Mr Blazey

Composer, musician, one fifth of Grey Tapes.

[mrblazey]

Short Kalimbo set from Noisefloor 2017, Staffordshire University

[mrblazey]

Live Kalimbo improvised performance at Culture Lab, 9/6/17

[mrblazey]

Kalimbo at NIME 2017 – Reflections and Feedback

At NIME 2017 I was fortunate enough to give a performance and a demo with my instrument/performance project Kalimbo. This was a fantastic opportunity to gain some really useful and insightful feedback from a diverse community of musicians/makers/engineers etc. who are all engaged in the world of new/novel instruments/interfaces.

Performance – Figura Ensemble - Royal Danish Academy of Music

http://www.dkdm.dk/Det-sker/Kalender/2017/05-2017/NIME-koncert

When I applied to perform, I was asked if I would like to do so in collaboration with Figura Ensemble percussionist Frans Hansen. I thought this would probably work well, especially with some form of tuned percussion to match the pentatonic tines of the instrument, so was happy to agree. One big benefit of this was that it inspired me to develop effects control features for a second audio input. At the moment, this includes high and low pass filters, controlled by tilting left and right (central position = no filter) and the same PD effects used for the thumb piano. These currently consist of a neutral and a pitched delay line, one reverb send for each delay and 4 cap-sense send controls, one for each delay and reverb. Overloading the capacitive sensors or tilting the instrument all the way in any direction causes various randomised parameter changes across the effects. This basically means you can easily control the sending of either input to any effect line and easily trigger changes in the effects’ characteristics, but the specifics of those changes are somewhat unpredictable. Mainstage handles the more precisely controlled effects and samplers, but for this performance I kept the second input within PD.

There is definitely a lot of potential for developing controls for external sound sources. For instance, besides playing with other musicians, this could be a nice way to introduce vocals to solo performances. However, the overlapping of some controls led to some unwanted changes in delay times etc. while performing: sometimes when trying to filter the external microphone through tilting, I would accidentally trigger a delay time change. I had tried to scale the accelerometer data so that the filters would work within a narrower range than the effects changes but there was a lot of unintentional overlap. I could either readdress the scaling or, for more deliberate control, introduce a system of buttons or capacitive sensors that would arm one set of control parameters each. That way I could safely filter the external source to either extreme without worrying about an unwanted jarring tempo change.

I had originally hoped to perform in one of the later, nightclub-based concerts and do a more beat-based, noisy set. However, the venue I ended up playing in was quite a fancy traditional concert hall and the setting was fairly formal. Frans also opted for the musical saw rather than tuned percussion as his weapon of choice so the situation ended up more suited to a restrained, tentative and explorative style of playing.

All the feedback I got about the performance was fairly positive: people enjoyed the calm, ambient approach, especially as many people had rushed over from the conference site and arrived just in time for me to play (very unfortunate for Se-Lien Chuang and Andreas Wexler who played first, as their Enhance Phenotype piece was amazing. Watch here - https://www.youtube.com/watch?v=xd2f2q6MB3Q ) . I feel pretty comfortable playing with the sounds I’ve developed so far, which were all well suited to this situation, but my main feelings after the concert were that I need to work on the controls for external input and develop more varied sets of sounds. When preparing for the performance, I realised that the PD effects side of things had always been a bit rushed, thrown together with minimal tweaks for whatever performance was upcoming. Now I have a better idea of what they do and what I would like them to do, I’ll be going back to tidy them up and fix up some odd behaviour caused by lazy patching etc.

Demo Feedback and Reflections

- It occurred to me that I don’t actually play the thumb piano tines all that much during performances, whereas most people expect that to be the main feature of the instrument. People also tend to play the tines much harder than me! This shakes them out of tune over time and makes them buzz a lot more than when I play. I have designed a better bridge to combat this, using the same tough steel bar as the current version, combined with the external wing-nuts from the previous installation version. This should provide consistent pressure on all tines, with easy adjustment from above.

