Pulsar: with audio

[neilbaldwin]

Quick video of my new NES tracker showing the kind of sounds that are possible when you run it at 240Hz

http://blog.ntrq.net/?p=338

[Bregalad]

Cool, I always through it could be interesting to run faster than 50/60 Hz. Many SNES games do stuff like that in their sound effects - it's the "standard" of the console because a 1,02 Mhz CPU is dedicated entirely to sound. This isn't the case on the NES.

[Dwedit]

I can't believe that a triangle wave was producing that!

[Jarhmander]

I can't believe that a triangle wave was producing that!

Frequency modulation gives always a surprising result, as the output spectrum is quite different of the carrier spectrum

[neilbaldwin]

Another (big) video update. Making good progress.

http://blog.ntrq.net/?p=350

[Gil-Galad]

I haven't looked into it much, but what's the difference between Pulsar and NTRQ? All I know right now is that both of them seem to be trackers.

Unfortunately, at the moment NTRQ kind of confuses me, especially since I don't know how to use a tracker yet.

[neilbaldwin]

They are both trackers but each with different approaches.

Fundamentally, what is driving Pulsar's design (heavily borrowed from LSDJ) is that I'm running the audio engine at (technically) 240hz i.e. 4 times per frame.

This provides the opportunity to explore some interesting sound modulation, not normally accessible running at 60hz. Watch the video and look out for me using the Nxx command on a continuous note, for example.

[Gil-Galad]

That's pretty awesome in the video and what you're doing. I may try to learn how to use the tracker later on. But for now, I'm just more comfortable using a hex editor and assembly text files.

I used some commercial music drivers to make some basic scales played simutaneously on each of the channels, but nothing more complex than that. But of course I really can't use them for games or music competitions. (I doubt it anyways). If you don't mind, I have some questions about Nijuu to ask you. I won't here because that's off topic. I'd like to send you a private message if you don't mind.

[neilbaldwin]

Ask away, no problem.

[thefox]

Out of curiosity, how does Pulsar time the 120Hz / 240Hz updates? Do you use a scanline IRQ or some sprite0 hit/overflow trickery? Or DPCM IRQ?

[Dwedit]

If it's DPCM irq, you're not going to be using any DMC samples. So I highly doubt it's that.

[blargg]

Maybe plain cycle timing? If you do it all from an NMI handler, you'd cycle time the first three 1/240 periods, then return from the handler after the fourth update for that frame, leaving about 1/240 second free to the non-NMI code.

[neilbaldwin]

Out of curiosity, how does Pulsar time the 120Hz / 240Hz updates? Do you use a scanline IRQ or some sprite0 hit/overflow trickery? Or DPCM IRQ?

Sorry, only just seen this question.

I'm just using timed code, as blargg said. In the NMI I just do;

refresh audio
delay
refresh audio
delay
refresh audio
delay
refresh audio
refresh sequencer

and space the delays out so that the audio refreshes are spaced evenly (as possible).

Then I have a flag (4 bits) that gets checked at each refresh to tell the engine whether to process that refresh or not. So by varying the value in the flag (which I can do from within the tracker itself as a command) I can make the audio engine run at 60/120/180/240Hz (ish).

[neilbaldwin]

Oh, I update the website with a new video recently too, mainly to demonstrate the flexibility of random jumps in Tables and also a new idea - multi-channel tables (which I've labelled 'MultiFX'.

http://blog.ntrq.net/?p=357

[neilbaldwin]

Pulsar V1.0 released today

http://blog.ntrq.net/?p=420