Archive for January, 2012


Jack and the StyloSim

My next programming project was designed as an experiment to see if Pure Data could deal with QWERTY keyboard input, USB input and audio input all at the same time.

I planned to make an audio effects device, using a Stylophone for the audio input, plugged into the laptop’s 3.5mm (1/8″) stereo mic/line in socket; the laptop keyboard; and a nice twin-joystick device I found on eBay.

The joystick device was called ‘USB Simulator’ (‘For Airplan Heli’) and had the model number FS-SM020.  The two joysticks were quite good to use – not the kind you grip in the hand, but bigger than the thumb-operated joysticks you get on a game controller, plus each of the 4 axes had separate adjusters which you could use to fix the centre of the joystick’s range of control.  There were a couple of buttons on the front, which you can see on the bottom right and left, but these proved to be cosmetic only once I opened the case and looked inside.


I couldn’t find out much about this device on the internet, except that it was, as its name implied, normally used for flight simulation – presumably to simulate a radio-control transmitter – and, in fact, it came with a CD, probably containing the simulation program.

The case was quite large, and there would be ample room inside for further circuitry.  I may look at this possibility later, but for the moment I made no alterations to the device, intending just to use the two joysticks: a bandpass (‘wah wah’-type) filter and volume for the left one, and a binaural panning control for the right one.

The first problem to be overcome was to make sure the audio input from the Stylophone couldn’t be heard, only the output from Pure Data, which would hopefully be the audio input modulated by the joystick effects.  This was achieved by the simple-to-use and very handy audio-routing program Jack.

Jack works by running an app called JackPilot.  When opened this presents a small window like this:

Actually, the top button on the left reads ‘Start’ when opened, but I took this screenshot after clicking it and starting the Jack server.  The normal procedure is to open Jack, start all the programs involved in your routing scheme, then click ‘Start’, and then, when it’s ready, ‘Routing’.  (Just occasionally, you may have to start some audio running through a program before Jack recognises it, but mostly that isn’t necessary).

The Routing window looks like this:


All the inputs are in the left-hand column, all the outputs on the right; it always shows the System audio in and out.  In this instance I was just practising, so I only used Pure Data, but I would frequently have Logic running, too.  To set the route, click on all the inputs and outputs in turn (they are shown in red when being edited), and click on the item(s) you want them to be connected to before moving on to the next one.

At the beginning of the Pure Data program I put the ‘Print’ and ‘Open’ instructions, as I did for the Theresynth and Cybersynth (see previous posts):

This configuration enabled me, once I’d plugged in the Stylophone (audio input) and Simulator (USB input),to check the Device number of the Simulator and open the Device so the ‘hid’ object would receive and deal with the output from the two joysticks.  When I clicked the ‘Print’ message box at the top, the following was printed in the Pure Data window, telling me that the Device Number was ‘0’:


(It also gave the Device’s name ‘FS-SM020’, which it hadn’t done with the particular game controller I used for the Theresynth and Cybersynth).

As the Device number was 0, I amended the ‘Open’ box to show this number, and clicked it.  Then I clicked the ‘Print’ message box at the top again, and it showed me the controls it could recognise:


The 4 axes of the joysticks can be seen (‘abs_x’, ‘abs_y’, ‘abs_z’ and abs_rx’, I think, and I can’t remember if any input was recognised from ‘abs_ry) – and also 3 buttons, but the button function hadn’t been implemented in this device.

I clicked the on/off toggle switch, applied the stylus to the Stylophone keyboard, and sounds issued forth – a very good sign!

The Pure Data StyloSim program had three inputs:


On the left is the [key] object, which receives input from the laptop keyboard.  In the end, I added only the box with [== 114] in it: 114 is the ASCII code for the ‘r’ key, so the program only responds to the letter r, turning a reverb effect on or off.

[Edit: I later changed this to 119, the letter ‘W’, to avoid conflict with another program that used ‘R’ for a different function]

In the middle is the audio input, [adc~] (analogue-to-digital converter), which leads to the output stage.

On the right is the [hid] (Human Interface Device) object, receiving input from the 4 axes of the joysticks.  The [route] objects separate the 4 inputs and send them in different directions for their various functions (bandpass filter, volume and binaural panning).

All the effects worked – although the panning was less binaural than left-to-right; I’ve heard better.  This is the complete patch, showing the objects I used, and the mathematical calculations required to get it right.


The patch can be downloaded from here.  (Click ‘Save Link As’).

This proved that Pure Data could combine information from the keyboard (or an external USB QWERTY keyboard) and [hid] to work on an audio signal entering simultaneously.   As  can be seen from the following picture of the set-up in action, I used ‘The Alien‘, my first Stylophone modification, as the audio source.


This instrument has a ‘drone’ function, so it’s possible to put the stylus down and operate both joysticks on the Simulator at the same time.  Both the joysticks, incidentally, are centre-sprung, meaning that all values return to the central ones, except the y-axis of the left one – the bandpass filter.  This means the centre-frequency of the filter can be set and left while other parameters are adjusted.

Here’s a sound file illustrating the StyloSim:

It’s more of an example than an actual piece, just to show what the StyloSim does.

It’s slightly edited, but only to make it a bit shorter, otherwise it’s exactly as it sounded when played. There’s no post production effects on it, I just hit the ‘r’ button to turn on the StyloSim’s reverb before I started.

There are three parts: the first part is the Stylophone played with the stylus in the conventional manner; the second part uses The Alien’s ‘drone’ feature and variable pitch and vibrato controls; the third part is played with the ‘feedback’ switch on.

[Edit: more modifications have been made to the StyloSim joystick (although not for this particular use).  See this post].


Cybersynth – the Theresynth improved

After finishing the Theresynth I thought of a way it could be significantly improved.

A little while before, I had experimented with alternative methods of inputting QWERTY keyboard information to the computer, by removing the circuit board and attaching different kinds of switches (See ‘A New Use for USB Keyboards‘).  I reasoned that the same might apply to gamepads/game controllers: as long as the USB chip inside was still attached to the USB output, it would still be recognised by Pure Data, no matter what kind of switches were used to control it.