- The fact that this current case can be easily opened is a definite plus, especially since several people had questions about the inner workings, as I was able to easily open it up and show them what was going on inside. However, several people found it too bulky for smaller hands. I like that the bigger scale allows for more tines but I still feel like the original sandwich tin version or the video boxes are much more ergonomic. They both had grounding issues and electromagnetic hum caused by proximity of pickup and arduino, but either one with a contact mic should work fine.

- A couple of people mentioned that they would like more control over the PD effects. Some also mentioned that they would like more control over delay/loop times. In PD you can change between a selection of set lengths for either delay, and Mainstage is currently limited to 3 pre-set BPMs. When people sent what they were playing to a delay line by using the cap-sense controls, many assumed that letting go of the control would set the delay/loop length. The variety of loop lengths/tempos has been plenty for my performances so far, but the last few have come out sounding a bit like one piece in various stages of development, rather than improvisations. Setting the loop length flexibly in the way that these people expected could help vary things quite a lot. The intended generative layers of polyrhythmic material could also be more flexible if this was the case, especially if you could add material to various loops, each set to its own tempo on the fly.

- I have realised over the various builds that the instrument really supports an exploratory style of playing rather than precise control over sound. Plenty of existing controllers offer precision though, so I’ve been happy to work with and develop the more unpredictable affordances of Kalimbo. Those who expected precision were sometimes unimpressed, but many saw where I was coming from, describing the instrument as an enchanted object and enjoying coaxing sounds out of it.

- Having seen the first 3 days of demo sessions, I felt quite unprepared for mine due to never really being involved with any proper conferences so far, let alone NIME: most demos had a poster with them, as well as business cards and/or fliers. I spent a lot of time explaining controls and intentions and justifying material choices instead of just letting people play and explore, and a poster documenting my process, findings, materials, successes, failures etc. could have saved me and the participants a lot of time. Also, most people seemed to use the posters as an indicator of whether or not to look more closely at demos, so this could have drawn in a few more people.

- One future development that I’ve had in mind since the beginning is to internalise as much of the controls as possible. Quite a few people suggested this, as well as  perhaps including speaker and amp in the body. This could lead to some interesting results along the lines of the feedback cello, with resonant frequencies feeding back through the tines, as I intended with my Multi Player Mode instrument last year. One great suggestion along these lines was to integrate the buttons and joysticks from a Playstation controller – they are already compact, ergonomic and potentially wireless. There are also plenty of cheap versions on ebay etc that I wouldn\t mind tearing apart for some experimentation.

- Influenced by Sabina Ahn’s piece, I’d like to try using the existing control data to manipulate visuals in the form of video, lights or a combination of both.

- Someone told me about the Margaret Guthman Competition at Georgia Tech. This sounds like a really interesting source of inspiration in general, but at least one modified kalimba has been entered in the past.

- Another reference someone gave me was the Brothers Baschet, who make beautiful sculptural instruments with metal tines/panels. Definitely a good aesthetic reference for upcoming work.

The lack of robust-ness, combined with conversations about NIMEcraft/digital lutherie and designing for longevity earlier in the week, made me strongly consider ways I could make Kalimbo more hardy and user-friendly straight from the box, possibly based around a Teensy so as to act as a MIDI compliant controller without needing PD. However, my supervisor pointed out that this will detract from a lot of the collage-based ideas that spawned Kalimbo in the first place, as well as negating a lot of the affordances mentioned above. We agreed that it could be good to pursue both approaches, making a commercially available, robust and simplified version that responds to the themes encountered at NIME, while continuing to develop my own personal version that facilitates or engages with live collage in various ways.

[mrblazey]

NIME 2017 - Personal Highlights

I’ve just returned from my first NIME Conference (New Interfaces for Musical Expression) and wanted to make a record of some of my favourite parts, mainly for my own future reference, but also in the hope that it might be useful to other musicians/makers/researchers.

