Ah man, haven't written any pico8 for a while so I'm trying to get my groove back. Let's warm up with another console's logo intro: Sega Megadrive's! (or Genesis', depending on your region)
Not really, I'm kind of a novice at sound myself. I just bodged it together a bit at a time, trying to figure out what sounded right, knowing what little I do of linguistics. :)
(I've looked into proper speech synthesis, and boy, that's complicated.)
Gruber and a couple of other pico-masters are (were?) trying to get text to speech working: https://twitter.com/gruber_music/status/956308869516414976
I'm really not sure PICO-8's pseudo-hardware has enough resolution over time or features to do decent speech synthesis.
Like, there's only 8 levels of volume and only 120 note changes per second. Also, no filters (e.g. lo-pass, hi-pass) at all, so no way to make pink or brown noise, only white, and that makes it hard to make sounds like "F". Edit: actually, maybe lower "pitch" white noise counts as pink or brown. I'm not an audio guru so I may not be understanding this properly. Pretty sure it doesn't, though.
You might be able to bodge something together that approximates speech the way I did above, but it would be very coarse and hard to understand.
Sampled sound is probably right out, since you couldn't approximate any waveform with a frequency higher than 60Hz.
Huh, funny coincidence.
StinkerB06 just added some new discoveries to the PICO-8 memory map:
0x5f40: Bitmask to half each audio channel's clock rate. Bit 0 (value 1) is for Channel 1, Bit 1 (value 2) is for Channel 2, and so on. The upper four bits are not used.
0x5f41: Bitmask to enable reverb for each channel. The bits have the same effect as 0x5f40.
0x5f42: Bitmask to distort the waveforms of each channel.
0x5f43: Bitmask to enable a lowpass filter for each channel.
(Nice find, @StinkerB06.👍)
Not sure if we can use these safely, but zep does sometimes say there's more hardware state yet for us to discover, so maybe it's legal...?
[Please log in to post a comment]