In order to test this theory I bought another game controller of the same type as the Theresynth (A ‘PCLine Rumble Game Pad’), took it to pieces and threw away the buttons it came with, leaving just the joysticks and the USB cable attached to the circuit board.

Just as I had done with the dismantled QWERTY keyboard, I scratched away the surfaces of the tracks where a couple of the switch connections had been, soldered wires to the board and push-button momentary switches to the wires.  I plugged the board into the computer via its USB cable and opened the Theresynth patch in Pure Data.

I followed the steps to set it up, as I described in the Theresynth post, and, sure enough, Pure Data recognised it, just as if it was the game controller, rather than a circuit board hanging off the end of a USB cable; when I pressed the test buttons I had soldered onto the board, the ‘hid’ object recognised them just the same as the buttons on the original gamepad.

Confident that I could operate the Theresynth with any type of enclosure for the circuit board, I considered how I could rehouse it.  An obvious thought struck me: I had recently bought a plastic Cyberman head (if you aren’t a fan of Dr Who, they have their own page in the Wikipedia at for about 50p from a local charity shop, and it was clear that the eyes were spaced apart almost exactly the same as the distance between the joysticks.  This was a clear sign of what I should do next.

In short, I transplanted the gamepad circuit board into the Cyberman head, with the joysticks in the eye sockets and 12 or 14 buttons round the outside in a vaguely ergonomic layout.  A chance visit to a Poundshop during the course of construction produced a nice USB-powered chain of blue LEDs, and I glued these at stategic points on the inside, producing an eerie glow when the instrument was plugged in.  Another spare USB Type A socket from one of my broken Apple keyboards took the place of the hard-wired output lead, and the Cybersynth was ready for action.


I took the opportunity to make some improvements to the Theresynth patch in Pure Data, adding, amongst other things, a facility to change the starting note from A to any other suitable pitch.  I automated some of the device number/control checks at the beginning, and the switching on of Pure Data’s audio.

The starting window now looked like this:


The Pure Data patch for the Cybersynth is here.  (Click ‘Save Link As’)

This photo guide, which I produced for myself before putting the circuit board inside the head, gives an indication of what the buttons are for, and what functions are included in the patch:

Cyberman Controls

I never figured out quite what to do with the former hatswitch controls.  In the end I stuck the switches on the back of the neck and used them for a sort of trill effect – the top button raises the pitch by a 5th when pressed, the bottom one lowers it by a similar amount.


The Theresynth

I recently blogged about using a gamepad as a drum controller (, and this was my next project using a Human Interface Device (HID).

HIDs are the devices we use all the time to interact with computers.  Keyboards, mice, trackpads, trackballs, joysticks, gamepads/joypads/game controllers and graphics tablets are all HIDs, and all can be turned to musical use with the aid of a suitable program to interpret their output.  I chose to start with projects using game controllers because they have a fair number of buttons – and often joysticks – and are designed to fit comfortably in the hand (or hands).

The gamepad I used for the drum controller had just buttons, but the one I used for the Theresynth had two thumb-operated joysticks.  I planned to use one of these for controlling the pitch of an instrument.  In addition, although I had used the pre-made MAX/MSP app ‘MultiControl’ for the drum controller, I planned to use Pure Data and do my own programming.

This was not as reckless at it seems, as programming in Pure Data is done graphically.  Many mathematical and audio functions are pre-programmed for you: all you have to do is draw boxes, write in them what function you want performed, and join the boxes together in sequences.  Easier said than done, as I discovered, but easier, I thought, than writing lines of code.

I have described progamming in Pure Data in my post about creating alternative musical keyboards, using QWERTY keyboards (  There, ‘key’ was the main Pure Data function used, as this is designed to receive key presses from the QWERTY keyboard; in the application I’m about to describe the special function ‘hid’ was the key to interpreting the output from the game controller.

There are many Pure Data instructional videos on YouTube, and this one in particular: deals with ‘hid’ and game controllers.

If you’re working on a project like this one, the first thing to do is to allow ‘hid’ to get Pure Data to recognise the game controller.  To do this, plug in the game controller, run, open ‘New’ from the ‘File’ menu and write (or draw, rather) the following into the empty box ‘Untitled-1’.

This is a short program which will show all the connected USB devices which Pure Data can recognise.  The ‘Print’ command is in a ‘message’ box, and is connected to the input of the ‘hid’ object.  Inputs are on the top of object boxes; outputs are on the bottom.  When you click ‘Print’, it sends the message, and gives a reply like this in the Pure Data window:


Each connected device has been given a number.  The first six, ‘Device 0’ to ‘Device 5’ are simply the different elements of my MacBook, keyboard, trackpad, and so on; but the seventh one, Device 6, looks like the important one: ‘4-Axis, 12-Button with POV’, ‘device type: joystick’.

Sometimes it will tell you the make and model; in my case, my ‘PC Line Rumble Game Pad’ is identified only by its ‘vendor ID’, 0x0583, and its’ product ID’: 0xa009.

We need to know the Device number, as we’ll use that in the program.  It doesn’t always come up as Device 6, so I always include in my Pure data programs a facility to check this number, and amend it if required.

The next step in writing a Pure Data program based on the ‘hid’ object, is to identify the outputs from the buttons and joysticks  in the controller.  This is done by a couple more additions to the simple program which produced the list of devices:

‘Open 6’ is another message box.  The small square on the left-hand side is an on/off ‘toggle’ switch, which I won’t be using now, but I’ll use it in a minute.

Click ‘Open 6’, and messages from Device 6 are routed to ‘hid’; click the ‘Print’ message box at the top again and the ‘Print’ object at the bottom will produce a list in the Pure Data window of all the buttons and controls in the device:


This shows 12 buttons, (‘btn_0’ to ‘btn_11’), the up-and-down and left-to-right axes of the left-hand joystick (‘abs_x’ and ‘abs_y’), the up-and-down and left-to-right axes of the right-hand joystick (‘abs_rz’ and ‘abs_ry’) and a hatswitch.