NIMECraft Workshop – Exploring the Subtleties of Digital Lutherie

Associated poster presentation - http://homes.create.aau.dk/dano/nime17/papers/0074/index.html

Getting thrown into group-based instrument making at 8.15 on the Monday was a great way to get started and meet a good chunk of people. The workshop started with a discussion on how the finer points of ‘digital lutherie’ (Jorda, 2005 - http://mtg.upf.edu/node/449 ) can be disseminated effectively – NIME papers tend to focus on technical details and developments at the expense of all the finer details and artistic choices that go into the look and feel of a finished instrument. My group got a bit sidetracked talking about the differences between making instruments for yourself vs. for other people; for example, when making for yourself you are likely to be happier to put up with material or programming flaws because you know the causes and how to get around the problem. You are also likely to test and problem solve individual aspects of the disassembled instrument, as well as incorporate programming and building developments into a tight feedback loop with your own playing/testing. This approach can afford to be quite haphazard and based on junky/recycled aesthetic and materials, whereas making for other people involves more craft and more engineering – you need to be confident that the instrument can withstand heavy handedness and that a stranger can operate it easily without your help/presence.

After the discussion we got on with modifying prototype DMIs (based on Bela boards) in groups, using a nice big mix of materials and adhesives. One great thing about this workshop was that we all started with a pre-programmed simple instrument. This meant that less tech-savvy folk like me could focus on fun ways to physically modify and actuate the instrument, but those who wanted to could get into the programming side as well – either way, everyone ended up with a working instrument and no one was left stuck at the programming stage. Another great thing that emerged was the way that various groups chose to work together – my group split into two pairs and made one instrument per pair. Some divided the instrument into four sections and worked on one each. Another approach was to think in terms of the ‘whole animal,’ with one group requiring all members at once to play the instrument effectively.

This Digital Lutherie thread is still being followed by Bela and Queen Mary University of London so keep an eye out on their twitters etc if you would like to take part in something like this.

@qmul_mat

@belaplatform

Papers

Designing a Multi-Touch eTextile for Music Performances – Maurin Donneaud, Cedric Honnet, Paul Strohmeier.

http://homes.create.aau.dk/dano/nime17/papers/0002/index.html

One thing I really liked about these guys was the open-source mentality – there was no academic hoodwinking or holding cards close to their chest, but rather a big emphasis on this being something that you can do yourself, including links to all the resources and materials you would need to do so ( https://etextile.github.io/resistiveMatrix/ ) ..not that I experienced any of this hoodwinking at NIME – pretty much everyone was very open to chat about their work and swap info, influences and experiences.

Self-Resonating Feedback Cello: Interfacing gestural and generative processes in improvised performance – Alice Eldridge, Chris Kiefer.

http://homes.create.aau.dk/dano/nime17/papers/0005/index.html

A big intention for Kalimbo has been to streamline all the elements of my performance ecology (acoustic instrument, effects, synths, samples and controls) into one object that does not require the player to remove their hands to manipulate effects etc. The Feedback Cello is a good example of an instrument that does this very well. The inclusion of a speaker and transducers in/on the body adds another dimension of feedback, with the added possibility of manipulating the audio between pickups and speaker with analogue or audio effects, or even another musician’s setup, as they did during one of the concerts with Thor Magnusson. They have also used it as a kind of resonating effects unit for live coding sets, adding some rich physicality to a sound world that can risk being a bit too ‘in the box’.

