Famous 3 voice chips
category: general [glöplog]
Hi
do you know about some famous 3 voices chips with some original examples of the classic tunes they played?
thanks
do you know about some famous 3 voices chips with some original examples of the classic tunes they played?
thanks
Texas Instruments SN76489 used in many game consoles, arcades and home computers, 3 square wave and one white noise generator.
Example: Leisure Suit Larry on PCjr
Example: Leisure Suit Larry on PCjr
SID, YM, POKEY! And that was it.
pokey is doesnt have a fixed amount of channels, you can combine 2 to make 16bit freq resolution instead of 8.
SID has only 3 voices? thats it? or maybe you could get more that just 3 kind of sounds?
more info please :) (& examples)
thanks
more info please :) (& examples)
thanks
LSL is a horrible example... Zeliard used all 4 voices simultaneously.
(4th channel hard to hear in that example but it is there, used for applying flange to a doubled voice)
ay-3-8910 ftw! \w/
YM2203 is a three channel FM chip used in various arcade machines and PC-8801 models.
Quote:
SID has only 3 voices? thats it? or maybe you could get more that just 3 kind of sounds?
Technically SID only has 3 voices.
However, there is a hack to play samples, which effectively adds a 4th voice.
Most music is just 3 voices though.
Quote:
Technically SID only has 3 voices. However, there is a hack to play samples, which effectively adds a 4th voice.
1) You can use 4 channels (or more) if you go the samples-only route and use software mixing. F.e. Polly Tracker uses 4 PCM voices.
2) Also you can make use of multi-speed tunes that practically also mitigate the limited number of parallel audio channels.
3) And finally: There are dual-SID solutions and some software to support that too.
(also 3 voices is quite a lot if you pack those channels properly; usually all the percussions and bass go into one channel, the next one has all the pads and arps, and the final one can be stacked with all the leads etc. If you combine that with fast speed, you easily get the impression of a relatively rich sound with a very limited number of concurrent audio channels; It also helps that all of this is mono and not streched out over a stereo spectrum...)
It also helps that the SID doesn't have fixed waveforms for each voice.
You can also get boards to have 2 SIDs or even 3 SIDs in the same c64. (like we did in Party Bit ) There are trackers around that support both.
I think I've seen exactly one tracker that supported 3 SIDs and it looked awful. :P In that scenario, I'd rather go with conventional SID-emulating VSTi instances (QuadraSID f.e.) on a PC host.
SID Wizard supports 3 SIDs now and it's very good. We used Goat Tracker Stereo & a Ninja tracker song playing together at the time because this didn't exist.
That's good. I was thinking of this one from the same author.
i think we need to clarify what is the difference between voices & channels
in the examples above i think there were more than 3 instruments no?(3 voices are 3 instruments that the song has & no more, no?)
in the examples above i think there were more than 3 instruments no?(3 voices are 3 instruments that the song has & no more, no?)
No, you can have as many instruments as you want but only as many of them playing simultaneously as there are channels. It's not like General Midi or something where you have a midi channel per instrument. (and then limited by the max amount of hardware sound channels) Most of the '8-bit' sound drivers will change instrument per frame on the same channel to give the illusion of more sounds playing at once. For example on the SID chip a modern bass drum sound might go:
So you hear a complicated drum sound that sounds like it has a hi-hat at the start and some depth to the audio, but in reality it's happened so fast it's near enough to the sounds being played simultaneously by the chip. Another example would be arpeggio chords, where the next note in an arpeggio is triggered every couple of frames to make one channel appear to be playing a chord of 3 notes or more.
C64 can put any waveform on any channel, other chips like the AY or SN76489 have a dedicated noise channel as well as 3 basic tone channels so you have to use that noise channel to play noise. You can get the AY to do a few other tricks (like buzzer) too.
Most 8-bit chips can be tricked into playing samples as well, sometimes this uses one of the existing tone channels to do a type of PCM modulation using an oscillator, other times it's done by rapidly changing the volume control register to simulate a 4-bit (usually) sample playback. As some of these players also do software mixing of multiple samples you can end up with more than the chip would usually play. (For example in the 'Vicious Sid' demos the SID chip is playing 4 virtual sample channels + 2 SID channels together)
Code:
frame 00 - noise waveform with high pitch (hi-hat)
frame 01 - pulse tone waveform sliding downwards (bass drum)
frame 02 - low triangle waveform sliding downwards (triangle gives it some more sub)
frame 03 - noise waveform (lower pitch end of the bass drum)
So you hear a complicated drum sound that sounds like it has a hi-hat at the start and some depth to the audio, but in reality it's happened so fast it's near enough to the sounds being played simultaneously by the chip. Another example would be arpeggio chords, where the next note in an arpeggio is triggered every couple of frames to make one channel appear to be playing a chord of 3 notes or more.