Now you know this, you can separate each of the inputs as they come out of ‘hid’ and get them to do various things – make noises (as I did with the gamepad MIDI drum controller), control effects (like filters, for example), and so on.  Other Pure Data objects (boxes like ‘hid’ and ‘print’) can do mathematical transformations along the way.

The way to separate them is in stages.  First of all, look at the left-hand column above.  There are two types of entry here: ‘abs’ and ‘key’.  Using a Pure Data object called ‘route’, we can get the ‘key’ information (from the buttons) to be separated from the ‘abs’ information (from the joysticks and the hatswitch):

Then, looking at the second column, we  can separate out the 4 joystick axes and the 12 buttons:

While I was testing this out just now I did two extra things, which you can see in the screenshot above: firstly I added the small box at the bottom, which is called a ‘bang’.  If everything’s working properly, this will flash on and off when I press the button ‘btn_0’ – which is very likely marked ‘Button 1’ on the game controller – and secondly I clicked the on/off switch (the cross means ‘on’) so that ‘hid’ would recognise and output button presses and joystick movements.

Because the MIDI note number and control system lends itself easily to mathematical manipulation, this  would be a common way to use Pure Data.  In this particular application, however, I decided to use some of its many audio functions.

I’d spent some time modifying stylophones, as you may have read in previous posts.  The thing about stylophones is getting great sounds from them is easy, but getting sinewaves out of them is well-nigh impossible.  In Pure Data, however, it is extremely easy, so I decided to create an instrument which would produce ethereal sinewave sounds, rather like the Theremin.

The Theremin is a wonderful electronic musical instrument, which you play without touching it!  If you don’t know what it is, you can read about it on the Wikipedia at

My plan for the ‘Theresynth’ was less ambitious, but would involve using one of the game controller joysticks (the left one) to control pitch and the other (the right one) to control volume.  In the event, I also added a kind of vibrato (more of a fast auto-pan), controlled by the other axis of the right-hand joystick.  This added a little more depth and interest to the sound.  As well as this, a programming error I made along the way also enabled me to add a second tone to the instrument, which was a kind of frequency modulation/ring modulation – very effective in the lower register.

The exact method by which I achieved this is rather lengthy to explain, and I’ll attach the ‘patch’, as bits of Pure Data programming are called, so you can see, if you want to study it.  In principle, though, it went like this:

The output from the the joysticks is in the form of numbers from 0 to 255.  Divide this in half and you get all the MIDI note numbers (they run from 0 – 127); convert the MIDI numbers to frequencies and use this as the input to an oscillator, feed the oscillator into an output amplifier and you have a simple musical instrument!  All these functions are available in Pure Data.

Using ‘abs_x’ (left-to-right on the left-hand joystick), as I did in the Theresynth, that part of the circuit would look like this:

‘/ 2’ means ‘divide-by-2’; ‘mtof’ means ‘convert MIDI note to frequency’; ‘osc~’ is an oscillator; and ‘dac’, the digital-to-analogue converter sends audio out to the computer’s sound system.  The number 64 is is in a ‘number’ box – it doesn’t do anything, it just tells you what MIDI note is passing from ‘/ 2’ to ‘mtof’.  Useful if you want to check what’s going on.

This program works – I’ve just tested it (but don’t forget to turn ‘audio ON’ from the ‘Media’ menu of Pure Data) – but the problem is that the travel on a gamepad joystick is not that far, and you don’t want to go from the lowest MIDI note possible to the highest in such a short space.  So a lot of what I had to do was to restrict the range of notes available.

In the end, as you’ll be able to see in the finished patch, I did this by dividing again by 10, severely restricting the range of notes available to the joystick at any one time; and then using one of the buttons as a freeze-pitch button, with a very short time-delay to allow the joystick to spring back to the middle.  In this way it’s possible to play a few notes, ending higher or lower than where you started, freeze the pitch there, let the joystick go back to the middle at this new pitch, and gradually move up and down in steps, as required.  After a little practice, this came to seem quite natural.

I never quite got the hang of creating a nice interface for my Pure Data programs, so the window that opens looks like this:

Following these steps gets the Theresynth ready to use.  Steps 1 – 4 just operate the checks I described above: identifying the device number, checking the buttons are all recognised by the program, opening the device and switching on ‘hid’.  Step 5 reminds me to turn the audio on – although in later programs I found out how to do this automatically when the program loads.  Step 6 may or may not be necessary, but clicks some toggle switches on and off to guarantee everything is ready to work, but stays silent.  Step 7 turns up the volume and you hear the initial tone, which is A (440Hz).

The actual programming is visible in two windows, ‘controls’ and ‘oscillators’, which can be opened by clicking the boxes in the bottom right.

In my next post, I’ll describe how I improved this design, but the instrument definitely worked, and seemed to me to be a good use for a second-hand joystick.



Drums with MultiControl v0.4

Here’s a method I used to play a virtual synth drum patch with a cheap (£3 or £4 off eBay) USB game controller and the program MultiControl v0.4.

Here you can see the game controller I used (the type without joysticks), and the MultiControl window.


To set it up, plug in the game controller, open MultiControl and select ‘USB Gamepad’ or similar in the top right of the MultiControl window.  There’s a drop-down list of devices that the program has detected, including the keyboard/mouse/trackpad of your computer, but it should be obvious which one is the game controller you want to use.

MultiControl is a MAX/MSP-based application, and will also open the MAX/MSP Status window.  To start with, this window will probably show you some information relating the keyboard/mouse/trackpad of your computer, but once you’ve selected your game pad at the top, all you need to do is press each button: the program will detect each one and allocate it to a separate line in the MultiControl window.

There’s a video on YouTube explaining how to set up MultiControl, and how to use it.  If you’re intending to do this, it would be good to look at that, here:

This is the MAX/MSP Status window.  On opening the program, it registered the information in the top half of the window; once I selected ‘USB Gamepad’ it began with ‘hi: focussing on USB Gamepad’ and waited for me to press the buttons; as I did so, it listed them (there are 11 on this device) and allocated them to lines 1-11 in the MultiControl window.