Fragile Instruments: Constructing Destructable Musical Interfaces - Don Derek Hadad, Xiao Xiao, Tod Machover, Joseph Paradiso.

http://homes.create.aau.dk/dano/nime17/papers/0006/index.html

Some people say that laptop or electronic sets can be too clinical or stark and therefore err towards chin-stroking appreciation and away from more abandoned enjoyment. Xiao pointed out that this probably stems from the expense and associated preciousness of all the equipment involved. Guitars can be expensive but they are readily available, especially to the high profile musicians who have famously smashed them to bits. This paper demonstrates a great method for bringing some destruction and danger into electronic performances. Some people questioned the authenticity of this danger, as the bits being destroyed were basically a proxy for the actual expensive, precious bits of equipment, but I still found the sentiment inspiring. I’m hoping to start work on a performance approach wherein the building of a performance ecology is integral to the performance, in a way that provides an instant narrative as well as legibility of form for the audience. I had already envisioned the deconstruction/disassembly of the ecology as a good way to end the performance, but after this paper it seems so obvious that smashing it to pieces would be way more engaging and fun, not to mention cathartic!

Gibberwocky: New Live-Coding Instruments for Musical Performance - Charles Roberts, Graham Wakefield.

http://homes.create.aau.dk/dano/nime17/papers/0024/index.html

While I have enjoyed quite a few live coding performances by this point, I’ve always seen it as something other people do, rather than something that would benefit my own practice. One reason I have felt this way is that I know from years of experience how to get the kind of sounds I want from certain equipment and software. Some purists might not agree with the approach afforded by Gibberwocky, but during the talk I had a definite lightbulb moment of “I could use that!’ Basically it enables you to easily tie in coding instructions with existing programs like Max or Ableton. For example, I might have a synth that I like to use, knowing that automating 2 or 3 parameters will have a pleasing effect. With Gibberwocky, this could be set up in a generative/algorithmic way. I’m yet to try any live coding, so I don’t know how experienced coders would feel about this approach, but it certainly seemed like something I would like to try out.

Current Iteration of a Course on Physical Interaction Design for Music - Sasha Leitman.

http://homes.create.aau.dk/dano/nime17/papers/0025/index.html

I think this paper is well worth reading for anyone that teaches in the realm of sound art, digital instrument design or any kind of digital creative practice. These areas tend to be approached by people from very different creative backgrounds, with different intentions and most importantly completely different base levels of technical knowledge. The main thing I took from this paper as useful to my own teaching was Sasha’s approach to dealing with this – students are first given a quiz on technical knowledge, complete with answers and directions to online resources which will allow you teach yourself how to get them. This is followed up by another quiz without the answers and finally, depending on individual weak spots, further help and one-on-one tuition to level the playing field.

MM-RT: A Tabletop Musical Instrument for Musical Wonderers  - Akito von Troyer.

http://homes.create.aau.dk/dano/nime17/papers/0035/index.html

Again, connections with my own past and future work made this instrument stand out for me. Part of Kalimbo’s appeal is the less-than-deliberate control system – you may not be able to create precise rhythms for drums or melody, but you can navigate soundscapes and ’find’ beats through gestural exploration. MM-RT also promotes an exploratory approach, but using sounds from physical materials. My upcoming work with the performative ecology building is going to involve individually controlled, non-quantised motorised percussion and un-synced tape loops. Akito’s demonstrations mainly involved short rhythmic loops, but he cited john cage’s generative works as an influence, and when talking to him at his demo he did say these kind of generative polyrythms are possible with MM-RT. The legibility of form is also a big factor in how engaging this instrument is, as the audience sees you pick up various objects and materials and inevitably gets drawn into how each one is about to sound.

Design for Longevity: Ongoing Use of Instruments from NIME 2010-2014 - Fabio Morreale, Andrew McPherson.

http://homes.create.aau.dk/dano/nime17/papers/0036/index.html

I have been developing Kalimbo entirely as a tool for my own performance. However, along the way, a couple of people have asked if they could have one. This is an exciting prospect but got me thinking about what would actually be necessary to allow me to hand one off to someone else and expect it to work. This paper looked at just under 100 instruments presented at NIME, before whittling these down to a tiny handful that became commercially available, regularly used in performances and sold to the public. This, along with the discussion from the digital lutherie workshop and lots of useful feedback from my demo session, gave me the inspiration to develop the instrument into something worth selling on, as well as a pretty good set of blueprints of the requirements to make this a possibility. If you make DMIs and would like them to be successfully sold on to the public, definitely read this and learn from what has worked or failed for others.

