diy-e-percussion

[diy-e-percussion]

Digital Conga test with multiple sensors

Following on from my two previous tests, I've incorporated using both an analog and digital Infrared sensor to detect different hand positions for my digital conga / hand drum. 

This now enables me to trigger six different sounds from the one pad - still just a normal electronic drum pad with one piezo. I also coded it to use a sequence of notes to play a 'Tumbao' pattern, one of the most common conga patterns.

The hardest part of the coding was the mode for sending notes in a specific sequence from an array so I could play the set eight note pattern, seeing as the sensors couldn't differentiate between some of the hand techniques, and reset if I deviated from it at all.  

It's not perfect yet but I'm very happy with the progress. 

From 1:43 I demonstrate using electronic drum samples instead, triggered by the different conga techniques. 

I’m still having trouble with the Analog sensor calibration, but hoping to fix that soon!

[diy-e-percussion]

Using an analog IR sensor to get more notes from the pad!

After the success of the last test using the digital infrared sensor to get two notes with the electronic pad, I tested out using an analog infrared sensor that can return a range of values based on the distance between an object and the sensor, based on IR reflections. This enabled me to divide the pad into rough ‘trigger zones’ and get an extra note out of the pad, although it wasn't so reliable with the sensor readings so sometimes triggered the wrong sound! I think it would work better on a large pad so the sensor trigger zones can be larger and further away from each other. Latency wasn't much of an issue though which was good. 

The next step is combining this with more sensors to see if I can replicate a full range of conga sounds / emulate the normal playing techniques.

[diy-e-percussion]

Using a proximity sensor to change the note of a drum pad

Here I'm testing out using an infrared proximity sensor to change the midi note that is sent by an electronic drum pad. This is my attempt at solving the problem of wanting to trigger different sounds from the same pad for using electronic hand drums, which would normally be achieved with a different type of hitting technique. 

Velocity sensitivity isn't suitable for accurately switching between sounds, so I decided to try this method based on moving the hand further down the pad to change the note, which is an exaggerated technique of what is common to do already to achieve a different note.

THE FULL STORY

One main problem that I have been trying to solve for the electronic-bata drums, is being able to get different sounds out of the same pad. For real drums, you can get many different sounds by using different types of hits. For the bata - to properly play the rhythms, you need to be able to play two different types of tones on the larger head of two of the drums.

Previously I had tried achieving this by using velocity sensitive triggering - it’s actually the thing that first led me down the whole MIDI route! So in that scenario, if you hit the drum loudly, it would be a different note than if you hit it at a soft or medium velocity.

This was OK although pretty unreliable.

I’ve seen one guy use a foot pedal, that he would press to change the pad’s note, similar to how you can play different notes on an electronic hi-hat.

Now that I can use a microcontroller to control the electronic bata drums, I have the option of using sensors and more custom ways of triggering MIDI notes

I kept thinking if there was anyway I could change the note played by using a similar motion to existing hand drum technique - this lead me to thinking about proximity sensors. Infrared sensors are used to detect objects by emitting and reading infrared light reflections - they’re commonly used for robots to detect obstacles.

In regular conga drum technique, you do move your hand forward on the drum  a fraction to achieve a slap sound, so I figured, if the sensor can be sensitive enough, if there is one on the edge of the pad, it can detect your hand moving forwards and backwards to change the note.

This way you can play a lot more naturally, and get any velocity for either note.

I bought a cheap sensor that could detect objects between 2 and 10 cm, soldered it up to some wires and blue-tacked it on to my drum pad.

Combing some basic code for using an electronic pad with an Arduino micro-controller, and using an infrared sensor, I created this test.

Really happy with the results! Thing it’s going to be a workable solution :D Here is the code I used for anyone interested!

The pad triggering code is based on the code 'Simple Sensing' from

Ryo Kosaka's HelloDrum arduino library

and the IR sensor code based on this

IR Obstacle Detector code from Muhammad Munir

[diy-e-percussion]

Hi-Hat Controller Code

Seeing as I found it so hard to find good code for a hi-hat controller pedal online, here is the code that I came up with, from some of my own ideas and other code I found online.

