Fewer Oscillators = More Fun

[nickbailey]

I guess you're not too interested in this any more.

Fact is, Bernd (before he got so successful at film making he went part time) put a lot of DSP and programming in place which means I can delete a lot of the filters stuff. Also, the VLSI guys got some new kit and are doing VLSI properly so I can delete the lookup table stuff finally.

This is an interim state so there's things need fixing (exercise numbers etc). Also, there's going to be a GUI programming thing soon with Qt widgets and audio callbacks. So they will be calling methods written in C++ via swig to respond to MIDI events and controlling things like resonator Q with sliders. A bit more like audio programming? Better late than never.

N/.

PS: I didn't merge your changes first. Sorry about that. Bit rushed. Funeral tomorrow; lab Wednesday.

[gperciva]

I wouldn't say that I'm not interested -- I'm certainly more interested in "teaching" than the "publication" side of academia! But the second part of Feb is really busy for me. I'll take a look at detangling the branch conflicts in March, unless you tackle it sooner. :)

[nickbailey]

On Tuesday, 21 February 2017 09:14:26 GMT Graham Percival wrote: I wouldn't say that I'm not interested -- I'm certainly more interested in "teaching" than the "publication" side of academia! But the second part of Feb is really busy for me. I'll take a look at detangling the branch conflicts in March, unless you tackle it sooner. :)

They are nothing much. You changed some variables from class to object ones (i.e. self.x -- not sure the exact Python terminology). I've also done a new page with a simple PyQt GUI and a MIDI receiver so they can make their oscillators play. I'll put it all on the Git repo when I've written the lab up. If you're busy (glad to hear that!) you might want to ignore everything until the next pull request. I'll be deleting the lookup table and filter-by-polynomial stuff soon because it's now covered elsewhere. About time really. The downside is that possible the only true example of lookup-table length calculation on the internet will disappear! Everyone else seems to be parrotting the lies I was taught. Until you actually tested it of course. Still, good riddance. I'll leave it to the VHDL guys. In other news, it turns out that the CS department is also having trouble with new undergrads who can't formulate algorithms ─ even trivial ones. The difference is, they have developed a new course to teach it! I'm trying to go and see the chap who gives it in the next few days. N/.