Audio tools

A personal overview of the software I use more or less regularly for music production, sound design, and related tasks. These are my own opinions, experiences and applications. I am not infallible so it might contain errors, which of course I'll correct if you find any.

DAWs

FL Studio

I've spent the most time with FL Studio, since when I was 14.

Obviously cracked back then (FL Studio free download in my search history), but now I own it legally along with all my other tools. Useful plugins, good for quick sketches, testing out stuff, lame trap beats, lots of electronic sounds. Lifetime license, which is nice. It has a good stem separator (IMO).

It just makes sense to me, as I've grown with it. It's been my only workhorse for a long time. Patcher is worth mentioning: it is a plugin that allows "patching" and routing together other VSTs, creating "your own" synth as a combination of other plugins, with macro knobs. The killer feature is that FL can be used inside another DAW as a VST plugin. This is incredibly useful, because it means I can stick with Bitwig as my main DAW and still use FL anyway.

Bitwig

My new go-to workhorse. Basically it's an Ableton that I actually understand and enjoy. I think it's actually ahead of Ableton for some features, still behind for others.

The Grid is a node-based environment (you'll find them a lot in this page) that allows for lots of creative uses. I find it very good for sound design, probably a little bit more than actually making music. However, it is my choice for live audio and FX routing, because of how easy the interface is to work with.

One notable gap: it doesn't have stem separation. For that I use FL Studio.

Ableton

I only got it because it was on sale, specifically because I wanted Max4Live. I really needed to get my hands on some exclusive devices, namely stuff made by Fors.fm, Kentaro.tools and others. Max4Live devices are a GODSEND for inspiration, sound design, creative FX, etc. and the only notable reason I would use Ableton for.

At the moment I mostly use it for "offline work" (rendering audio through the devices), or with BlueCat AudioConnector to send audio to/from another DAW and an Ableton track, because I don't really like having Ableton as my main DAW (at least for now).

Reaper

Best in terms of price (around 60€ lifetime license), reliability, flexibility, performance. Ideal for "actual" sound engineering use: mixing and mastering, and probably live recording.

If you can code even a little bit, and you know audio processing, ReaJS/JSFX is perfect for experimenting and/or quickly making my own custom processing.

Reaper also allows dubbing audio in videos. A real pro-grade piece of software. I still haven't got myself familiar enough with it, though.

Renoise

Renoise is an incredibly powerful piece of audio tracker with a steep learning curve.

Similarly to Reaper, it is relatively cheap and powerful, with few to no frills, but it will be scary to most people, exactly because it is a tracker. If you come from the piano roll world it will make little to no sense to you at first. You'll have to watch tons of tutorials and make very little progress at the beginning. It is frustrating for many genres and ideal for others (notably drum intensive stuff like jungle/dnb/idm). Moreover, to master it you will need to learn the note commands, the keyboard shortcuts and all the like. It has a lot of useful internal tools, and since recent updates, LUA scripting can be used for example for coding note sequencing.

I haven't used it much. Man, if only I were good at desktop trackers the same way I'm good at handheld trackers...

Synths & Sound Design

Reaktor

Reaktor is a node-based environment that allows users to create (or use other people's) patches that act as sound generators or effects, basically a creative audio playground.

I was initially unsure it would have been helpful since I know Pure Data already, however the amount of useful ready-made devices I found was staggering. I was convinced to get it at some point because it was on sale and because I found some devices that were only compatible with the full Reaktor software (not the player version).

Among my most used devices so far:

• Ruidoz (sample sequencing/mangling)
• S-layer (sample mangling/layering)
• Ultraloop (loop mangling/layering)
• Lazyfish TG-8H (radical audio FX).

I find it incredibly useful for foleys, sound design, creating new loops from existing ones. It is almost mandatory for making IDM.

Vital

I freaking suck at making good sound design with digital oscillators but Vital is really... vital. Not only is it great as a free alternative to Serum, but it would pack a punch even if it were paid software. The only thing I dislike is having to register to use all of its features.

It's great for EDM sound design, subtractive synthesis, unison "swarm detuned" oscillators, trance stuff, etc. I haven't yet spent much time on it... it goes deep and can be overwhelming. But there are lots of tutorials.

Synplant 2

Synplant is a VERY useful sound design tool. I have not yet seen anything close to it in terms of intuitiveness and usefulness.

You feed in some short audio recording, and it will leverage neural networks to perform parameter estimation on its own sound engine to replicate the sound given as input. You basically allow it to "learn" how to recreate that sound using its synth engine. From there you can grow random mutated variations of that sound, visualized as plant branches, and you get a "DNA editor" which modifies the generated sound radically.

Note: it is NOT generative AI! Moreover, it is local, runs on your computer (same thing as STEM separation), and is supposedly not trained on content ripped without consent. So it's a pretty ethical use of neural networks.

