herr_prof wrote:

If you want to skip between two different wavetables, just make a wave with two wavetables back to back


That's a good idea. But ya know, that requires planning and editing. I want just random offset access... uh...wait...does lpof just loop around if you give it a big number? I don't even have my PS P charged right now. Oh well, something to try.

The docs say

LPOF 0001 adds one to both values, LFOF FFFF removes one (so values > 0x800 moves the loop backward)

But when I try going backwards, I get weird, ducked up sounds. All clicky.

Last edited by breakphase (Jan 31, 2017 8:38 pm)

That makes sense because lpof will move your loop points and probably create zero crossing problems, ie those clicks.

i never use plof for osc instruments, I feel like it's more suited for drum loops, or stuff with really short transients like speech. lpof works really nicely with samples like monowave because that is 256 samples long, so you could have a loop at x800-x900 and then plof 0001 in a table to slowly scan forward a waveform at at a time, and not have to worry about clicks

Last edited by herr_prof (Feb 14, 2017 5:03 pm)

herr_prof wrote:

That makes sense because lpof will move your loop points and probably create zero crossing problems, ie those clicks.

Yeah but in theory, if  i start on wave boundaries, then I go forward 2048 smps, I will be at the start/end of the next wave; then I go backward 2048, I should still be landing on the wave boundaries.  I should be right back where I started. So I propose that either LGPT is fucked up, or I'm not understanding the lpof command.

herr_prof wrote:

i never use lpof for osc instruments, I feel like it's more suited for drum loops, or stuff with really short transients like speech. Plof works really nicely with samples like monowave because that is 256 samples long, so you could have a loop at x800-x900 and then plof 0001 in a table to slowly scan forward a waveform at at a time, and not have to worry about clicks

Yeah it's good for many good effects. but with it would just be good to get random access in oscillating mode.

Last edited by breakphase (Jan 31, 2017 10:00 pm)

Piggy has no protection from changing a sample ON The zero crossing. Its a quirk, not a bug big_smile

http://kyma.symbolicsound.com/qa/1710/s … wavetables

herr_prof wrote:

Piggy has no protection from changing a sample ON The zero crossing. Its a quirk, not a bug big_smile

http://kyma.symbolicsound.com/qa/1710/s … wavetables

Sorry, I feel like we're not saying the same thing. I'm saying, I put the loop start at 0 (a zero-crossing) and loop end to 0x800 (the end of the wave,thus zero crossing), then move forward by 0x800 like this:
00 LPOF 0800

So now loop start is at ^^ 0x800, and loop end is at 0x1000 (a zero crossing)

Then I do

LPOF F800

Theoretically, I should be back where I started right? Still at zero crossings, because I move only by intervals of the cycle period. But it clicks... I don't get it.

are you moving the waveform at a point in playback of the actual audio when it is playing at that zero crossing?Thats the only way to ensure it won't click.

Imagine the Loop point is quick and brown in this sentence:
The | quick brown | fox

If I stick a plof AT the zero crossing to move it to:
The quick | brown | fox

It shouldn't click! But if I do a plof WHILE THE WAVEFORM IS GREATER THAN ZERO it will probably click

The | quick brown | fox
                 | LPOF F800

Make sense? I dunno, maybe Im wrong!

Last edited by herr_prof (Feb 1, 2017 9:22 pm)

Oh no, I get what your saying. I'm okay with that discontinuity. That's fine, sounds like the WAVE channel in LSDJ kind of. I mean the waveform is kucfed up, and has a 'buzz' in it. I guess click is the wrong word. Periodic discontinuity. tongue and only when I go back, not forward.

Last edited by breakphase (Feb 2, 2017 4:42 pm)

Maybe thats CPU spikes on your PSP from running backwards

You are a silly goose

Im serious! record an example!

Actually that's a good idea.

Well, Im having the same issue when going back with the sample. But my knowledge is ancient and I read somewhere about this issue.
But it was supposed to work? Never got backward plof to work with big samples. I could do some pwm with shorter samples and it would go back and forth, but with these wavetables it doesn't work. It just goes to the end of the sample and don't go back.

breakphase wrote:

I'm saying, I put the loop start at 0 (a zero-crossing) and loop end to 0x800 (the end of the wave,thus zero crossing), then move forward by 0x800 like this:
00 LPOF 0800

So now loop start is at ^^ 0x800, and loop end is at 0x1000 (a zero crossing)

Then I do

LPOF F800

Theoretically, I should be back where I started right? Still at zero crossings, because I move only by intervals of the cycle period. But it clicks... I don't get it.

This is absolutely how it's supposed to work. I just tested with monowave and it works perfectly in both tables and patterns. Build 1.3n_052 in win32.

I tried it and it worked, dunno wtf I was doing wrong. No clics either