Ay -3-8910 psg trick

Page 1/3
| 2 | 3

By PingPong

Prophet (4093)

PingPong's picture

12-01-2017, 00:09

By ARTRAG

Enlighted (6923)

ARTRAG's picture

12-01-2017, 08:01

Maybe on msx 2 the rtc in test mode is a better option than hblank because iirc this latter does not trigger while borders are plotted

By PingPong

Prophet (4093)

PingPong's picture

13-01-2017, 19:54

umh, how can you trigger the ints with RTC?

By Louthrax

Prophet (2436)

Louthrax's picture

13-01-2017, 20:25

ARTRAG wrote:

Maybe on msx 2 the rtc in test mode is a better option than hblank because iirc this latter does not trigger while borders are plotted

Wouldn't that change you current time/date settings ?

By ARTRAG

Enlighted (6923)

ARTRAG's picture

13-01-2017, 23:06

@Pingpong, no, you can only do a loop that waits a given rtc increment to cycle
@Louthrax, yes this would change the time and date

By Louthrax

Prophet (2436)

Louthrax's picture

13-01-2017, 23:14

ARTRAG wrote:

@Louthrax, yes this would change the time and date

Argl, too bad! So we have a high-res timer on all MSX2 machines but it can't be used...
Was thinking about saving the RTC time before switching to test mode, then adjusting the time back after use by adding the delta divided by some constant (you just need not to reset the MSX in between!). That could also add a small delta in the current time...

By Grauw

Ascended (10699)

Grauw's picture

14-01-2017, 03:55

ARTRAG wrote:

Maybe on msx 2 the rtc in test mode is a better option than hblank because iirc this latter does not trigger while borders are plotted

It does, I use it in VGMPlay to get a 300Hz interrupt. However, the 1chipMSX is buggy.

(Note at 60 Hz there are 262 lines and at 50 Hz there are 313, and you can only specify line interrupts for 256 lines, so there is still a little gap.)

PWM is also done with the AY3 (YM2149 actually) on Atari ST. I think it is possible on MSX2, but tricky.

By Manuel

Ascended (19270)

Manuel's picture

14-01-2017, 10:52

Wow, that sounds pretty awesome! Is it only possible with the power of the ST's CPU?

By ARTRAG

Enlighted (6923)

ARTRAG's picture

14-01-2017, 11:39

and its timers... this makes the difference
http://dev-docs.atariforge.org/files/MC68901.pdf

By hit9918

Prophet (2927)

hit9918's picture

14-01-2017, 11:27

the sega demo does wobble when sliding the frequency, it sounds like something has lack of granularity and the notes would be out of tune.

the 9938 rate has potential. look here http://insidious6581.blogspot.de/
scroll to "The audio updates are done with the CIA timer at 200 times per second", check out the drum.

this rate below khz is about adsr vibrato arpeggio kind of things taken to increased rate and then new things can happen.
the nicest would be to run a whole player at this rate. players tend to spike at notes, and the note rate would not be increased, maybe the cpu load isnt that bad.
and maybe 150hz without opened borders is already touching this interesting zone.

By Grauw

Ascended (10699)

Grauw's picture

14-01-2017, 12:39

Grauw wrote:
ARTRAG wrote:

Maybe on msx 2 the rtc in test mode is a better option than hblank because iirc this latter does not trigger while borders are plotted

It does, I use it in VGMPlay to get a 300Hz interrupt. However, the 1chipMSX is buggy.

(Note at 60 Hz there are 262 lines and at 50 Hz there are 313, and you can only specify line interrupts for 256 lines, so there is still a little gap.)

I want to add, instead of using line interrupts, you can also poll the HR flag. Nice 15kHz oscillation.

Page 1/3
| 2 | 3