Have a rudimentary song generator at familiar-synths.js.
The song still isn't quite to my liking as all tunes it generates still sound "same-y".
Gibber (alpha playground) is still having problems with some of the presets, so the preset/instrument library that can be drawn on is pretty limited.
- figure out how to make some reasonable generative synths instead of relying on presets
- remove drums at end so they stop
- find instrument sound levels with each other so they play well with each other
- add extra effects (pads, background texture, more complex beats/rythms) to score to help flesh it out
- figure out why the songs sound so stale
| priority | low |
Boards is generated but unordered and the test is unassembled.
- test out with a nano/arduino with a CR2032 cell to see you can vibrate the motor
- order boards
- create software for it
| priority | medium |
Jacket is assembled but the backing board needs to be paired down.
If there's some way to make the backing easier, this would be ideal.
- Put EVA foam used for floors as a backing plane for the LEDs
- Put EVA foam in both front corners as backing
- left is for electronics, suitably shoved in there to try and shield it from use and offer some strain relief
- right is for battery
- might need to elongate the wire from battery to electronics
- Need to fit to see if the positioning, EVA foam pruning and corner backing is reasonable
| priority | high |
Boards are ordered but need to be soldered and tested
- Tested boards
- Big hangup was the 'sense' that needed to either be tied to ground or go through a 1ohm resistor
- Let the smoke out of one by connecting it to a large BLDC motor
- Ordered more motors
- Test with encoder
- Create two motor system with arms to create something interesting
- rev version (?)
- on board atmega
- on board encoder (integrated)
| priority | high |
Have a set of 10 boards (put on Tindie) but without a good method for the diffusors.
The diffusors are laser cut from EVA foam but maybe they would be better served by 3d printing them. Seems like kind of a waste of material but who knows.
- create a clock out of them (or something) as a standalone project
| priority | medium |
The 3d printing solution I was pursuing is not going to work.
I think a better method is to have cloth/Tyvek hinges sandwiched in between laser cut/3d printed/whatever thin strips which then go to a central shaft.
Ordered some tiny stepper motors. Hopefully they can be driven direct from an Arduino or, in the worst case, driven by some small transistors. With luck, I can connect the shaft directly to the platform.
- get tiny stepper to run (direct off an Arduino)
- construct test rig as proof of concept
| priority | high |
Still don't know how to approach this.
Some key points:
- Each layer does not have lights in between them, it's completely backlit by the back most panel, diffused from LEDs
- The spacing can be done with foam
- Put it in a frame
- figure out what the scene should be an implement it
| priority | low |
Create a full node with lightning support.
None.
- Either buy a pi + hdd or rent one that has enough space
- stand up node and play around with it
| priority | high |
Software is mostly created, though it needs to be tested a bit.
- run some outdoor tests on the solar + battery to discover feasibility
| priority | medium |
Highly speculative, need to research more.
A modification to sox to allow for logarithm frequency spectrogram here (gh).
- find a workflow to create spectrogram, output some type of constellation map
- go back from constellation map to sound to see what gets picked up
| priority | low |
Initial investigation is good. DeepSpeech is faster than real-time.
- test with usb microphone
- create interface (maybe with visual feedback)
- add some external trigger mechanisms