NovaKordo

Must be time for version 3: this could be the final one, but don’t hold your breath.

The rabbit is a symbol of prosperity but if ever one was serious about becoming successful, one wouldn’t be inventing variations on the accordion. There are 3 areas that I am playing around with at the moment, although through the latter part of 2022, I became more serious about improving my actual accordion and vocal skills and now have a live set that I have been intimidating audiences with. Along with percussionist Andy Fraser, we are creating a very full sound in a very small space – you can wedge an accordion and cajon in almost anywhere!

And talking of cajons, I have a preliminary design for a tunable drum that is also based on wrist movement – put simply, one or two playing surfaces are moved out in front on wrists (as per the Novakordo) and attached to the main box via a duct. Chambers in the main box will have ports so the effective volume (and hence) pitch of the box can vary – bongos and tables rolled into one!

On the NovaKordo front, there are 3 main areas that I am working on. First, is an upgrade to the differential wrist. Whilst version 1, surprisingly worked ok, I felt that the belts (known as T2 and found widely in 3D printers – hence being cheap) would be too flimsy for playing live. With too much pressure the belts could be made to slip over the teeth of the gears.

There is a more robust belt that is common in robotics (T5), and I have considered a rebuild with these, but I think there is an easier way. With the belt system, it is possible to rotate the top part 360 degrees or more. In practice, the keyboard assembly will only turn through about 120 degrees max. This means that instead of using belts, we can use ropes. Very small ropes. And I have identified a fishing line that is incredibly strong (intended for sea monsters). Because this can be attached to fixing points on the wrist, this will eliminate any chance of slippage. Another advantage is that the parts can be thinner, rather than fatter if we used the T5 belts

Here is a prototype of the new wrist. Incidentally, I have moved from Tinkercad to Fusion 360 for the more complex designs. I still use the former for a quick and easy print as Fusion is still somewhat baffling.

You can see that cords can be attached from the horizontal wheel (on the right) to the vertical wheels (behind). These, in turn, are connected to yokes on the stepper motors (not yet attached).

Because one can create 3D changes in sound via hand movements (remember that the object of the wrists is to allow the keyboard assemblies to follow the movement of my hands via “9 dof” sensors and thus get variable timbres, modulation, and volume changes, I have been using the 3D joystick from the earlier version of the Novakordo, to play around with potential sound setups. Particularly interesting are:

• my Hammond organ experiments where drawbars can be moved dynamically together with the speed of a Lesley cabinet. Using the UVI Acoustic Samples B5
• a mix between upright and electric basses. Playing with various Ample Sound basses
• a mix between Bossendorfer, Rhodes and Wurlitzer pianos. Here I am a longtime fan of the Moddart Pianoteq plugins
• movement between the soprano, alto, tenor and bass parts of a choir – in this case the Spitfire Audio Epic Choir
• and similar to above with various sax sounds (and enhanced with a breath controller) using Audio Modelling SWAM plugin

More on this with demos in a further blog.

Finally, I want to start playing with the lights in the frame again soon. So far, I have just used a simple demo program but I need to model the lights to the shape of the frame, try out various effects, and then have a means of controlling them easily for different songs, and probably have the lights triggered by the sounds themselves. I’ve just discovered a new set of tutorials here by Scott Marley and here is a link to a set of tutorials from Dave’s Garage.

I made a major breakthrough in the early part of the year when I decided that we would become even more un-accordion like – that is a keyboard and a buttonboard (Launchpad) for each hand. In addition, and possible a world first, I proposed a scheme whereby the keyboards would track the movement of the hands. They would do this by monitoring the hand positions (using sensors attached to the palms). Robotic “wrists” would perform the actual movements to maintain the alignment.

In the previous post, I showed a simple proof-of-concept to test out various sensors and ways to be able to move the keyboards in the appropriate planes. And what I have been doing since, and no small feat, is to learn robotics, and test out a system that would be able to move the keyboards (and at nearly 2kg for the keyboard/Launchpad combo have sufficient oomph). Here are some views of my first prototype – they work but I am not convinced that they are robust enough so I am currently working on version 2.

