On Sun, Aug 21, 2011 at 3:04 PM, Renato <rennabh(a)gmail.com> wrote: There's quite a range of options for working with a controller of that type. None of the solutions are perfect, and all require some adapting of the workflow to the tools. Eg: SooperLooper & JackRack can do quite some damage, as can a combo of Seq24 (with clip launch buttons) and some softsynths running Ardour busses for automateable effects.. Also I'm working on coding a live looping tool myself for a while now... :) Although I quite like the midiFighter, its price is astronomical for 16 buttons. Ya could build it using an Arduino and some parts for 1/2 the price. Cheers for the reply, -Harry

On Sun, Aug 21, 2011 at 5:48 PM, Renato <rennabh(a)gmail.com> wrote: Actually this isn't true, the new "Uno" model of the Arduino (as opposed to the ancient "Duomilanova") as it uses an ATmega8U2 USB interface chip rather than the FTDI chip the "Duomilanova" uses. Basically, you can reprogram the 8U2 to appear as a "standards complaint" USB MIDI device, and then it will just pop up in ALSA's MIDI tab. Neat eh? Then all that you need to do is write "midi data" (integers) to TX... For a "hardware" MIDI port aswell as USB MIDI just solder a 5 pin DIN to the project, using the Arduino's TX pin :) -Harry

For the sake of completeness & the archive: The AKAI APC 40 works perfectly with Linux, it appears as a standard ALSA MIDI device. This is in line with the expectations since it is marked as a "standards complaint" MIDI device on Akai's homepage. There is a mention of a "proprietary handshake" there too, so I thought I'd clarify that: The handshake are special messages sent between the APC and Ableton's Live software. It has nothing to do with the MIDI compatibility of the device under any other circumstances. As far as using the device goes, it is (IMO) quite brilliantly engineered, with some parameters remaining static, and other controllers having "banks" to switch between. Eg: Master fader & crossfader stay static, while the lower right set of 8 endless encodes switches MIDI channel, while the CC's stay the same. The Grid works in a similar manner, with each button (horizontally) sending the same note on, but on a different channel. Working with the device due to this fact is very easy, as one can do like so: parameterSetFunction( 144 - midiByte1, midiByte2, midiByte3 ); // first arg is the track you want to call on The controller / grid LEDs respond to the same MIDI message they send, wether that be a Note on or CC. Note that one must use the right channel to write to the right "bank" on the device. Hoping to post a quick demo video using it with Linux soon, -Harry PS: On a side note, its built like a tank, and the feel of the knobs & faders is fantastic :)

On Sun, Aug 21, 2011 at 09:26:09AM -0500, Gabriel Beddingfield wrote: I got a LP half a year ago, and all it takes is plugging it in. And the Novation site provides excellent documentation on the MIDI messages to/from the LP. Meanwhile I've written a C++ Python extension that connects to it using libusb (bypassing ALSA), and that also hides the MIDI details and provides access in terms of x,y, color, and up/down events directly. Works like a charm. Greetings from sunny Crete, -- FA

On Sun, Aug 21, 2011 at 04:26:20PM +0100, Harry van Haaren wrote: I found only one minor problem with the LP: it transmits MIDI using running status, the top row uses a controller while the rest uses note on/off. Even after a soft reset the LP will not tranmit the status byte until it changes. So when using the ALSA interface you don't get any events until the status changes for the first time. From then on all is OK. Actually there are only two buttons of the 80 which have the same data except status. The interface I wrote exploits this and will find the missing status byte except for those two buttons. -- FA