Author
| does the PSG have also different "sounds" / instruments ?
|
norakomi msx professional Posts: 870 | Posted: May 10 2006, 18:23   |
Hey hey hey !!
when listening to different kinda games, I seem to notice that the PSG sound is not always the same.
It almost seems as if the PSG has different instruments.
Question:
If I play an "A" at 440 hertz using the PSG register (1 channel, no noise used), will this ALWAYS sound the same, or can I indeed assign a certain characteristic to its sound ??
|
|
ARTRAG msx guru Posts: 2222 | Posted: May 10 2006, 18:29   |
Using the sole PSG HW, the only "enhancement" you can get is the modulation of the amplitude
resorting to one of the standard waveforms provided for the amplitude envelop.
no instruments, all you can get are square waves at different frequency and variable amplitude
|
|
Edwin msx professional Posts: 716 | Posted: May 10 2006, 19:19   |
You can also code your own frequency and volume envelopes.
|
|
PingPong msx master Posts: 1286 | Posted: May 10 2006, 19:21   |
By using evelope HW you can go near the sound of some instruments, like bells.
|
|
wolf_ online
 msx legend Posts: 5178 | Posted: May 10 2006, 19:23   |
Norakomi, mess around a bit in basic with the M and S commands in a play string.. that'll teach ya!  |
|
GhostwriterP msx addict Posts: 339 | Posted: May 10 2006, 19:50   |
Ever listened to Seleniak? That sounds like a heavy dirstorted square wave
|
|
pitpan msx master Posts: 1567 | Posted: May 10 2006, 20:16   |
You can basically play with three parameters: tune frequency, amplitude (volume) and noise frequency. For the amplitude, you can also use the hardware envelopes, but you should be very careful not to desynchronize the song.
My recommendation: get a PSG tracker and load a bunch of good PT3 songs to discover what can be done with it.
|
|
dhau msx master Posts: 1384 | Posted: May 10 2006, 20:34   |
It was discussed to death in russian ZX Spectrum forums. Basically GI AY chips and Yamaha YM chips are slightly different, where AI uses 4 bits for duration and YM uses 5. This make sounds a bit shorter and so same melody sounds a bit harsher on YM. Mr Bulba Wrote software AY/YM emulator that accounts for all those differences.
AY can be found in most MSX1 systems. YM and integrated YM is in all MSX2 and higher. Notable exceptions are Arabic MSX1 systems and Brazilian Expert+ with Toshiba MSX Engine chip. I think it sounds more like YM hten AY.
|
|
boblet msx lover Posts: 69 | Posted: May 11 2006, 11:24   |
I have one hx-10 with the ay chip, and one hx-10 with the ym, guess they just used whatever the felt like at the time.
|
|
norakomi msx professional Posts: 870 | Posted: May 11 2006, 13:23   |
Quote:
| mess around a bit in basic with the M and S commands in a play string.. that'll teach ya!
|
I never really played with the play statement.
Could you give an example?? Quote:
| By using evelope HW you can go near the sound of some instruments, like bells.
|
I dont understand how changing the volume can alter the "instrument sound".
If you could change the volume of certain frequenties, then this would be like an equalizer.
Lets say you EQ more BASS into a PSG instrument.
But I guess thats not possible right?
For instance, when I use SEE (sound effects editor, Fuzzy Logic) then I can use the volume channel to change the volume of the PSG, but I see no clear way to change the "instrument sound".
Is this possible with SEE ??? |
|
wolf_ online
 msx legend Posts: 5178 | Posted: May 11 2006, 15:32   |
play "m20s10l4c"
Quote:
| I dont understand how changing the volume can alter the "instrument sound".
|
Imagine the psg being a square wave of 440hz
-- -- -- --
-- -- -- --
If you alter the volume of this waveform at a rate below 20hz then you get amplitude-modulation, sometimes called 'tremolo' in synths. If you alter the volume at a rate above 20hz then you'll get yourself in the tonal domain, the domain where we perceive these changes as being 'changes to the sound', rather than 'changes to the amplitude' (the latter would require a rate <20hz).
So, if one of the waveforms of the PSG (didn't check) is a sawtooth, then, when applying this sawtooth to our waveform above at 440hz, would get you something like this:
\ \ \ \
/ / / /
whereas a slower ampmod rate would result in something like this: (zoomed-in view ^_^)
--
--
--
--
--
--
--
--
So the frequency with which you modulate can be quite important. Actually, for stable tones you need to sync these modulation-values to the note somehow or you'll get nasty crap. As there's, at a certain point, a lack of detail in what you get from these modulationvalues, notes at some pitches can sound slightly different, as if they'd be detuned orso.
SEE?
Yes! That's how, for example, I did the laser sfx for UU. One of the most right columns should be set to VWave orso.. some keycombi orso.. shift-w? |
|
mars2000you msx master Posts: 1723 | Posted: May 11 2006, 15:35   |
Quote:
| whereas a slower ampmod rate would result in something like this: (zoomed-in view ^_^)
--
--
--
--
--
--
--
--
|
It looks like Nemesis/Salamander !  |
|
AuroraMSX online
 msx master Posts: 1383 | Posted: May 11 2006, 17:50   |
Quote:
| Imagine the psg being a square wave of 440hz
-- -- -- --
-- -- -- --
If you alter the volume of this waveform at a rate below 20hz then you get amplitude-modulation, sometimes called 'tremolo' in synths. If you alter the volume at a rate above 20hz then you'll get yourself in the tonal domain, the domain where we perceive these changes as being 'changes to the sound', rather than 'changes to the amplitude' (the latter would require a rate <20hz).
So, if one of the waveforms of the PSG (didn't check) is a sawtooth, then, when applying this sawtooth to our waveform above at 440hz, would get you something like this:
\ \ \ \
/ / / /
|
It's probably me, but I don't get this one. I would have expected something like
\ \ \ \
\__ \__ \__ \__
(provided that PSG waveform and sawtooth are perfectly in sync)
|
|
wolf_ online
 msx legend Posts: 5178 | Posted: May 11 2006, 17:57   |
Well, there my psg knowledge ends..
Your diagram would be true if the psg square isn't a signed one, e.g. it's output alternates y and 0, rather than y and -y.
Anyway, it was just to point out that amplitude-modulation can be used to transform a waveform, no matter how many errors there are in the story.  |
|
ARTRAG msx guru Posts: 2222 | Posted: May 11 2006, 20:37   |
Aurora is rigth : the psg has only positive output!
|
|
|
|
|