MAX Status window


This is a general screen example, not from this Gamepad application, but you can see that the first column, ‘Input’, registers button presses – the shaded bar moves across the box as the button is pressed – and the 7th one, ‘Values’, shows (in the case of this Gamepad application) ‘0’ when the button is not pressed (i.e. ‘off’), and ‘1’ (i.e. ‘on’) while it’s pressed – except for the hatswitch buttons, which can register 0, 0.5 or 1 at different times.

The 4th column shows each button’s control number, ‘c1’ being line one, ‘c2’ being line 2, and so on.  This information will be useful later on.

The rest of the left-hand side of the MultiControl window is not important for this purpose, but the right-hand side, the MIDI side, is.  To allocate the 11 controls to particular MIDI drums, I had to turn MIDI output on by clicking the small box at the top marked ‘MIDI on’; change the first 11 boxes in the next  column to ‘noteout’ rather than ‘ctlout’; and set the MIDI note number for each in the next column ‘MIDI range’.

You can see this in the instruction video, the range being set by clicking and dragging; but in the video you see a rather vague range being set.  The boxes in the ‘MIDI range’ column are so small that it’s very difficult (well, I found it impossible) to set precise values – it was OK starting from 0 on the left-hand end of the box, actually, but getting the upper value correct was the hard bit.  Unfortunately, this upper value is critical, as it represents the exact MIDI note number of the drum sound you want to make when you press that button.

However, this can easily be got round.  This is because it’s possible, using the ‘Save file’ box at the bottom, to save your settings and then retrieve them later, so you don’t have to input them all again every time you want to use a particular gamepad or other device for a particular purpose.  I used this feature to input the exact values I needed, in the following way:

First of all I worked out which buttons I wanted to operate which drums.  Practising with the virtual synth I intended to use, I noted which drum was which, and which note of the keyboard produced that particular drum sound.  Using my chart of notes and MIDI numbers I listed the 11 buttons and wrote the appropriate MIDI note against each one.

MIDI Int MIDI Note No Chart

By clicking and dragging from the left hand side to somewhere in the middle of each of the 11 boxes in the ‘MIDI Range’ column, I produced a number in the next column, ‘values’, that was pretty close to the MIDI note number I needed for that particular drum sound when the button was pressed.  In every case I had been able to get it to show 0 when the button was not pressed, which is what was needed.  I then saved the settings file with an appropriate name.  These files have the extension ‘.xml’.

I then opened the .xml file in a text program (e.g. BBEdit).  Apart from some general stuff at the beginning, each button had a long entry between the tags <patcher name => and </patcher>.  They didn’t come out in order from 1-11, but this is the entry for the button described on line one, as you can see from the line near the bottom, which says “text edit” value = “c1”.


You could, I suppose, edit any of the values in here.  I’m not at all certain what they all are, but the one that needs editing in this instance is the 4th line: “X Range” value = “0 51”.  The meaning of this is that pressing the button moves the value output from o (no note) to 51 (D#3 – can’t remember off the top of my head what drum sound that represented in the virtual synth I was using).  If you go through each of the entries and make sure that the first number is 0 and the second one is the MIDI note that button is supposed to output, you can save the file when you’re done, with all the right numbers in it.

Next time you want to use this controller for this application, you can just click ‘Open file’ and ‘Load’, and all your saved settings will appear.

The only remaining thing you need to know is how to set up the inputs and outputs.  MultiController has a window for this, which appears when you choose ‘MIDI Setup . . .’ from the ‘File’ menu.  My notes to myself say to set it up like this:

MIDI setup window

(Although it seems likely that I meant “to MultiControl 1” or “to MultiControl 2”  and  “from MultiControl 1” or “from MultiControl 2”.  If no other MIDI devices are being used at the same time, it almost certainly doesn’t matter).

You might need to use the IAC Driver Bus 1 or 2, depending on the number of MIDI devices you’re using at the same time, but whatever you end up doing, you’ll need to change the ‘MIDI Device’ setting at the bottom of the main MultiControl window, and the input device setting of the synth you’re using, so that they agree.

If you need to change the ‘Audio MIDI Setup’ (on a Mac, at any rate), clicking the ‘System Setup . . .’ button in the bottom left of the MultiControl MIDI Setup window will open that for you, instead of you having to go to ‘System Preferences’.


A new use for USB keyboards

In the course of creating alternative keyboards, and interpreting USB inputs, I had an idea for a project which could prove to be useful.

It occurred to me that the computer was able to recognise the input from QWERTY keyboards because of the chips and circuits inside the body of the keyboard – the actual switches themselves, how they were laid out, what exactly they represented, were irrelevant.  So, if you had access to the circuit board inside the keyboard, and left the USB connection intact, you could do what you liked with the switch connections.

The keys of a QWERTY keyboard are arranged in a matrix – I had no idea how many connections there would be in each axis, and it probably differs from one model of keyboard to another.  But I had an old Apple keyboard I was about to throw out because of sticky keys, and I decided to experiment on this.

Blueberry keyboardsm

Here’s what I did:

First of all, I took it apart, threw away the casing, put the switches to one side, and detached everything from the circuit board, except the USB lead.

As it happens, the circuit board in this keyboard is unusually large – covering the whole width of the keyboard, and with extra USB sockets on each end.  Normally, the mouse would be plugged into one of them – and they aren’t the same: the one with the lead attached is the one that has to be connected to the computer.

The extra USB connectors on the ends of the board would probably be useful later, but more important was the circuit board itself.  Here you could see the two series of tracks for the two different axes of the switch matrix.

In theory, each track in the first series – which I called Series A – would give a different output when connected to each track in the second series – Series B.  There was only one way to find out: I plugged the USB lead in to the computer, opened Word, and got to work with a lead and crocodile clips.

If I remember correctly, I had to scratch off the track surface a little to get a good contact, but starting on the left I clipped the lead to the first track, A1, and progressively touched the other end to B1, B2, and so on, noting in a table what letters or symbols appeared.  When I reached the end of the Series B tracks, I reattached the lead to the second track on the left, A2, and repeated the procedure, touching each of the tracks in Series B and noting the results.

