This repo is (initially) a few loosely organised thoughts for a macropad that I've been thinking about for some time and which (in lieu of having built it!) is replaced with keyboard macros and slightly inconveniently placed shortcut buttons.

Other approaches I have used and which are "good, but could be better" - a cheap USB HID device as a transcription start/stop button (mapping: F13). This can also be a surrogate PTT button. I also have a side button on my vertical mouse set up for the same purpose. And a couple of keys assigned on my keyboard. I also tried a foot mouse and ... wasn't a fan.

While such hardware was of very peripheral interest before (get it?!) with the advent of Whisper, STT has dislodged text as my default mode of entering text into a computer. And as I do that quite a bit, I find myself looking for something more

Why a voice pad?

If you use STT to do a substantial amount of typing then you know why voice related peripherals exist.

But the question of what functionalities are probably still somewhat workflow specific: I need stop and start buttons but if you're a professional transcriptionist ... you probably are well-served by existing hardware.

So this voice pad reflects my needs. Which are...

---

Bindings / Shortcuts Plans

Start And Stop

Press a button to start STT and press a button to end the record stream / input.

This can be one button, two or three.

• 1 button is start/stop
• 2 buttons are start and stop
• 3 buttons is start, pause stop - but most STT apps dont support pause.

Current preference: stick with 1.

Voice Note

Besides STT

Voice Recording

Design assumption/idea: separate buttons for real time and async recording

Design Notes

Context constriction is key: if the recording keys bind to F13 and that in turn binds to software level functions we will end up with overlapping text inputs