Curious what notable personalities in the scene think was great this year? So is TGEFM! So we reached out to some of our favorite luminaries ranging from musicians, label personnel, and more for their “Best of 2023” lists. Now, listen: TGEFM is not a taskmistress. Contributors can write these out however they want. So if it doesn’t actually look or read like a list… and sometimes it really is just a list with no other observations! Who cares?
HU$H invites listeners to take refuge in a hypnotic hybrid of grunge, hip-hop, and electronic music underscored by his own unbelievable triumph of will. The vocalist, multi-instrumentalist, and producer has survived an often-brutal upbringing marred by bullying, homelessness, confusion, and alienation to emerge as a dynamic and disruptive sonic insurgent. Signing to Strange Music and generating nearly 20 million streams and counting, he provides cover from life’s darkness via his 2023 LP SHUNKWORKS. Today HU$H shares their favorite music technologies of 2023
First off I wanna thank you for asking me to write this up. The first thing that came to my mind was music technology. This is the first time in a long time that I’ve seen any major changes in music technology. Some of it is terrifying and some of it is just boring and I didnt even notice it was happening, but all of it together is going to create really interesting new ways for creating and performing music.
Midi 2.0 Took Flight
So midi 2.0 dropped in 2020. This year, Native Instruments joined the world of MIDI 2.0. In my mind, this solidifies its existence in the future of where music technology is going and means other software companies will follow. Without getting too technical, here’s a few points about MIDI 2.0 that’s gonna make life easier:
Auto Configuration: midi devices can now have two way communication through one cable. I know, it took long enough. No more insanely large daisy chained midi networks and no more messing with a ton of drivers and controller profiles to get it to work. More room for that automatically working technology we love not setting up for 3 hours.
High Resolution Knobs: If you have ever worked with midi you know they only have 127 values from the lowest value to the highest. If you’re coming from analog, that’s just insanity. On the technical side, when building a plugin I have to do what’s called “value smoothing” so it doesn’t sound square and digital when you turn the knob. This introduces latency that makes it less musical. Higher resolution means smoother more responsive knobs, it also means if you’re building a midi network you have a lot more bandwidth to work with.
Audio Generation
This is the one everyone is terrified of, because it is totally terrifying and I love it. It is definitely for the hardcore DIY’ers. Throughout the year I’ve been keeping up with all the AI technology, from training an image generation model on the bunny from my album cover for Skunkworks (image) to training one that makes myself into anime (video), and lately training language models to write code and call functions. In the AI audio world, there are a few things that have come out that as a musician I think are incredibly useful.
Steinberg Backbone: Less on the DIY side but still the first to enter the game at this level, Backbone is a drum designer with a ton of features and a drum resampler that uses AI to generate samples using their model trained on 300,000 samples of kicks, snares and cymbals. The results are really incredible, and the interface steinberg built is amazing for drum design.
Harmonai: This is a research lab under the company Stability AI. For context Stability AI created the open source AI image generation model everyones been using called Stable
Diffusion. They have now created an audio model called Stable Audio and it is incredible, it has not been released as open source yet but Harmonai has announced plans to release open-source models based on the Stable Audio architecture. These models, along with the training code, will enable artists to fine tune their own model on their own music. I have a library of original music that I would love to run through one of these and sample the output, throw one of my plugins I created on it, go crazy. There are so many ways to assist in the creative process and help people through those creative blocks with this stuff it’s really exciting.
Sample Diffusion GUI: Dion Timmer, who is an incredible producer and dj, has created a GUI for Harmonai’s Sample Diffusion. This allows you to train your own sample diffusion models on your computer using your own samples, and generate new samples using a prompt, without having to code anything. I like getting weird with audio so this was something I wanted to start doing as soon as I heard about AI Audio. I plan to make a dataset of pots and pans and mix them with glitch noises. Hybrids, like you see people making tiger frogs with ai images. Will definitely be playing with this one on my next album.
In wrapping up, 2023 has been a landmark year for music tech. With the advances in MIDI 2.0 and AI in audio, we’re at the brink of a new era in music creation. There is a lot more on the technical side, like chatgpt can write audio processing algorithms and they work amazing, and frameworks like JUCE make it a lot easier to build audio plugins. I won’t burden you with even more tech talk but I think with all this we’ll see a lot more boutique plugin companies and hardware companies creating things we’ve never seen before, and it will enable artists to be creative. This year has been extremely productive for music technology moving forward and I’m excited to see what happens next. Expecting a new genre to drop any day now.