Not all the combinations seemed to work, but eventually I produced the following table:

Apple keyboard matrix screenshot

This list seemed to contain all the letters of the alphabet and all the numbers 0-9, so I considered this a great success!

It seemed to me that tracks 1, 2, 4, 5, 6, 7, 9, 10, 11 and 12 in Series A, and tracks 2, 3, 4 and 11 would be the useful ones – 14 connections giving 40 outputs.  It wouldn’t matter what kind of switches were used to make the connections, their output would be predictable and the ASCII codes they produced could be interpreted reliably by programs such as Pure Data.

All I had at this point was a circuit board with a cable attached to it, so I decided to put it in a permanent box.  I had a suitable one to hand, which it would just fit in, and decided to use this, together with sockets on the rear connected to the 14 tracks.  I gingerly soldered ribbon cables to the two series of tracks, and this photograph of the inside of the box shows how it looks:

MIDI CPU insidesm

You can see the leads from Series A on the left-hand side and the leads from Series B on the right-hand side, leading to a circuit board with PCB connectors.  These are connected to 3 different sockets on the back of the box.

There are 3 sockets because I decided that not every potential application would require the full 40 connections.  There’s a 15 way DB15 HD socket (the type used for VGA) for any application requiring the full 14 connections (10 x 4 matrix covering all the letters A – Z, numbers 0 – 9, and a couple of others); a DB9 socket for applications requiring only 7 or 8 connections (4 x 3 or 4 x 4 matrix, e.g. a numeric keypad); and a 4-pin socket (the type used for S-Video) for an application requiring only 4 connections in a 2 x 2 matrix.

MIDI CPU Box rear viewsm

The original circuit board, as described above, had a hard-wired USB out lead and a USB A socket on each end.  As you can see from the above pictures, I took off  the hard-wired output lead and replaced it with one of the sockets from one end of the board.  I know you’re supposed to have USB B, not USB A sockets in this position, but it’s easier to come across USB A to USB A leads, and the functioning is still the same.

The large DB25 socket, and the two 5-pin MIDI sockets on the left are for a future project, not yet finished, for a fully-featured MIDI controller.  I’ll describe this in a later post, when I get round to constructing it.

I also bought a supply of leads, ready for when external switching devices need to be attached.  In some cases this was easier than others.  I chose these formats partly for the availability and inexpensiveness of the leads, but was unaware that the 15-pin HD leads commonly seen are designed for VGA or S-VGA, which don’t have all 15 pins connected one-to-one.  In the end I found the ones I needed, but if you’re doing the same kind of thing, take care.

The final thing to note is the circuit on the left-hand side of this picture:


Although not strictly necessary at this stage, I decided to add this feature while I was working on the box.  The MIDI controller, planned to be in the same enclosure, would need a USB converter in order to be attached to the computer.  The only USB converter I have at the moment might be required in future for another device, so I decided to add a MIDI to USB converter to this box, which the MIDI controller could use if necessary.

I took a chance on a cheap MIDI converter, which was £5 on Amazon.

I had read bad reports of these things, but they’re so cheap compared to the brand name converters, I thought I’d see what they were like.  You can get them for even less than £5, direct from China.  When I got this one, I took the circuit board out of its plastic casing, chopped off the the in/out leads and rewired them to the sockets on the back of the project box.  It was powered via USB, and appeared to work fine with my Sonuus G2M Guitar to MIDI converter, so I took the final step of cutting off the USB lead and attaching the USB output to the spare USB input on the Apple keyboard PCB.  (It was spare because, as mentioned above, I took the socket off to create a new USB output socket instead of the hard-wired lead).

It still worked, so I put the lid back on the box, and called it a day on that project for the time being.  When I get round to creating some alternative switching devices, or start on the MIDI controller, I’ll post something on them.

A brand of keyboard with a smaller PCB would probably be more practical, and any keyboard will do, if you want to try it yourself.  There’s an excellent video on YouTube, here:, which goes through the steps very clearly.

It just so happens that I saw a job lot of broken Apple keyboards for sale like the one I used here and bought them for about £1 each, so I’m stuck with these for future experimentation!  The faults they had were various – sometimes it seemed to be the circuit boards, sometimes it was just the odd key, so all of them looked usable in one way or another.

I may not need the circuit boards – although I could make self-contained instruments if the cases were big enough to hold them – but I have plans for the keying mechanisms, which could be interesting.


Alternative Keyboards 4

After creating the isomorphic alternative keyboard with a Cherry USB QWERTY keyboard and the program MIDI Integrator (see my previous post ‘Alternative Keyboards 3‘), I started to think about the possibility of using two keyboards, like the Jammer.

The Jammer uses MIDI Integrator, but it can do this because it uses two AXiS-49 MIDI keyboards.  It wouldn’t be possible for me do the same with two QWERTY keyboards, for two reasons: a) MIDI Integrator only has the facility for mapping one QWERTY keyboard at a time, and b) MAX/MSP does not recognise alternative key codes with the ‘Shift’ key pressed – i.e. both ‘q’ and ‘Q’ are interpreted the same, even though they have different ASCII codes.  This would mean that however many keyboards you had, they would all output the same notes to the same synthesiser or sound card.

So, I needed to find an alternative program for this application, one that would recognise two QWERTY keyboards, be able to interpret their outputs separately, and potentially route them to two different synths.

A program that will do this is ‘Pure Data’, a relative of MAX/MSP.  Like MIDI Integrator (but unlike MAX/MSP itself) it is free to use, and programming is done graphically.  Full details are on the Pure Data Community site at

A piece of Pure Data programming looks like this:


You create the different types of boxes, write in them the values or functions and sequence them by joining them together with lines.  In the above example, the message box at the top contains the number 60, the MIDI note number for middle C, and a little lower down is a box with the function ‘mtof’, MIDI-to-frequency, which converts the MIDI note number to a frequency measured in Hertz.  There are 4 boxes marked ‘osc~’, which represent sine wave oscillators, mixers (‘*~’) and a graph which will display the complex waveform created by the summing of the 4 sine wave oscillators.

