Comment by neltnerb | Hacker News Reader

Comment by neltnerb | Hacker News Reader

As someone who is doing an awful lot of work in the supply chain, I completely agree. Open and free to copy reference designs are the only reason I can get anything done without having impractically broad and deep knowledge.

Simple example -- I make LED lighting for fun and profit. I adopted the arduino platform years ago because while I could spend the time to make sure I'm laying out all the components properly (I've done boards for lots of other microcontrollers), I hadn't done USB programming before. So instead of needing to find a partner who could write me an entire bootloader system for doing USB in circuit programming, I can just copy the arduino design, add my secret sauce, and redistribute it.

Or, more recently I've been migrating to ARM. I know zilch about ARM processors, but with the mbed platform I could easily make my light into a USB MIDI device, a DMX platform, and so on. If I had to learn myself how to do all that, it would be impossibly time consuming and I'd never get anywhere.

mbed

Oh my goodness, this is fantastic. I had been exploring another platform for this but there wasn't nearly as many examples/tutorials etc., plus it was closed source.

mbed + ethernet is wonderful. If you need some firmware for such a thing, may I suggest my own github:

  https://github.com/mfassler/eFirmata

The code is still a bit messy, but I'm still actively working on it. (I've got a bunch of changes that I haven't checked in yet.)

Thanks! I'm extremely interested in the possibilites for embedded MIDI processing, as my music-making is hardware-based and I prefer to minimize the computer use.

Sounds like a great project! I hope you share it when you get done =) I'm a big fan of electronic music, although I don't care for MIDI due to some quirks.

Are you thinking of the traditional route of making new instruments to do off-board audio generation to your mixing board?

One of my favorite projects so far was to make our LED lights MIDI devices so you could sequence them using something like Ableton directly.

http://blog.saikoled.com/post/49255468343/myki-show-tell-epi...

The second application is my favorite.

No, doing instruments is hard and at most I would build from a kit there. I already have a bunch of good quality instruments and get the sounds I like from programming those. What I'm interested in is hardware sequencing and in particular combinatorial methods for sequence generation, eg where you combine a rhythmic sequence of one length with a note sequence of another to generate changing patterns. At present I do this sort of thing with a Nord Modular but it gets unmanageable for more complex structures. Also, I don't love the sound as much as I used to but I can't give up the control over MIDI in hardware XD

I approve of the audio clips on your website BTW, we share some similar musical tastes. I'll keep an eye on the project.

Thanks! I can't say I fully understand what you mean; I was mostly taught to do electronic music using Max/MSP, so most of the stuff I make is all about generating sounds without much direct conception of things like notes or patterns. It's a little unusual of an approach, myself I do things like build simulations of throat singing based on research papers.

Here's a video I did with my throat singing simulator accompanying an LED art piece I did a long time ago.

http://web.mit.edu/neltnerb/www/artwork/ultraluminous-2008-0...

I guess I don't understand how a rythmic sequence would get so complicated as to overwhelm a computer =) Is this like Conlon Nancarrow stuff? His work is incredibly bizarre.

https://en.wikipedia.org/wiki/Conlon_Nancarrow

Here's my embedded MIDI/audio platform on the (ARM based) Raspberry Pi: http://www.joebutton.co.uk/blog/tags/baremetal-pi-audio/

I've neglected it lately but I'm currently picking it up again.

I've a feeling MBED would have been much bigger if they hadn't started off with the pure online compiler approach. That really scared off a lot of hackers. Thankfully they've fixed most of those issues (though I don't they've opened up the 'magic interface chip' firmware).

I think that the NXP LPC11U2x based boards with on-board USB don't require the second interface chip. I haven't implemented it yet, but they allow direct USB programming via the same kind of system.

http://mbed.org/platforms/mbed-LPC11U24/

The chip itself is under $5, so if you're implementing it yourself I think it comes in pretty cheap. You still need two USB ports though, which for me seems pretty irritating to explain to a customer I'd end up integrating it for.

The best I can think of is to have a USB port for the interface, and then have a hidden internal USB port that you have to take off the case to access for reprogramming the firmware. That's ugly though, compared to the Arduino style where you can use the same port for both.

Reply on news.ycombinator.com