody can make a plugin now… Just not very well.

Are we going to see an analog revival, and go back to more hardware, built with talent by humans. Personally I’d love to see it.

Along with the crap there have been some really good plugins that have come to market of late, that has always been the case however

As for the so called analog revival that has come and gone and peaked several years ago. The near bankruptcy and then sale of Moog when they were selling $10,000 poly Synths and $5,000 mono synths really put a bookend on it, if it ever existed at all

Seems like most of the hardware synths that get interest these days are digital and plugins in a box.

And no doubt many of them are running code that was created with AI

[_leras]

It's just another synthedit era. Most will flash in a pan. A handful will grow something meaningful. It's not that big of a deal in the current form.

There's quite a barrier to making audio tools. You need to handle multiple format types on multiple types of computer, multiple audio rates for calculation, incoming audio and midi, audio dsp, graphical UI (even if simple) etc

Things like Juce have no doubt helped a lot and made this simpler, and I believe clap can also help by being the only target format (as it can be wrapped) but that is still a huge amount of work before you even start working on the idea itself, or to complete afterwards to make a full release.

If ai tools can take away that plumbing part, like Juce, (and realtor and synth edit do to an extent), then whats not to like? More people will be inclined to make things.

And who cares if not everyone goes on to multiple releases. Many would open source and share code.

A lower barrier to entry can only be a good thing.

[_leras]

I’ve noticed a lot of fly-by-night developers putting out sub par products of late. Seems anybody can make a plugin now… Just not very well.

Are we going to see an analog revival, and go back to more hardware, built with talent by humans. Personally I’d love to see it.

As for the so called analog revival that has come and gone and peaked several years ago. The near bankruptcy and then sale of Moog when they were selling $10,000 poly Synths and $5,000 mono synths really put a bookend on it, if it ever existed at all

Seems like most of the hardware synths that get interest these days are digital and plugins in a box.

And no doubt many of them are running code that was created with AI

What are you both on about? 'if' we see an analog revival...

There have been as many or more new models of analog synths put out on the last 5-10 years, than perhaps in all the time before that.

I doubt many of the hybrids are running code created by ai. AI will no doubt help in the future, but it doesn't do anything new it just helps write out existing code.

If anything helps digital synths it is the cheap raspberry pi type processors people can put inside them these days.

[IvyBirds]

I’ve noticed a lot of fly-by-night developers putting out sub par products of late. Seems anybody can make a plugin now… Just not very well.

Are we going to see an analog revival, and go back to more hardware, built with talent by humans. Personally I’d love to see it.

As for the so called analog revival that has come and gone and peaked several years ago. The near bankruptcy and then sale of Moog when they were selling $10,000 poly Synths and $5,000 mono synths really put a bookend on it, if it ever existed at all

Seems like most of the hardware synths that get interest these days are digital and plugins in a box.

And no doubt many of them are running code that was created with AI

What are you both on about? 'if' we see an analog revival...

There have been as many or more new models of analog synths put out on the last 5-10 years, than perhaps in all the time before that.

I doubt many of the hybrids are running code created by ai. AI will no doubt help in the future, but it doesn't do anything new it just helps write out existing code.

If anything helps digital synths it is the cheap raspberry pi type processors people can put inside them these days.

So I say the Analog Revival has come and gone and your counter argument is taking about Synths from 10 years ago

Great job!

And the winner of the so called Analog revival is a German company that sells analog poly Synths cheaper than a copy of Omnisphere because nothing says analog revival like having to sell analog on the cheap while Moog goes tits up

As far as the AI Purity tests, the AI Amish have created the environment where developers and designers who build hardware and software will use AI and just keep quiet about it, because there is no advantage or incentive not to

And Hybrid Synths are not analog, they are digital and generate sound with ones and zeros

[kaamos]

I am making this. It is AI assisted. It is not vibe-coded. A lot of very deep and detailed human work has gone into this project. The computer is my mathematician, my babelfish, and my librarian. I am the human who has made this artistic creation and you are invited to join me if called to the void by its echo.

Teen tätä. Tekoäly on apuvälineeni — ei tekijä.
Tämä ei ole ns. "vibe-koodattu" projekti. Ihmistyötä on tehty paljon, syvällisesti ja yksityiskohtaisesti. Tietokone on matemaatikkoni, babelfishini ja kirjastonhoitajani. Minä olen se ihminen, jonka käsistä tämä on syntynyt.
Jos tyhjyyden kaiku kutsuu sinua — olet tervetullut.

https://github.com/jmcgill-public/kaiku

[Seafire Mk2]

Devs have been copy-pasting code in the wild for years. This is just the next evolution.

Indeed. Its the same as open source plugins.

[TechHaus]

Devs have been copy-pasting code in the wild for years. This is just the next evolution.

Indeed. Its the same as open source plugins.

No.

[] Peter:H []

Just little update.
1.) I love the distinction between vibe-code versus ai-assisted that was mentioned above.
2.) I actually really have started the moog filter experiment which I mentioned above, https://github.com/huberp/phu-moog-filta
It compiles on windows and works in bitwig ... it still need "bit" of work until a 0.5.0, algorithm, code, ci, releasing.
It was actually mostly only done writing instructions from my mobile phone during away from home for a week. I work openly, you can peek into my PRs, Issues, ...
3.) Fun and more elaborate example of my ai-asissted endeavours is https://github.com/huberp/phu-beat-sync-multi-scope

[Funky40]

The flood of new releases *right now* is remarkable !

Three more until end of next week and somebody should start a "new Subforum request" thread, haha.

Seriously, it`s nolonger funny.


It will cause damage to this market i think.

[Urs]

The flood of new releases *right now* is remarkable !

I just saw a video where someone was selling trivial stuff with utter confidence. I guess AI is a Dunning Kruger amplification system. People with no idea start making uninspired stuff and sell it as if it was the hottest shit on the planet.

[Funky40]

People with no idea start making uninspired stuff and sell it as if it was the hottest shit on the planet.

But this is to a certain -much lower degree of course- also true for common established Plugin makers. YMMV

And the Flood of plugins is not only related to newcomers.
The established guys are throwing it as well.....can`t be a coninsidence either.