SALTO: A System for Musical Expression in the Aerial Arts - Christiana Rose.

http://homes.create.aau.dk/dano/nime17/papers/0058/index.html

Scoring choreography of any kind to music, or writing music for any piece of choreography, is bound to include matching up sound and movement in perfect timing (not always, but often enough…). The approach in this paper allows you to use the performers’ movements, along with things like muscle strain and speed etc., to trigger and generate sounds directly. It’s almost as if the compositional structure is given to you for free, just leaving you with sound design choices. Lots of scope!

Cyther: A Human-Playable, Self-Tuning Robotic Zither - Scott Barton, Ethan Prihar, Paulo Carvalho.

http://homes.create.aau.dk/dano/nime17/papers/0061/index.html

Not too much to say about this one apart from it sounds great and works great – the robotic capabilities are very versatile and dynamically expressive, and the design elegantly places all of the robotic workings beneath the strings, leaving the playing surface completely open to a human performer. As mentioned above, my own work so far is heavily tailored to me being the performer, whereas this instrument can be played by various people to get drastically different results, as demonstrated in the Expressive Machines Musical Instruments concert on Wednesday; Ben Taylor used live coding to generate patterns that would be impossible for a human to achieve, whereas Scott Barton incorporated a lot more human interaction and extended techniques in collaboration with the robotics.

Demos/Posters

Sounding Architecture: Inter-disciplinary Studio at HKU - Álvaro Barbarosa, Thomas Tsang.

http://homes.create.aau.dk/dano/nime17/papers/0010/index.html

Beautiful, large-scale sculptural instruments based on architectural designs.

Live Coding YouTube: Organizing Streaming Media for and Audiovisual Performance - Sang Won Lee, Jungho Bang and George Essl.

http://homes.create.aau.dk/dano/nime17/papers/0049/index.html

Quite a practical, technical paper and not something I’m likely to use myself, but included in my highlights because Sang’s performance on Wednesday with multiple jabbering Donald Trumps was brilliantly terrifying.

Design Considerations for Instruments for Users with Complex Needs in SEN Settings – Asha Blatherwick, Luke Woodbury, Tom Davis.

http://homes.create.aau.dk/dano/nime17/papers/0040/index.html

Asha was in my workshop group when we discussed how instruments intended for people other than yourself need to be self-explanatory and hard-wearing, particularly in SEN settings where explaining how to use an instrument and how not to break it can be difficult.

Robotically Augmented Electric Guitar for Shared Control - Takumi Ogata, Gil Weinberg.

http://homes.create.aau.dk/dano/nime17/papers/0092/index.html

Really cool looking instrument, and again, judging from my observation of a few demo participants, really versatile from player to player. I also really liked the drum sequencer style of the control program, complete with randomiser for a nice variety of precision or chaos.

Performances

I think most of these were filmed, as was my own performance, but they aren’t yet online. I’ll add links when they are available.

Anthony T. Morasco – Listening – Composition based around a very cool homemade piano-toll/music box and soprano singer.

Hans Peter Stubbe - Spatial Piano - Improvisation where the disklavier acts like a second player, reacting to player input.

Sabina Hyoju Ahn - Breath - Amazing performance where sound, light and visuals were generated and controlled with Sabina’s breath, using DIY circuits and a lighting rig made from e-waste that looked like a mini post-apocalyptic city scape.

D. Andrew Stewart and Sang Won Lee - Disappearing: Live Writing – Stream of consciousness typing, generating sounds and seamlessly evolving into beautiful visuals and minimal beats.

Matthew Steinke – Robotic Musical Performance - All of the robotic performances were impressive but this one had the most charm in my opinion. Musically engaging, visually reminiscent of Victorian tinny automatons and with great collage-y visual and audio snippets.

