https://youtu.be/ahts3efFzFI
watch this video and figure out how iterate through the circle of fifths for clean, procedural audio.

There are a plethora of chord progressions and scales. figure out three or four for the user to pick from. The more different moods/feeling to them, the better.

figure out how to change the pitch of an audio file without changing the speed at which it is played.

either get pitched samples (what they use in music productions, already come sin a variety of pitches) or get a handful of different pitches on a roughly equal distance, that can be pitched up and down without too much distortion.

Play a specific audio file, when a button is pressed.

we can suggest these to the user before meditation. allow the user to add custom ones.

ask the user what they did while listening. exercise? read a book? art? sat with eyes closed?

figure out where to save all the data about the user. also figure out how to store it (data schema and the like)

Figure out how to parse all the data. we have how a user felt, what they did, how they felt afterwards. need to figure out, based on a user feeling that they should be recommended .

design interface for asking the user how they feel.

instead of asking the user "do you feel angry, depressed, bad, good", ask them instead more specific things, like "I feel lethargic and unmotivated." limit to 5 max, maybe?

slow fade between pastel colors

get together at least 5 different sound files for each sound type:

fire crackle
rain
thunder booming (hopefully without rain)
wind
ocean
seagulls
brook/river

make sure each file is at least a minute. it can be pulled from a video, as long as the video isn't looped.

randomize length, and starting point, for a set of files, to have an 'infinite' sound of that kind.