The movement in 2 planes is achieved by a differential mechanism (see here and below) using 2 stepper motors. Whilst the system works fairly well, I don’t think that the belt drives I chose are strong enough. Because I don’t need 360 degree movement in either plane, I am working on a modified system where the belts are replaced with strong cords and fixed in place on the pulleys. All will become clear soon.

New year, new design. I’ve had quite a few breakthroughs with the Novakordo since I last blogged. First off, the new frame is now nearly complete. Think of the frame as a personal, wearable keyboard stand – the old one was made out of 38mm drainage pipe connected with 3D printed parts. The new one is lighter and uses transluscent Acrylic tubes. These were just begging to be lit up, so I am running LED strip lighting throughout. Each of the LEDs is addressable so I am keen to see how I can let the music change the colour patterns formed. In this video, I am just using a test sequence to change the lights but I think it shows what can be done.

My previous version had a piano keyboard for the right hand and a Novation Launchpad Mini for the left – the idea being to sort of mimic an accordion. For various reasons, I am now going to start experimenting with a keyboard and Launchpad for each hand. They are attached to the frame with a bar normally used for holding a pair of video monitors. This is how it looks (see top image for right hand, below for left) with the newly discovered XKEY 25 which, although it is rather clattery, is lightweight and a delight to play.

The next big innovation is to do with expression, by which I mean changing parameters such as the volume. Again this comes from my accordion-like ideas where I wanted something to simulate the bellows without having to pump air. The natural thing seemed to be with hand movements and my previous version had the Launchpad mounted on a 3 axis joystick. Moving the Launchpad around would allow bellow-like volume control and/or any other changes in expression such as filters, modulation or … anything. Now I have changed this around completely.

“Gloves” worn on each hand will contain sensors that monitor the position and orientation of the hands (these sensors, rather like the ones found in smartphones, measure position in 3 planes – XYZ) and information is fed back to a motorised platform which makes sure that the Launchpad is alway under the hand. I’ve tried hacking a commercial “pan-and-tile” mechanism intended for use with cameras, but found that it was neither fast nor accurate enough so am now working out how to build my own. Here’s a very crude prototype that is giving me a feel for what I need to make – a couple of stepor motors creating a pan and tilt, plus a couple of the “glove” sensors (one on top of the motors, and another far right), controlled by an Arduino Nano. It works, but some improvement needed – will report back in next blog.

And of course it may be that the hand movements will also have an effect on the frame lights – so many possibilities, so little time.

I’ve missed out on updating you all (well both of you) but there has been a lot happening – most importantly, starting to build Mark 2. Rather than write a lot, I thought it would be best to put it all onto video. So here we are.

Meant to say also:

• I thought the breath controller was USB Midi, but I was wrong – the USB is for power only and one needs the old 5 pin DIN connector, so it meant buying a Midi to USB device (only £12). I think, eventually, I will make my own.
• I should have said more about the elastic cord – the current joystick is held in place at 4 points: I am going to try 8 points for the new one – think it will make it smoother to move around.
• I will say more about the foot pedal in a later video when I decide on something more practical. With the foot pedal, it means a cable has to run up to the frame – which will keep me in one place. You could say the same with the “umbilical” (USB cable and power) that goes to the computer but that is easier to manage. But current best ideas are either pressure sensors in feet so achieve change by shifting weight from left to right feet. Or, possibly add a gyroscope to the box so that the angle at which the player turns provides the control function.

This week I will continue investigating the possibilities of my 3 axis controller. I wanted something simple to try out but challenging in what can be achieved. I already knew that there was a saxophone plugin that, as a “modelled” instrument relied on the expression applied to the notes. The instrument in question is from Audio Modelling and hopefully, I can give you a demo by the end of the week.

I have also purchased a breath controller, something I was keen to have as a Novakordo add-on. Will be keen to see how this works in conjunction with the joystick.

I have also ordered the parts needed to build a second joystick assembly – I want to keep the existing module working whilst I play around with a few ideas. Plus it will give me the opportunity to document everything properly. I imagine that as some stage I will have 2 instruments on the go – the one that I can play, and the one that I experiment with.