(EDIT - excerpt here - https://youtu.be/f8KkhRJ2Ltc )

Sang Won Lee, Jungho Bang and George Essl – Live Coding YouTube - What’s more terrifying than Trump? Lots of Trumps. Turning them into elements of a piece of music helped take the edge off...

Jeff Snyder – Ghostline - Four performers have their movements tracked by webcam, interacting with an on-screen ‘ghostline’ to trigger sounds. Visualisations and sonifications merge and shift over time in a composition where the sights and sounds are merged perfectly.

Yemin Oh – Time Discontinuum - The sound and video of a piano performance is recorded over one minute. This material is then fragmented, repitched and replayed along with the live performer as his movements across the keyboard dictate how the original material gets regenerated.

Jonghyun Kim – Vehicle Music – Noisy, cheeky and fun. A radio-controlled car generates synth signals through its movements, parading around the performance space between occasionally smashing into cymbals, drinks and most importantly a loud, distorted guitar laying in the centre of the space. An older performance can be seen here  - https://vimeo.com/36848457    

There were a few timing clashes with my performance, sound check and demo, and I opted to prioritise late night concerts over early morning paper sessions so I did miss a handful of talks and performances. I’ve also focussed on things that related to my own work and/or inspired me the most so this list is by no means comprehensive but still, plenty of inspiration for me to be getting on with. Overall, NIME was an amazing event full of great people and I really hope I can be part of the conversation and return in the future.

[mrblazey]

Collage Music Club

I’ve started a collage music club in dialogue with my research. Kind of along the lines of secondhand sureshots / rhythm roulette etc, it will consist of various collage-based beat making challenges. I’m hoping this will demonstrate benefits of applying visual collage techniques to music making. It’s open to anyone so if you want to play, visit here -

https://www.facebook.com/groups/226917777712952/

or check out some results here -

https://soundcloud.com/collagemusicclub

[mrblazey]

Visit Original Link

Demo of a studio composition made entirely with collaged layers of Kalimbo recordings.

[mrblazey]

I’m now working on an installation using some of the idea from these recent Kalimbo projects. The criteria for the show is very open, i just knew it was likely to be based on a plinth so used that as the starting point. 

I decided to keep things simple and make the installation static/fixed down, so I will not be using an accelerometer to map midi to Mainstage sampler instruments etc., but just focussing on the capacitive-sensing aspect and staying within Pure Data for the audio processing. I noticed in the last Kalimbo build that the capacitive sensors seem to interact with and multiply one another, so this project gives me a chance to figure out some creative exploitation of this behaviour which can then be transplanted into the programming for the handheld/live versions of the instrument.

The idea for the piece is fairly simple and self-contained; the acoustic instrument, fitted with a contact mic, will stand on top of a box (possibly an old hard-body suitcase or something similar) containing an arduino, audio interface, mini Mac and speakers. It can be played by any number of people, but ideally two facing each other. The sounds played will contribute to an ambient bed generated within PD, with capacitive sensors in 4 corners affecting various aspects of the audio and sound bed. I’d like the sound bed not to be too static but also not too intrusive, ideally drawing in passers-by to interact and leave an audible trace of their playing on the ongoing bed. 

The above photos are of the physical instrument, which is what the viewing public would see and interact with. It consists of two opposing sets of tines and four capacitive-sensing bolts. I tried the wing-nut idea mentioned in a previous post - it works well in that it can be tightened from outside the box, but once the tension reaches a certain point the bolts inside rotate with the wing-nuts, so it would be good to find a way to keep these still (maybe set them within wooden recesses or use spiked washers to grip them in place).

In the pictures, both sides are tuned to the same pitches, but after some testing I felt that this didn’t really sound very good, so I’ve since rearranged the tines to different ranges on either side. This works a lot better with less clashing but I might still try some other tunings, perhaps with each side having alternating notes within one range. 

I was originally intending to paint this along the same lines as the video box Kalimbos, but there may be something in the idea of romantic duets and the Romeo y Julieta branding that could become part of the piece’s concept.