C64 can put any waveform on any channel, other chips like the AY or SN76489 have a dedicated noise channel as well as 3 basic tone channels so you have to use that noise channel to play noise. You can get the AY to do a few other tricks (like buzzer) too.
Most 8-bit chips can be tricked into playing samples as well, sometimes this uses one of the existing tone channels to do a type of PCM modulation using an oscillator, other times it's done by rapidly changing the volume control register to simulate a 4-bit (usually) sample playback. As some of these players also do software mixing of multiple samples you can end up with more than the chip would usually play. (For example in the 'Vicious Sid' demos the SID chip is playing 4 virtual sample channels + 2 SID channels together)
If you really want to get into explaining the most basic thing for someone who knows absolutely nothing about anything, then you need to explain why you need three channels, isn't stereo enough. And the C64 only has one output channel, how can it have three? Does it have a hidden surround output somewhere or is the third one for the subwoofer?
I also bet that the OP wouldn't know what "frames" are - assuming she is not a troll.
I also bet that the OP wouldn't know what "frames" are - assuming she is not a troll.
now it is more clear thanks
so was there an "only 3 instruments" chip in the past?
thanks
so was there an "only 3 instruments" chip in the past?
thanks
I guess the NES chip where you have a fixed triangle and noise channel is something like that, but you can alter the other two tone channels to have different waveforms, though only with 4 different settings.
However early home computer games from the 1982-1984 era usually set the sound chip's channels up to only play one instrument for the duration of a song. (eg: channel 1 : bass, channel 2 : an organ, channel 3 : a flute) If the hardware only allowed pulse tones than they were made to sound different by changing the ADSR (or note duration) on each channel. (eg: 1 short bass, 2 : long melody tone, 3 : shorter accompany tone)
Music on those chips moved on quite rapidly though to try and make more interesting sounds (by updating the chip parameters as the notes played) because it gives the music much more character when you're dealing with simple waveforms. You might find this seminar video by LFT interesting, it goes through the progression of chip music through the years. Again, while the hardware is fixed the progresison came from the software evolving to allow new tricks.
However early home computer games from the 1982-1984 era usually set the sound chip's channels up to only play one instrument for the duration of a song. (eg: channel 1 : bass, channel 2 : an organ, channel 3 : a flute) If the hardware only allowed pulse tones than they were made to sound different by changing the ADSR (or note duration) on each channel. (eg: 1 short bass, 2 : long melody tone, 3 : shorter accompany tone)
Music on those chips moved on quite rapidly though to try and make more interesting sounds (by updating the chip parameters as the notes played) because it gives the music much more character when you're dealing with simple waveforms. You might find this seminar video by LFT interesting, it goes through the progression of chip music through the years. Again, while the hardware is fixed the progresison came from the software evolving to allow new tricks.
Nice seminar video, thanks. I hadn't seen that one. The point about LFO rate ("effect rate") vs. audio rate ("pitch rate") modulations is a good one. In my opinion, if you can use an audio rate interrupt on PSG/AY/YM/SN like they do on the Atari ST, you can do a pulse width modulation effect, which isn't possible with plain vblank interrupts, so it's a different "platform" in that sense. Still I wouldn't say that such Atari ST tunes aren't chip music. But when you do PCM samples at something like 4+ kHz modulation, how about that? And then again, if PCM isn't allowed, then nothing coming from Amiga's Paula would be "chip music", even including AHX?! If I remember correctly, at Assembly '94 there was a chiptune compo where the limit chiptune vs. non-chiptune was defined as the size of the MOD file. How about the Sega Outrun arcade machine, with FM + PCM? Or if you have a 6-operator FM chip, it could do very thick sounds like the Yamaha DX7, is it not chip music if the timbres are too nice?
messy post... the asm'94 chiptune limit was 32 kB