Bruno de Souza Lino

"If I tell you the next piece of music you make it's not gonna play in the radio, nobody will listen to it - your friends, family - no one will notice or listen to it, would you still do it? And if the answer is anything but a clear and concise 'yes,' then maybe you shouldn't because you don't want to. You must make music because you have to and you must have no expectations because, in the end, expectations are nothing but stress. And when you have no stress, you can do anything. And eventually your craft will find its audience." - Steve Vai

SpectraLayers, RX, Spleeter, Acustica and other similar products use the same MusDB dataset, so you'll get similar results regardless of which one you use. Demucs V4 Hybrid tends to produce better separation and was what I used to do the two mashups I posted a few days ago to the forums. The output still isn't perfect and you'll be missing quite a bit of content above 11k as it still uses the same dataset with some 800 more songs added, but it's a different algorithm.

Don't buy anything from Unison Audio. Not that I even did. This is more advice for future me.

I'm only curious why the Mac version is 5 times larger than the PC one. Is it because you have to stop an AU that needs to support both Apple Silicon and x86?

I'm fairly confident everything this plugin does can be replicated with one instance of Sonitus EQ and one instance of Sonitus Compressor as they're describing roughly 98% of what the plugin does in the description. Only thing left for you to find is the amount of slope and resonance of the high pass filter, the amount of boost of the high shelf, the slope of the low pass filter and the compression ratio.

I'm glad you said "most plugins," because that will never be true for anything Unison Audio puts out, no matter how long it has been released.

I think I found one with a whooping 3 million inside:

Give up and get an actual job.

There's two types of music producers: The ones that design the sounds they put into their music and the ones that just need the correct preset. NI targets both with their product line with Komplete. From one side, you have Kontakt and Reaktor, which allow you to create your own instruments and plugins like Tim Exile, which ended with two NI products (The Mouth and The Finger) which were based on his Reaktor stuff he used for live performances. From the other side, you have the Play Series stuff, which are ready made presets you can do minor tweaks on to suit whatever music you're making. The latter used to be less prevalent or non-existent in earlier versions of Kontakt and Komplete, but NI saw the market for the "preset producer" and dug into that.

They also have a 5,000$ plugin.

This is a design flaw with the export dialog, which is track selections on the project override settings and presets of the dialog. That "Entire Mix" category can be completely ignored currently unless you have nothing selected.

Sometimes fit to improvisation doesn't do anything, which can be a fun exercise in frustration. There doesn't seem to be a solution for when you encounter stuff with time signature changes. Taking the tempo division would set it at a tempo that would change when you change the signature. Can't line it with the grid because the grid changes once you change the tempo or time signature.

Breaking news. That hasn't happened yet.

That might be remotely illegal. Stop.

But it costs money and requires time, money, discipline and dedication. Can't ask people that these days. They have jobs and all that guff.

This seems similar to the same situation you have with Audio Damage plugins.

Maybe they used RX to clean those references out.

While RX does give you some plugins your can use in your DAW which replicate some of its functionality, it is fundamentally a standalone piece of software. And it has been like this since RX 6, the first version. Yes, the first version of RX is version 6.

You also have to agree with whoever licensing terms the voice bank you're using have as well. In Solaria's example, you can only use the paid version in commercial projects and attribution is optional (although they probably want you to do so). The terms also dictate how you have to refer to the bank in case you want to, you also cannot use it to create other libraries which use it as an element. You also cannot use it "as the input to any singing synthesis or speech synthesis model, algorithm, application or any data-driven workflow that generates singing voice or speech." (verbatim from their EULA). It's also important to note that both Synthesiser V and Eclipsed sounds will collect your data. While it does say what kind of data it gets, it also says: "If necessary, Dreamtonics or Eclipsed Sounds will collect more information for the implementation of this Agreement. The User confirms the authorization to Dreamtonics and Eclipsed Sounds to do so, and Dreamtonics and Eclipsed Sounds shall keep such information confidential and shall only use such information for the target of this Agreement." (verbatim from the EULA). I don't know about you, but this is remotely not compliant with GDPR, IMO as it doesn't outline exact what kind of data it will collect. As per using the product, don't be surprised if you have to test the thing yourself and determine what works for your use. There's barely any documentation for these, the documentation is not comprehensive enough or some features of the program require external knowledge that's neither explained nor pointed at in detail in the documentation. One of those being which phonetic language or variation of it the software uses, for example. You'll also be hard pressed to find video tutorials of these outside of basic knowledge already present in whatever sparse documentation it has. Think of this as a product made by VOCALOID users to VOCALOID users, much like how most AI algorithms suffer from the same issue as well. Ease of use and intuitiveness don't seem to be a priority for them despite some of them being commercial products.

Which could maybe also work in favor of the "room" treatment if you make it a bass trap instead of just a "solid wall."

Having to manually select what you want to export when you want the whole thing you mixed defeats the purpose of even having presets (this probably feels like the 100th time I'm saying this, even though it's probably the 99th), especially when a simple selection can override them. And then you end up having this tool tip lying to the user: Because you made a selection and what you're getting is this instead, when the simple act of selecting that preset should honor the promise said on the tool tip regardless of what you have selected before opening the export window: And this is a design flaw because you don't get exclusive solo by just selecting a track. If that was the default behavior, then it could be argued that you should be more careful with your selections before exporting.

I suppose this is some sort of deal to make us think their subscription plan is attractive, as this is not part of even the Collector's Edition of Komplete 14, which should in theory have everything NI makes.

Well, I'd think a library called "Fantasy Brass" would be conducive to making brass sounds that may not exist which is what the Arteria one is settling for.

Well, there's Arturia's Augmented Brass, which is priced at 69 bucks atm.

And this considering CbB comes with 4 reverbs (BREVERB 2, Sonitus Reverb, REmatrix Solo and Style Dial SPACE). I'd add only maybe something you can use to load impulse responses (like ConvologyXT or Melda MConvolutionEZ), Airwindows Nonlinear Space and Valhalla SuperMassive. That should give you almost every Reverb flavor available to men.