Ah, OK, then yes, a source of internal modulation of octave would be possible. BTW, I also tried to implement the Tides scheme of integer ratios of the input frequency, rather than just octaves, but couldn’t get it to work reliably. I may revisit that at some stage.
Someone on MW asked for a video demonstrating the Auto-didact Peaks-playing-itself self-patch used for the SoundCloud track above. So here is a quick video showing just that:
beautifully done bannelong! can’t wait to test this out - we’ll use those samples another time
A little patch using the PLO for FM as described above: https://www.youtube.com/watch?v=Ycvo5g7lcY8
@BennelongBicyclist Not sure if you don’t want people posting youtube clips featuring DMC prior to the 1.0 release so I haven’t tagged it with any information.
@freelivela: I will use your samples, maybe the weekend after next. When you have the opportunity, a nice pair of open and closed hihat samples would be great. The tail of the open sample can’t be too long, though.
@Flohr: nice! The PLL slew and occasional quick glitch gives it a nice “humanoid” feel, I think - it isn’t perfect, but isn’t meant be. I’d fiddle some more, as time permits, to try to get PLO working more reliably, but I think it is worth persisting with it. Thanks for the idea!
Posting clips etc is fine - DMC isn’t a commercial product that I am trying to keep under wraps until its release. I’ll post a link to your clip on in the relevant thread on MW, if that’s OK.
Glad to hear it! Sharing on MW is fine with me.
Really enjoying the random lfo modulation in addition to the sine btw. Currently listening to a nice drone from the two lfo’s at audio rate with slow random modulation of their waveshapes.
Discovered by chance on YouTube:
@bannelong - no prob - i’ll have them to you by that time. i have an idea for alternate samples pertaining to the nomenclature of the firmware as well that i’ve been tinkering with in the studio that may yield some interesting results…
also, do you have a file size target in mind? i can clip the tails to whatever is needed, but with cymbals, tails are everything so i’d like to go with something requiring a minimum amount of butchery in the daw. we may potentially a better result if i play with mechanically dampening the cymbal in lieu of using exponential fades… i was getting some really strange artifacts post bit-reduction depending on the curve of the edit - strange stuff.
My assessment of the PLL oscillator:
As a traditional sub-osc, It’s not that great. Its attempts to track are admirable, but the slew and glitchiness kind of limits its usefulness for this purpose (for me).
A shame because I really like those waveforms!
As a glitchy noise-maker or modulation source, it’s pretty cool!
Try piping in an analog VCO at high-frequency and modulate the PLL’s frequency and other controls. Sequence the source osc.
Peaks totally can’t keep up, so it puts out bursts of noise, chirps, and other wild sounds.
By the way, the random hi-hat mode is great.
@glitched, yes, I agree, the PLO mode has a certain glitchy charm. However, it can be made to work as a sub-oscillator quite reliably, but only with very well-behaved inputs, which limits its utility. I’ll try recasting it as an Easter egg mode, which only tracks one external oscillator on a single input channel, but distributes the derived pitch data to both output channels, which could then output signals shifted by musical intervals (e.g. by a third or a fifth or a seventh etc), not just by octaves (as Tides does) - thus making it into a chord generator (as well as a sub-oscillator). But all that is presuming that lack of processing power is the cause of the glitches. However, the code is very similar to the PLL code used in Tides, and that seems to track audio frequency inputs a bit better, but is also far from perfect - which is also part of its charm. So I think PLO mode should be retained, not matter what, but lots of caveats given about what it can and can’t do. Thanks for the feedback, though, very helpful! It is very easy to fall into the trap of finding something works in the constrained setting of my own rack, and to then assume that it must therefore work equally well in other settings and with all other inputs.
No problem. Ii’m happy to give honest and direct feedback. Often, developers (actually, coders, musicians, and creators of all types) get so excited about what a certain device or medium can do, but don’t assess whether it’s “good” or usable.
Yes, PLO mode is indeed useful and there is much fun to be had. Even changes in the source oscillator’s timbre can cause enjoyable fluctuations in Peaks. I haven’t tried percussion, but I think that’s a great idea, too.
I like your idea of using both outputs!
Update: percussion, such as bass-drums, aren’t very good for the PLO.
Yes, that’s to be expected! tHe way it works is that it collects some data about the period (inverse of the frequency) of the signal it is trying to follow, then tries to match it, then iterates a few times to see if the attempt at period matching was correct, and adjusts if necessary, and so on. With a percussive signal, it’s all over before it can lock onto the right frequency. Also, there needs to be a single dominant frequency in the signal. Percussive sounds usually have a mix of all sorts of noise at lots of different frequencies, so there is nothing for it to lock onto. And it isn’t an envelope follower - it isn’t able to follow the amplitude envelope of the input signal, because the trigger input is binary, not analogue.
Ahh, makes sense. Thank for explaining that.
At this stage, I think I might leave the PLO mode as it is - it is only a bonus extra. Trying to refine it may be a rabbit hole down which I do not wish to venture. There are still lots of other things I want to try in Dead Man’s Catch, plus I am very keen to converge on v1.0 of it ASAP, because there are other projects I need to attend to (esp. refactoring Bees-in-the-Trees).
Totally understandable! I, as well the the others, really appreciate your efforts on lil’ Peaks. You could have chosen to work on Braids full-time, but you put a lot of effort into this module–the one I have!
Sorry a wee bit off topic but i just received my Peaks module for the Call to Arms Charity auction.
Thanks MI and Luftrum …i now have 2 Peaks ok now lets DMC one
More DMC work scheduled for the forthcoming weekend. But you know, as Robbie Burns wrote:
The best-laid schemes o’ mice an’ men / Gang aft agley
it just occurred to me that a dual clock divider/multiplier would be a great addition to Peaks, and very much in the spirit of the original firmware.
I’d envisage each divider having two controls: division/multiplier and pulse-width.
I think this would be super useful!