News:

11 March 2016 - Forum Rules

Main Menu

How Do You Compose Chiptunes?

Started by Quantam, June 13, 2022, 04:04:09 AM

Previous topic - Next topic

Quantam

Question pretty much says it all.

FAST6191

I have only ever really converted songs between systems with a light bit of simplification but I get the impression from many of those I talked to that do that such things are not so very far from asking what text editor a programmer uses, or I guess hex editor that a hacker uses, which is to say while there are bad choices it is so very far from the whole story.

You could probably draw a line from the would be composer's background -- musicians probably going more for midi (if I can play a keyboard and have it appear then great), those that would have came up via trackers then that, those that compose more generally (usually meaning an affinity for sheet music/musical notation) and move sideways then reflecting that and any kind of language probably being those that push it to the limit (I don't write assembly by default but when I do...) or the super rare programmer turned musician.

Said limits, much like a lot of good art and challenges within, are also what matters more in said conversations it seems -- while the DS I spend much of my time with has 16 full wave capable independent channels then older stuff... being able to play two chords at once (or overlay sound effects and music) was a notable transition in power for some of those or took some seriously creative steps (basic but I really like it https://www.youtube.com/watch?v=roBkg-iPrbw ).

PolishedTurd

It's very important to me to be able to perform the sounds I am composing with, meaning I can play them in real time on a keyboard. Punching in numbers and then listening to the playback is much less satisfying. Nearly all of the chiptunes I hear are flat and unexciting in terms of dynamics and modulation, which I attribute in large part to the decoupling between physical performance and production of musical "data". To really flex a track and make a chip sing, putting subtle modulations on it or performing "envelopes" on the modulation, is probably too numerically complex for most people to bother calculating and inputting manually, whereas it's a flick of the wrist when performing.

I owned a Midi-NES cartridge for years and still have a GBA nanoloop cartridge somewhere. Wonderful as it is to hear the real hardware, both were too compositionally cumbersome.

These days, I use the Plogue emulations and absolutely love them. The Genesis/Mega Drive and SNES plugins are glorious, and I recently made a smartphone app where the UI sounds were 100% NES (2A03) from a Plogue plugin.

That being said, it is also of great interest to me and others to be able to bring new music to old NES games. It would be tremendous if MIDI could be converted to MML, and then if the MML could be compiled into assembly for these specific music engines that you expertly deconstruct. Then someone could compose a song and import it into a game without learning the engine inside out or getting covered in bytes trying to translate an instrumental track into assembly commands.

Sliver X

I composed them on a guitar, then translated that into the music data format for whatever game it was I reversed engineered; I would then manually insert this back into the ROM via a hex editor.

(I realize this is insane)