I remember ZX-81, which had no sound. In 1983, that problem was overcome by flickering colors - it affected the TV's electronics instead :) Or rather, all TV sets had electronics to split audio from video on the RF band, and such electronics could never be made perfect. So it might be that an imperfection in the C64 RF causes video noise to leak to audio, and by making the right noise...
@inphanta it actually is. You have to tune up the audio of the TV quite a bit. Inside the C64, not much is shielded against other influences except of the VIC itself. The loopback comes from the TV which radiates itself. (dont go too close with your Tape recorder :D)
Oops, wait... those are solid horizontal lines, so the frequency of the samples will actually be the PAL line rate, 15625 Hz. Then the resulting luma output is bleeding over onto one or both of the video outputs, resulting in the sound you hear. (That's also why it's so quiet, since usually the goal is to *minimize* that bleed!)
Your comment is right. The routine is basically using the video output generated by the VIC as an extremely rough PWM channel. Of course, you're probably talking 1-bit samples (maybe 2-bit since there's some gray in there) generated at whatever the pixel clock's rate is (not sure what that would be on a PAL system), plus there's going to be distortion during the blanking intervals. What's amazing is that this is incredibly similar to the way sound was FIRST generated on an IBM 704!
the sound is not produced by the flickering bars, its the other way round: the digitized sound causes these flickering bars. of course there's no Sid-chip used because its digital and the sid only produces anaolg waveforms
I remember ZX-81, which had no sound. In 1983, that problem was overcome by flickering colors - it affected the TV's electronics instead :) Or rather, all TV sets had electronics to split audio from video on the RF band, and such electronics could never be made perfect. So it might be that an imperfection in the C64 RF causes video noise to leak to audio, and by making the right noise...
powderyorgan 2 months ago
What the fuck is this?
sacrottt 4 months ago
@sacrottt do you honestly care? :)
rvounik 3 months ago
@rvounik No. Fuck this.
sacrottt 3 months ago
Isnt that actually a proof AGAINST the sound quality of the C64 that such distortions are actually possible? :D
AmstradExin 10 months ago
@AmstradExin: No, since this isn't using the SID. It's using the graphics chip to play audio. That's the whole point of the demo.
inphanta 8 months ago
@inphanta it actually is. You have to tune up the audio of the TV quite a bit. Inside the C64, not much is shielded against other influences except of the VIC itself. The loopback comes from the TV which radiates itself. (dont go too close with your Tape recorder :D)
AmstradExin 8 months ago
Blinding.How on earth do you program somthing like this?
Amazing.
SPFXsynthesisers 1 year ago
Oops, wait... those are solid horizontal lines, so the frequency of the samples will actually be the PAL line rate, 15625 Hz. Then the resulting luma output is bleeding over onto one or both of the video outputs, resulting in the sound you hear. (That's also why it's so quiet, since usually the goal is to *minimize* that bleed!)
gregly 1 year ago
Your comment is right. The routine is basically using the video output generated by the VIC as an extremely rough PWM channel. Of course, you're probably talking 1-bit samples (maybe 2-bit since there's some gray in there) generated at whatever the pixel clock's rate is (not sure what that would be on a PAL system), plus there's going to be distortion during the blanking intervals. What's amazing is that this is incredibly similar to the way sound was FIRST generated on an IBM 704!
gregly 1 year ago
This demo takes the piss (in a good way). A technical triumph!
inphanta 1 year ago
awesome!
MiracleKD18 2 years ago
holy crap! Amazing...
TimTrimT 2 years ago
the sound is not produced by the flickering bars, its the other way round: the digitized sound causes these flickering bars. of course there's no Sid-chip used because its digital and the sid only produces anaolg waveforms
samplesmasher 2 years ago
@samplesmasher no
mirabilis 1 year ago
Amazing!
BaconBricks 2 years ago
W glowie sie nie miesci co na tym sprzecie ludzie wyczyniaja =) Alleluja :)
Zgrywus666 2 years ago