>and you were getting music out of the internal speaker usually known for just going "beep", then you've got something impressive.
Nah, just need to dedicate CPU cycles to the task.
we had digitized speech (SeaDragon) on the Apple ][ and music (Goonies), even classical. (Not linking to the obvious Castle Wolfenstein)
At 1 MHz you can get fake 2-voice music. (Karateka)
1-bit sampling can produce anything (albeit at low quality) when you dedicate even 1 MHz to the process.
Similarly on the PC @ 4.77 Mhz, you can get fake mutl-voice music (8088 MPH by Hornet + CRTC + DESiRE)
On the PC, since you had ~8 MHz you could get fake 3 voice music. (Sorry, can't find a demo)
8-bit DAC at 16MHz sounds SO much better.
So yeah, this is old news.
Still impressive for a calc. :-)