Amongst the many functions available in Pure Data are ‘key’, ‘keyup’ and ‘keyname’, which are designed to accept input from a QWERTY keyboard.  The function ‘key’ accepts and outputs the ASCII key code when a key is pressed, ‘keyup’ does the same when the key is released, and ‘keyname’ tells you the name of the key in use.

Most importantly, Pure Data accepts all ASCII codes, and can therefore distinguish between simple key presses, and key presses + Shift.  This means you can have two whole keyboards-worth of different key numbers, making it worthwhile using two keyboards, with no ASCII code being duplicated in both of them.  As long as the program could be made to understand which numbers belong to which keyboard, it would be possible to treat each keyboard differently, and route their outputs to different MIDI synths.

So what I needed was a) a second Cherry Keyboard, and b) to learn how to program Pure Data to distinguish key presses from one keyboard or the other; divide them into two streams and deal with them separately; and use the ASCII codes received to output appropriate MIDI note numbers.

The second keyboard was easily found on eBay.  It had a PS2 connector, and it was from this experience that I gained the knowledge of PS2 to USB converters which I wrote about in this post:  It functioned perfectly with the aid of a ‘Blue Cube’.  I also removed the keycaps, cleaned the (second-hand) keyboard, and  painted the keycaps blue.  It wouldn’t matter which keyboard was used for the left hand and which for the right, as they are identical models, but I tend to use the blue one for the left hand and the red one for the right.

Dual Cherry Keyboardssm

They’re set up here for testing purposes only.  Having two separate keyboards for left and right hands means that in practice, they can be placed in a more ergonomically suitable fashion for comfortable playing.

(The eMac they’re connected to is almost worthy of a post of its own.  If you’re a Mac User, and you’ve got room for an old-fashioned CRT-screen desktop, it’s worth considering an eMac: nobody wants them, so they’re very cheap: this one, which was £20 off eBay has the same OS as my laptop – not the latest Mac OS, but a perfectly respectable one; it has firewire as well as USB and Ethernet, and a DVD ROM/CD burner drive.  I’m able to use it as a dedicated MIDI machine, and use the laptop for soft synths and recording.

Upstairs in the Music Room I have 2 iMacs, which are similar in appearance, but slightly smaller.  These aren’t as good as the eMac, and aren’t able to run such a recent OS, but are still able to perform some useful tasks as VST instruments hosts, MIDI control my Behringer V-Amp, and so on – and they were £5 each!)

The USB hub I installed in the Superstylonanphone came in handy at this point: although the eMac has 3 USB connectors, I was beginning to run short, as both of the Cherry keyboards needed to be connected individually, as well as various other devices I was using at the time.

Not shown is a vital piece of equipment required for the blue keyboard, which is a clip to hold the Shift key down.  In order to take advantage of the two keyboards-worth of ASCII numbers recognised by Pure Data, it was necessary to do this – Caps Lock only affects the letter keys, not the others – and I haven’t yet found the best way of doing it.  The next thing I’ll be trying is a drawing board or table cloth clip (it has to be thin and flat on the bottom); but in any event, it’s vital that one of the keyboards outputs the usual numbers and the other one outputs the numbers with the Shift key down.

Turning to the software, the two difficulties would be a) getting the program to distinguish the numbers received by one keyboard from the other, and b) getting it to output the appropriate MIDI note numbers.

The entire program looks like this:

Cherry Double screenshot

The function called ‘key’ in the top left receives the numbers from the keyboards.  It neither knows nor cares whether these numbers are coming from one keyboard or two.  The function ‘keyname’ is just there for the sake of interest – originally for the sake of testing! – and doesn’t actually contribute to the working of the program (or ‘patch’, as they call it in Pure Data).

The number from the keypress is passed to a very complicated sifting system underneath, the purpose of which is to divide the numbers coming from the red keyboard from the numbers coming from the blue keyboard.

Unfortunately, as I said in the previous post referred to above, there is no simple way of distinguishing the ASCII numbers – they’re pretty random, it seems.  I’ve used a function called ‘moses’.  Like Moses, who parted the Red Sea, this function parts the stream of numbers coming into it: you put a number in the box and it sends this number and any higher numbers out of one output and all the lower numbers out of the other output.  In this way I was able eventually to divide the numbers and send them in the right direction.

In the middle part of the program you can see the two separate streams of note numbers moving from top to bottom; the settings on the right-hand side of the program allocate MIDI channel, velocity, duration and octave information.  At the bottom, the Pure Data functions ‘makenote’ and ‘noteout’ convert all these details to the kind of MIDI signal understood by a synth.  You just need to tell the program where exactly to send these MIDI signals in the program ‘Preferences’.

That just leaves one question: how does the program know how to convert the incoming ASCII numbers to the correct MIDI note numbers?  The answer is in a similar way to MIDI Integrator in the previous post: by means of a pre-written text file.  In the bottom right of the program, there is a list (very incomplete at the moment as it only contains ‘Wicki-Hayden’, ‘Janko’ and ‘System C’, and only ‘Wicki-Hayden’ has so far been implemented!) of layouts which can be used.

When you select the ‘Wicki-Hayden’ option, the instruction at the bottom gets the program to read the text file ‘wicki.txt’ into a table (or array) called ‘notes’; the function ‘tabread notes’ in the middle of the program uses the information in this table to convert the ASCII numbers it receives into the MIDI note numbers you want it to output.

The format of this text file is quite different from the text file in MIDI Integrator.  It looks like this:

wicki txt screenshot

On the face of it, pretty meaningless!  The way it works is this: the table is 128 numbers long.  This is because there are 128 possible MIDI note numbers, and why it says in the instructions at the top of the program, in the middle, ‘Open ‘table notes’.  Set array size to 128′.  The button for opening ‘table notes’ is right down near the bottom of the left-hand side of the program window, not right next to this instruction, as you would think it should be.  Don’t ask me why.