You can use this to start with a reference sound and recreate lots of variations on it. It is extremely great for chimes, percussions, stabs, any short sound really, and possibly also leads and basses. Its engine is basically sort of FM-ish, so it will almost always collapse to FM-sounding patches whenever you input something that can't be exactly replicated, like a voice or other complex sounds. In that case you end up with "best effort" recreations, or it could hit the wall and go mad, generating FM hellscape sounds. Great for quick discovery of sounds starting from references.

VCV Rack

I had little use for it so far, however it is great. I also made a couple of experiments creating plugins for it (Calvario). It will surely work great with my ES-8, as I could potentially output CV signals from it. Very good for sound design, random messing around, the unlimited modular synth experience. Especially good for testing modular things out before buying something eventually.

The free version is great for just playing around and recording. The paid version includes inserting as VST in a DAW, which I almost never need. One caveat: you have to navigate through a lot of YouTube videos, VCV library, check what others are doing to find what you might want to use. Also there's the issue of being stuck not knowing what to do, because downloading large libraries leads you to "own" lots of modules and you never know which one to use. Start slow, don't download everything.

Cardinal

A good VCV Rack alternative. I haven't explored it yet.

PaulXStretch

Paul Nasca! The man, the legend. This algorithm is simply epic. Moreover, it is the reason this video exists: Justin Bieber slowed 800%. I owe so much to this tool. It's one of the best creative uses of a phase vocoder and has become a standard in sound design. PaulXStretch is basically the "PaulStretch" phase vocoder algorithm in a good GUI.

You throw in a sound, any sound, and stretch it to the extreme using an algorithm that preserves spectral (tonal) information, trading off transients (percussivity), making everything sound smeared, lush and ominous. It's the one-for-all recipe for turning anything into ambient.

Valhalla Supermassive

Needs little introduction. Simply one of the biggest sounding algorithmic delay/reverbs of all time, and it's free. It can turn your small plucky sound into a WALL of sound. It's actually quite hard to make it "stay in the back" and not steal the spotlight.

Airwindows

Chris Johnson is a legend. He maintains a suite of free FX plugins called Airwindows. The imperative here is no distractions, no fancy GUI, no bullshit, just plain functionality.

They are mostly creative takes on stuff that is otherwise kind of ordinary and boring (channel strips, drum bus processing, dynamics, etc.). Check it out if you're aiming for a "new sound" when mixing/mastering.

OSTirus / JE8086

IF you own an Access Virus device then you might be legally entitled to use OSTirus. Which is basically an emulator, and as most emulators, it follows the rule of "if you own it, you can emulate it". It emulates a Motorola DSP, and that allows you to run the Access Virus devices firmware and have the editor in VST format. I can't help you in finding the firmware but I just know it's out there somewhere.

Same goes for JE8086. IF you own a JP8000/JP8080 you might be legally entitled to use the JE8086 instead. I can't help you in finding the firmware but I just know it's out there somewhere. These two synths are simply the best and most immediate approach for EDM and trance goodness.

Node-Based Environments

Pure Data / Plugdata

Pure Data, my beloved... Pure Data is another node-based audio playground for generating or effecting audio.

Probably the most notable one, at least historically. I have never made anything remotely stable or really useful in the long term in PD, but it has always been very useful to learn audio stuff, especially years ago. There's lots of didactic content.

By using Plugdata instead of "vanilla" PD, you mitigate the harshness of the tkinter interface, replacing with something more user-friendly, and it adds "musically useful" plugins. Plugdata can be used as a VST, which makes it very effective for designing your own processing, similar to ReaJS, but node-based, hence somehow more user friendly.

It's also very useful because of the heavy compiler for creating/porting algorithms to hardware devices (e.g. the Daisy ecosystem), but I haven't explored that yet.

Max

Max is another node-based environment. Notably, its history is intertwined with that of Pure Data. Max vs Pure Data is kind of the Apple vs. Linux debate. They are both extremely useful and capable, but appeal to different demographics.

I'm not much of a Max fanboy, however I recognize its value in some areas. Max excels at user friendliness, but it's paid. It allows for Ableton integration, but it's paid. It has better "out of the box" documentation than PD, but it's paid. You get the idea. It also does a lot of ML-oriented stuff.

Note: IMO, Max is currently very much challenged by Plugdata in terms of user friendliness. Plugdata has closed the gap in many ways, plus it's free and open.

CDP/Soundthread

Recently some guy has started developing a GUI for the Composers Desktop Project (CDP) that is a sort of individual audio mangling tools that run in the CLI.

Using Soundthread makes it easier to perform this form of sound editing using the various tools. There are several useful CDP tools currently supported (over a 100), notably the spectral and PVOC stuff, but don't expect the intuitiveness of modern interactive node based environments.

This shit is free and was used by AFX, though, and this is a great selling point for many people, but don't expect to have it easy.

Chiptune

LSDJ

I wouldn't be writing any of this without LSDJ. I started listening to electronic music because of chiptune, and LSDJ was among the first tools I encountered. I kept on making music because of LSDJ. I am who I am because of LSDJ. It is THE gameboy tracker, no objections. You can make incredibly powerful and detailed music and I have been just one of its most humble users, scratching barely the surface despite all the years spent on it.