[mrblazey]

Kalimbo 2(c) - Trying to address the few problems from the last build. This came together a lot quicker as I reused the wooden panel from the last one. With the box being clear this time, measuring, lining up and cutting holes was a lot easier and I also had a much better idea of how the inside needed to be laid out etc.

I used a few layers of primer to try and avoid the flakey edges but it didn’t really work, so I had to go with the other plan of finishing as much as possible before painting, on the assumption that once it is painted it won’t be opened again. The third picture shows all the insides in place just before painting.

Lowering the bridges left me with bolts that were too long to fit inside the box. I should have sawed them down to size but since the holes were already lined up i decided to keep drilling through the back to allow the bolts out the other side. This ended up working nicely as a little stand - very useful when layers of paint and varnish were drying!

Once I’d put it all together I realised that the tension on the tines wasn’t strong enough so they were slipping about and not being very resonant. Quite a big design flaw since the inside of the box was now inaccessible, but I fixed it by gluing back an offcut from the back bridge to make it higher again. This left the tines above the pickup at the right height while pointing the back half upwards, providing plenty of tension and also looking pretty cool!

I am generally pretty happy with this one but, as always, plenty to learn for next time;

- Flakey paint is still an issue along the edges that fold. I can stick to the plan of making sure everything is in place before painting, but there are still likely to be tweaks to make, and why use an openable box if you can’t open it in the end? I had an idea to make a cassette release of Kalimbo tunes packaged inside electrified video box kalimbas, so it would be great to sort this properly if I’m going to make things that end up in other peoples’ hands. One idea I had was to put the top wooden panel on the outside, hiding the peeling edge a bit. Another idea was to cut off the lid, making it effectively a removable acoustic instrument that attaches to a resonating box containing all the electronic elements (the lid would lift off rather than hinge open).

- If the box isn’t going to be openable, it would be good to have a tightening mechanism for the bridge that can be adjusted from the outside. One simple example i’ve seen that would work well is a wooden bar tightened from above with wing-nuts. 

- There is still weird interference from the arduino to the pickup. This is especially bad when using the capacitive sense controls. This could be because of their proximity, in which case relocating the arduino board within the box would help. A small shielded box or some other form of shielding between the pickup and arduino might also solve the problem. It could also be because I didn’t connect the instrument’s metal parts to the arduino’s ground, which seemed to help with the first Kalimbo. Another approach could be to use a good contact mic. This would give a lot more percussive/body sound but might rule out the magnetic interference.

- The capacitive sensors have a cumulative effect on each other; touching all of them at once gets much higher readings all round than any single one gets. (1000-4000 as opposed to around 100!). While this is unexpected, it could be a useful and interesting control feature with some unpredictable results. It occurred to me today that replacing the cap-sensors with simple buttons would probably make for some much more reliable midi triggers, but the combined ability to work as a trigger and a more fine-tuned control through proximity is really appealing, especially as part of what I am setting out to exploit is the fact that these various uses of collage approaches in my work afford unexpected results that would not be achieved with conventional methods; a collage approach should be ‘material-led’ - rather than ‘fix’ things until they work in a familiar way (like a button for example), it’s much more interesting to ‘collect’ unexpected results like this and allow them to suggest what use they might have creatively. 

[mrblazey]

Kalimbo 2(b) - Addressing the problem of the awkwardly-shaped box from the last build, this version is in an old video case. It’s a lot more comfortable to hold and manipulate. I’ve reinforced the lid with wood from a cigar box - this is to take the pressure of the thumb piano tensioners and to make the lid more resonant than just thin plastic. The two grills on the front help the acoustic resonance nicely. 

I love the look of this one - very collage-y as well, with the use of an obsolete medium that would end up in the bin otherwise.

Some problems;

- I painted the box early in the process. As I needed to open and close it a lot to finish putting it together and make adjustments, the paint along the edges started to chip and flake away, even with a couple of layers of spray-varnish. Next time I will have to do as much of the build as possible before painting to avoid this.