So the table is 128 numbers long, and what the program is doing when it receives a number is this: if it receives the number 60, it outputs the number at position 60 in the table; if it receives the number 113, it outputs the number at position 113, and so on.

This is more complicated than the method for MIDI Integrator, but can nevertheless be achieved with a little patience and the correct data to hand.  Here’s what I did:

First of all, as before, I filled out the spreadsheets showing the layout of the keys, the ASCII numbers and the MIDI note numbers (see  In fact, the sheet for the red keyboard was the same as the one I used in MIDI Integrator, it was the new sheet for the blue keyboard that was different.  To fill this in, I had to change the MIDI note numbers and the ASCII numbers.

In fact, I could have left the MIDI note numbers the same as the red keyboard, but I thought it would be more practical to have them an octave lower, so I entered all the MIDI numbers with 12 deducted from each one – that’s all it takes to lower the notes by one octave.

The ASCII codes had to be changed because the blue keyboard would output the number of the Shifted keys – i.e. ‘Q’ instead of ‘q’, ‘%’ instead of ‘5’, etc.  I used this chart to change the numbers on the spreadsheet:

[Edit: to be clear, this is a list of all the MIDI notes and standard ASCII codes from 0-127.

In the chart below, Column 1 shows the names of all the 128 MIDI notes available, from the lowest to the highest; Column 2 shows the MIDI number associated with each of these notes – these are also the ASCII numbers which my Pure Data patch uses; Column 3 shows the key that produces this ASCII number; and Column 4 gives a description of which key this is].

ASCII Note chart screenshot

[Edit: In fact, on UK keyboards the symbols ” (‘Quotes’) and @ (‘At’) are reversed, so ” = 64 and @ = 34]

The one obstacle I came up against was the F keys, which don’t have an alternative number when used with the Shift key.  This meant that the blue keyboard was restricted to 4 rows – but as this is more than 40 notes, it wasn’t too much of a limitation.

[Edit: looking at the chart above,  you can see that the F keys don’t have an ASCII number in any case, so neither of the keyboards can use this row of keys with the Pure data patch, even though the single keyboard could with the program MIDI Generator (see ‘Alternative Keyboards 3‘), based on Max/MSP, which uses a different, non-ASCII system.  This problem may be overcome with some keyboard remapping.

Edit 2: I’ve now used KeyRemap4Mac – now called Karabiner – to remap F1 – F8 as 0 – 7; I couldn’t find a way of remapping F9 and above].

Here’s how I created the text file:

First of all, the first position in a 128-position table is 0, and the last one is 127.  So I opened a blank text document and wrote a column of numbers down the left-hand side from 0 to 127.  These are the ASCII key numbers.  I then went through the two spreadsheets and for each key I needed to use I wrote down the MIDI number against its ASCII number in the text document.  When I finished, I put ‘0’ against every unused ASCII number in the list.

When I had a list of 128 numbers – many of which were now 0 – I removed all the ASCII numbers in the left-hand column, left one space between each remaining MIDI note number, and saved the text file as ‘wicki.txt’.  This file is now read at any time when ‘Wicki-Hayden’ is selected as the layout in the program, and the values don’t need to be entered again.

The patch and the text file are here: Double Keyboard patch; Wicki-Hayden layout text file.

See also this post: ‘Alternative Keyboards 5 – Postscript‘ for a little extra info on the Cherry Double Keyboard.


Alternative Keyboards 3

In my last post (‘Alternative Keyboards 2‘) I described choosing a suitable QWERTY keyboard as a 2-dimensional ‘isomorphic’ musical keyboard.  In this post I talk about the software I used to interpret the key presses and output MIDI information to the synthesizer (virtual synthesizer, in my case).

The first thing to appreciate is that every key press on a computer keyboard sends a number (its ‘ASCII’ code) to the computer via USB.  Every key has a different number, and the number sent out by the key is different when the Shift key is pressed at the same time.  Knowing this, and knowing which key sends which number means you can tell the computer to do a certain thing with each number it receives – for example to output a MIDI note number.

When you use a MIDI music keyboard the relationship between the number it sends and what you want done with it is very simple: press the note C in the middle of the keyboard, and it will send out the number ‘60’.  MIDI synths recognise 60 as the number for C, and C will be sounded; press C#, the number 61 will be sent, and C# will be sounded; and so on.  Unfortunately, this isn’t the case with the computer keyboard – the numbers sent out by the keys bear no relationship whatsoever to where on the keyboard they are: The ASCII code for Q, expressed as an ordinary decimal number, is 113; W, right next to it, is 119; E is 101, and so on and so forth.

So, in terms of software, if you’re using a QWERTY keyboard, as I was, you’re looking for a program that a) recognises ASCII codes input from a keyboard, and b) has a means of you telling it to do something unique with each different number it receives.  In this application the unique action would simply be to output a specific MIDI note number for every specific key number received.

Having read about Ken Rushton (‘MusicScienceGuy’)’s instrument, the Jammer, at (and his blog at, I decided to use the program MIDI Integrator he had devised to operate it.  MIDI Integrator is described, and can be downloaded here:  It’s free, and works on Macs and Windows PCs.  It’s an application based on MAX/MSP, and I had downloaded and installed the free version of this, ‘Max Runtime’ from

MIDI Integrator was designed to be used with the Jammer – and is potentially very useful because of that, as the Jammer has 2 separate MIDI keyboards; but inputs for MIDI keyboards weren’t useful for this application, as the QWERTY keyboard doesn’t output MIDI information.  However, there is another feature of the program that would do the trick.

Setting this up was quite an involved process, but consisted of just a small number of important steps.  Most importantly, when done once, most of these things didn’t have to be done again!

When you open MIDI Integrator, the interface you see looks like this:

MIDI Integrator screenshot 2

All the boxes with a ‘p’ at the beginning are boxes which can be clicked to enter information.  There are lots of boxes here, but I only needed to look at two of them.