If you don't like the real gameboy approach, run it in an emulator. To record tracks, it works best doing it inside bgb (with the split tracks WAV recording functionality), or use RetroPlug, a VST plugin to run and sync an LSDJ emulator in your DAW.

If you're afraid of dealing with trackers, you can resort to Fors Junior, an LSDJ-inspired synth that is able to reproduce most of the GB soundprint, and more.

Nanoloop 2 / Mono

I make most of my GB music on Nanoloop 2 nowadays. It's an FM synth with a minimal interface and unique sound. I use it exclusively in pattern mode as I'm not a fan of the song mode.

I suggest using it on an OG Nintendo DS (should have the best sound hardware-wise, but I don't have any comparison proof).

I also use Nanoloop Mono, which is great, totally different sounding as it has a completely different sound engine with analog filtering: raw, punchy and gritty.

Milkytracker

I've used it sometime.

Audio Routing & Utilities

VB Audio Matrix (Coconut)

The perfect tool to handle aggregate audio devices in Windows and route inter-application audio streams. It works by creating a "virtual" audio I/O matrix that sends and receives to/from multiple actual audio devices (WASAPI and/or ASIO). For example, in my setup:

It can also stream audio in a LAN context, but I didn't use that yet.

BlueCat Audio Connector

Similar to VB Audio Matrix. The advantage is that it can be inserted directly at some point in a track, can carry MIDI too, and can be more intuitive than VB Audio Matrix, though essentially it does the same thing. An alternative could be ReaStream, but I've never used it.

MIDI-AbletonSync Translator

A custom Python script I made for a very specific purpose: It creates an Ableton Sync signal synchronized to a MIDI sync signal, allowing devices that only support Ableton Sync (the case of Virtual DJ) to stay synced to MIDI. It leverages the great Carabiner library for dealing with Ableton Sync.

Virtual DJ

I used the free version when I was little. More recently I bought it by impulse, to use it with a cheap DJ controller, because it has great and fast online stem separation.

My aim is to integrate it with the rest of my setup to make "instant" or short-planned remixes of songs by stem separating and playing back the vocal part, staying synced with my hardware and the rest of my software.

Sometimes I sort of regret buying it because DJUCED is cheaper and probably has comparable stem separation performance in terms of speed/quality. I use it with either a Traktor X1 MK1 (used) or a DJ Control Inpulse 200 (used).

FFMPEG

This solves a LOT of batch processing things you would go mad dealing with otherwise.

Say, you want to convert the entirety of your sample library (that is maybe several GBs of WAV in various representation formats) to a uniform "WAV 44.1kHz 16bit" format, so that you can load it in your hardware sampler. How do you plan to do that without FFMPEG? No kidding, here's a script:

fd -e wav | split row "\n" | par-each { |f|
  let p = path dirname
  mkdir $"../converted/($p)"
  ffmpeg -n -hide_banner -loglevel error -i $f -sample_fmt s16 -ac 2 -ar 44100 $"../converted/($f)"
}

Moreover, FFMPEG deals with a lot of video encoding tasks. I usually record stuff with OBS in MKV and then rescale/compress it using FFMPEG. It has some of the greatest support for powerful CPU instructions or GPU acceleration, and it's usually still a one-liner in the CLI.

If you're a power user (not scared of command line) and want to automate batch audio processing tasks, FFMPEG is the go-to.

Audacity

A great piece of general-purpose audio editing software.

I started hating on it since when it was acquired by some private company, which is slowly turning it into no-longer-free-as-in-freedom software. I prefer to stick with version 3.0.0 (no networking, no telemetry, no proprietary stuff, prior to all this shit). Generally I'm not against using closed source software, but seeing a free tool turn into something no longer free is sad. Plus, it actually became really bloated since then.

I use it as a tool that should deal with the basic audio manipulation stuff. No extra shit. I don't need VST support if I'm using a DAW anyway. Trim, cut, mix, stereo separate, normalize, basic filtering, noise removal, batch processing etc. It should just stick to that, IMO.

Resonic

It looks like a great sample library browser, better than browsing sounds through a DAW, at least for me. I'm using the free version at the moment.

MiniMeters

Great audio visualizers. Not in VST form, it sticks some mini signal meters onto your desktop to show many values of interest when producing. To me, the most useful are the LUFS meter, the stereo field visualizer, and the frequency and waveform visualizers. Good to keep a literal eye over your audio, and as a visual aid to mixing/mastering.

Loopable Audio Tool

A Python thing I made. It's buggy but sometimes useful. Used to take in a WAV file and search for "good" loop points inside of it using various approaches (Manual/FFT/XCorr). Has experimental features like volume companding, amplitude modulation removal, timbre matching, etc. It does a decent job in helping the search of loop points, but in general it never got to the point of working flawlessly and making miracles.

Good in combo with PaulXStretch, as it makes it very easy to experiment in designing the start and end point of looping pads.