- The bridges are too high for the sunken pickup. I tried mounting the pickup outside the box, bringing it closer to the tines, but this looked ugly. Next time I’ll keep the sunken pickup but shorten the bridges.

- Ground hum is still there, even without the arduino plugged in. Connecting the pickup ground to the metal parts of the instrument seems to help so I’ll have to build this in next time.

[mrblazey]

Kalimbo 2 - Since I’ve been playing with the first Kalimbo I’ve found some new ways to make use of the capacitive touch sensor on the bridge, such as;

-triggering bass notes, with the accelerometer controlling pitch and filter

-triggering hi-hats etc more precisely, with kicks and snares triggered by the acelerometer as before

The accelerometer triggering everything on a pentatonic scale sounds nice but doesn’t allow you to skip notes, just as the original drum set-up doesn’t allow you to play kick (far left) then snare (far right) without triggering all the percussion mapped in between. Using the capacitive sensor as a trigger gives you the option of more controlled/deliberate triggering of sounds, which can be then be further affected (pitch, tone, volume, effects sends etc.) by the accelerometer data.

With that in mind, I started on a new Kalimbo prototype with four extra capacitive sensors (as many as would fit on the arduino nano). The arduino code is the same as before besides the extra OSC data from the new sensors.

I tested the circuit out in a cardboard tape box - works great! Bare wire ends poking through the cardboard triggered drums very responsively. I might be able to program the sensitivity to control midi velocity to replicate touch-sensitive drum pads, although plain triggering might be all that I can get to work reliably.

The only other main problem I wanted to fix from the first Kalimbo was a weird digital ground hum that seems to come from the arduino into the guitar pickup. Having tried a few fixes I thought that this was happening because of the metal casing so have gone for a wooden box this time.

Everything worked OK, the new tines (sprung steel drain cleaning strip) have a different tone and are more resonant than the bike spokes I used before but they could probably do with being thicker for a bit more volume and playing resistance. The other major problem is the box - I personally don’t like the look of it and it’s a very uncomfortable shape to hold while playing, especially if you are trying to play with the capacitive sensors and thumb piano at once. 

[mrblazey]

Kalimbo in action - extended thumb piano/midi controller.

[mrblazey]

(Photograph - Simon Bowen)

Last week I visited #FACT in Liverpool to talk about collage-based approaches to composition and instrument making, as well as my experience of working with the Kinicho ambisonic soundsystem. Here I am talking about material-led composition with field recordings, using an upcoming #beatpicnic release as an example.

http://soundspaces.xyz/

http://www.fact.co.uk/

http://kinicho.com/

[mrblazey]

Binaural mix of a piece created for Thinking Digital conference at The Sage Gateshead. Sounds were collected in and around Gateshead and the binaural mix was completed during the conference using the #kinicho #ambisonic soundsystem. (HEADPHONES REQUIRED!)

Original mix here - https://soundcloud.com/mrblazey/0519a

More information on the amazing Kinicho soundsystem - www.kinicho.com/

[mrblazey]

#columbo #kalumbo #kalimbo #colimbo #kalimba #collage

[mrblazey]

Made for Thinking Digital conference 2016, using field recordings taken around Gateshead by Tim Shaw and John Bowers. A spatialised version was created during the conference using the Kinicho Cosmos ambisonic soundsystem (binaural mix coming).

[mrblazey]

An ongoing collection of readings relating to Paolozzi’s collage-based approach. My current PhD research aims to apply his intentions and techniques to all levels of the musical ‘performance ecology’, including composition, performance and instrument building.

[mrblazey]

Early test of a collage-built kalimba. Made from a metal sandwich tin, cupboard door handles and bicycle spokes, the instruments houses an arduino with accelerometer and capacitive sensor. Audio is transmitted through a guitar pickup, while data from the arduino controls midi synth, samples and effects through pure data and mainstage.