What I had to do is ‘map’ the keys of the QWERTY keyboard – i.e. match them to MIDI notes – in the top half of the screen, and route the output to a virtual MIDI instrument in the bottom half – which involved telling the program where to send the MIDI signals to operate a synthesiser or sound module.

So, I began by clicking the box marked ‘p mapPCKeys’.

When I did this, it brought up the window for the QWERTY keyboard, which looks like this:

MIDI Int screenshot mapPCKeys

What this tells you is that the mapping instructions used by MIDI Integrator are in a text file called ‘MidInt PcKeys v1.txt’.  What I had to do is replace the text file ‘MidInt PcKeys v1.txt’ with one of my own, with my choice of notes.

To do this, I clicked on the third button ‘coll MidInt PcKeys v1.txt ‘, under the text ‘Click below to edit mapping table’.  This opened the text file, which looked (in part) like this:

MIDI Int PcKeys v1 txt

You can see from this example that each line has 5 elements, like this:

56 #8, L N 76.  This is interpreted as follows:

56 is the number (usually ASCII) sent out by the key;
#8 is which key that is (in this case, the number 8 on the second row down of the keyboard);
L is the virtual MIDI instrument in MIDI Integrator to which the information is sent (L, R, Lx and Rx are the choices MIDI Integrator gives us; it doesn’t matter which one you use if there’s only in use at any one time);
N means ‘note’ information is what’s being sent;
76 is the MIDI note number.

So all I had to do was identify the keys on the keyboard I wanted to play notes on, find out the ASCII number of each one, and write a line of code for it like the one above.

The problem is matching the right key number on the left-hand end with the right MIDI note number on the right-hand end, and the MIDI Integrator website made this easy, by providing a chart (an Excel spreadsheet), laid out in the same way as the QWERTY keyboard for me to write this information in.  It already had the key numbers in it, all I had to do was write the MIDI note numbers in next to each key.

MIDI Int Keybd Map

Not all of the keys on the keyboard have code numbers that can be used, but it’s only a small number round the edge that won’t work.  I experimented to check exactly which ones could be used, and labelled them or didn’t label them in the spreadsheet, as appropriate.

This chart shows which codes MAX/MSP allocates to which key, mostly based on their ASCII codes:

MIDI Int screenshot key nos

I made this diagram to show the 5-row note layout for the ‘Wicki-Hayden’ keyboard:

MIDI Int Wicki Layout

and this one to for the ‘Janko’ keyboard:

MIDI Int Janko layout

I figured that I might as well create several different text files for MIDI Integrator while I was at it, and try out some different layouts.

So I had the keyboard layout spreadsheet and the pattern of notes for my layout (I started with the Wicki-Hayden).  The final piece of information I needed was what MIDI note numbers corresponded to the notes in my layout diagram.

This chart shows the MIDI numbers which correspond with all the notes from C0 (the lowest) to G9 (the highest).  C4 is ‘middle C’.

MIDI Int MIDI Note No Chart

For mathematical reasons, 128 notes are available in MIDI and for mathematical reasons they are numbered 0 – 127, not 1 – 128.

Once armed with this information, I filled in the spreadsheet, went back to the text file ‘MidInt PcKeys v1.txt’, filled in my own numbers and saved it as ‘MidInt PcKeys AM Wicki.txt’ in the same folder as ‘MidInt PcKeys v1.txt’.

The file began like this:

MIDI Int screenshot wickitxt

(Large numbers like -999 are used for comments, and are not interpreted as instructions).

It doesn’t matter what order you put the notes in, so I started in the top left of my spreadsheet, so you can see the line of codes representing the F keys, followed by the line of codes representing the number keys.  The file just continues until the end of the bottom line (the one beginning ` Z X C V . . .).

There are other things you can specify in the file – for example, you can map some keys to MIDI control messages, and MIDI Integrator has some of its own commands.  I ended the list with the following:

-2 space, R C SUS;
-15 Home, R C OCT 2;
-13 Pg Up, R C OCT 1;
-14 Pg Dn, R C OCT 0;
-16 End, R C OCT -1;
-12 >, R C OCT -2;
5 Ins, R C KEY -1;
-6 Del, R C KEY 0;
-11 <, R C KEY 1;

Instead of ‘N’ for note as the 4th character, ‘C’ is for control.  MIDI Integrator’s command ‘OCT’ changes the octave of the keyed note up or down by the amount shown, and the command ‘KEY’ changes the keyed note up or down by one semitone.

Now my note instructions are permanently preserved in the text file and never have to be input again.  Whenever I open MIDI Integrator, I can click on the box ‘p mapPCKeys’,

MIDI Integrator screenshot 2

and in the box that opens up, click on ‘read’ next to the text ‘Click to change mapping table’,

MIDI Int screenshot mapPCKeys

This brings up a typical ‘Open file’ dialogue box in which I can select ‘MidInt PcKeys AM Wicki.txt’, which immediately maps all the keys of the QWERTY keyboard to the Wicki-Hayden layout (or the Janko layout if I choose ‘MidInt PcKeys AM Janko.txt’, or any other layout for which I’ve created a text file).

The layout files I’ve so far created can be downloaded here: Harmonic; Janko; Wicki-Hayden.  (Click ‘Save Link As’).

There are a few other things to do to get MIDI Integrator working – for example, the keyboard volume can be set via the slider on the right-hand side – but these are all explained in the detailed MIDI Integrator Manual, available along with the program, the spreadsheet and other things at

The main thing is to select the instrument routing.  This is done by clicking the box ‘p Instrument_L’, ‘p Instrument_R’ – or whichever one you’ve chosen on the lines of the text file – at the bottom of the main screen:

MIDI Integrator screenshot 2

The boxes on the left-hand side, especially ‘noteout l’ (or ‘noteout r’, etc., for the other instruments) are important.

MIDI Int screenshot instr box

A QWERTY keyboard isn’t quite as good as the AXiS or the Jammer, of course, but it’s a pretty cheap way of getting to know isomorphic keyboards and alternative keyboard layouts.


January 2012
« Dec   May »

Enter your email address to follow this blog and receive notifications of new posts by email.