I’ve noticed that on every AI site that creates new songs, the result you download from various attempts (MP3 in my case) is a file/song where all the elements/instruments are very muddled together, and, above all, the file is very, very, very compressed.

Well, it might sound strange, but that’s exactly what I’m looking for…

I don’t know which plugs those sites use to do that, so my question is precisely about this: “Does anyone know of a specific setup that can produce exactly that kind of blending and compression, so I can achieve that same sound on my Bounce Bus?”
I need it for an audition…

Thanks to everyone for the advice and attention.

P.S.: Please, gently, specify the plugins and settings you use (better with a screenshot).

 

Last edited: Jul 15, 2026

haha strange, but with an specific audio this is hard to get done... gonna try saying that use the fastest compressor u have, and push the audio minimum 4db, from there to 12db would be crazy but in that range u can achieve what u r looking for. the attack and release vary for the audio source... but again, it depends, i would never use a fast attack on a bass audio but maybe its what u need... and remember u can use https://mvsep.com/es to get good separation stems and hear them separately so u can deduce with fresh ears how far u can or need to compress any audio file u treat.
FET compressors maybe are the one u need, its the most fast, but if its too much than u need, u can search a little back.
again cant be specific even with an screenshot cause all audio files even hard compressed are different to treat.

 

My mind went to UrsaDSP Boost into God Particle and I can see Garuc Audio MacHead in it too.

 

during the last 2 years or so the questions asked in audio forums are getting worse and worse.

 

Just put your audio through a lot of plugins, add aliasing and maybe export it in 192kbps, print it on a CD, pee on it and throw it to the trash.

 

Or one of those LANDR plugins, Musik Hack Master Plan, SSG Optimus. Any of them would be helpful if all you want is to wreck a mix.

 

The incredible thing is that hundreds of these sites are springing up here and there like mushrooms, yet we know NOTHING about how they put their recipes together... (plugins, FX, hardware, vocals, backing vocals and so on). Don’t you find that strange?

Nevertheless, there must be loads of people working in this massive industry nowdays… and not a single one to spill the beans?
Isn’t there even a single person who’s been sacked and is leaking information out of spite? Hmm…

And yet, to become so popular so quickly, they must be getting their ‘know-how’ from somewhere... but where?

Please, only provide useful answers... anything else is unnecessary and a waste of time.

 

Last edited: Jul 15, 2026

Maybe provide short audio snipped we can process, instead of just shooting random settings and name dropping stuff for unknown material at hand.

 

It's a moot point.

Everyone bangs on nowadays about the mixing and mastering process, which is important, but the most important thing is the music itself.

Where's all the great music gone?

 

If this is not just trolling:
AI music is not mixed, it‘s generated. No plugins involved.

 

AI music is incredible. It will only get better.

Suno is the only platform that I'm aware of that supports your input and AI creations. Suno offers an "AI Studio". Look into that.

 

Aberrant Digitalis or Lese Codec

The responses too.

this is it. lol

its no secret how its made. They have the worlds leading mixers on call. Serban, CLA, all those guys just run the AI stems through their template.

 

I understand, but this is a post about Mix, not about how to write and record great music (if you have a look on the Forum, I'm sure, you’ll find about that).

Anyway, two nights ago I was talking to my dad and I said to him: «Unfortunately, we’re losing the battle against AI and… we’ll lose. We’ve won a few skirmishes here and there, but… we’ll soon lose the war.»
He looked at me with a sad expression without replying.

EDIT: Now, rereading my own words, it’s as if I’m seeing scenes from the film ‘The Terminator’ all over again. LOL

 

Last edited: Jul 15, 2026

those AI tools are generating spectral images, so you can't really mimick it in casual mix process,

speaking of mixing/mastering, there are various limiters and clippers,
take a look at for ex. what modules TDR Limiter 6 has:

 

This is what I use.

Clip to -8 momentary.

Limit to -8 short-term.

 

That's a brave man right there

Imagine the answers... all considered, could have been much worse

 

you are already the second person who said that, but it is not true.

those online mastering services basically use AI driven audioprocessors just as the ones you can use

locally.
it is not like image restoration which generates a new file.

 

AI music is generated, not mixed (like @timer said).
There are no plugins, no template, no synths ... just audio.

Some reading :
https://sybase91.medium.com/suno-th...erate audio token by token using transformers.