I really love how willing everyone is to share their projects, code and ideas in the online DIY communities, it’s a beautiful thing! Here is the pedal in action:

and here is the code :)

[diy-e-percussion]

Using a microcontroller + learning basic electronics

Initially for my experiments, I’ve been using an existing drum brain to receive signals from the piezos or pads and send MIDI notes to the computer.

However when I was planning out the design of the MIDI Marimba, I realised that if I wanted to have a large number of keys, I’d be plugging in a huge amount of cables every time I wanted to use it - it was looking like it was going to be 16 cables for the piezos, two power cables, two USB / MIDI cables... Way too much trouble, effort, and cables that could get damaged or unplug. So I decided to look into other options. I had initially wondered about using something called a microcontroller, which you can read more about here: https://electronics.howstuffworks.com/microcontroller.htm

I had asked around on the forum of a local electronics store and described my project and what I wanted to do - and had received the answer that it would be way too complicated and costly - so I initially abandoned the idea.

Luckily I decided to look into it again and ask different people, and it turned out that it couldn’t be further from the truth!

Microcontrollers are pretty inexpensive and there is an absolute wealth of information out there on how to use them and an insane amount of cool projects that people make with them.

Here’s an overview of using Arduino based microcontrollers, mainly for creating MIDI controllers

If i used a microcontroller to build my marimba, I would connect all the piezos and everything else to the controller, and then the ONLY cable I would need to plug in to use it would be the USB cable for power and receiving the MIDI signal, OR a separate physical MIDI cable if I wanted.

This sounded like a way better option and also opened a world of possibilities - I would be able to add my own knobs, buttons, faders, pedals, sensors - anything really to the marimba. 

While it was going to be a steep learning curve without any background in electronics, I was very excited for all the possibilities and amazing things I could add to the instrument - so I went ahead and bought a cheap Arduino Uno copy board, some basic parts and started doing online tutorials.

ARDUINO TUTORIALS

I started with these tutorials by amandaghassaei and they were the perfect introduction! There are heaps of other ones on the instructables site as well as all over youtube! 

https://www.instructables.com/Beginner-Arduino/ https://www.instructables.com/Arduino-Sensors-and-MIDI/ https://www.instructables.com/Intermediate-Arduino-Inputs-and-Outputs/

Following the tutorials was great, but as soon as I started trying to modify or build my own circuits I found I really needed to learn more electronics knowledge.

LEARNING BASIC ELECTRONICS

Sparkfun was my saviour, their tutorials are really straightforward and they use great demonstrations - it seems like it’s all focused more at kids and schools but it’s really well done. I found I really needed this knowledge to make any changes to builds and figure out what parts I needed to buy!

https://learn.sparkfun.com/tutorials/what-is-a-circuit/short-and-open-circuits

https://learn.sparkfun.com/tutorials/voltage-current-resistance-and-ohms-law

I bought some extra parts to play around with, and experimented with potentiometers ( often you will you see this as knobs ), buttons, LEDs, MIDI ports and piezos. In the beginning all the parts could fit right into a breadboard, which is a board that you use for making prototype circuits. The great thing about these is that they are electronically connected in a way that you can mostly plug parts and cables right into them, without any need for soldering anything. Here’s a compilation of some of the tests I did with MIDI 

STARTING TO SOLDER

The next test I wanted to do was to connect a 1/4″ jack to connect a sustain pedal or external pads. This meant that I needed to solder, so that I could connect the jack to wires that could plug into my breadboard.

This was a great place to start as soldering is a skill that is key to building anything with microcontrollers that is more permanent and stronger than a prototype.

Again I started with sparkfun tutorials

https://learn.sparkfun.com/tutorials/how-to-solder-through-hole-soldering/all

My first solders weren't very neat, but they worked!

Building on other projects that I found online, I then created the code to use a sustain pedal - which could be used to add one to a keyboard that doesn’t have one, or as part of a project like my marimba :D After that I moved on to using the same pedal as a hi-hat controller pedal, to change the note being played by a piezo if the pedal is pressed. I was surprised at how hard it was to find similar code online, but with some trial and error I was able to combine different bits and pieces I found and come up with some bits of my own.

Here it is in action 

It was definetly a steep learning curve for all of this, but I got a lot out of it and the potential for customisation is so much greater now, my design for the marimba has really changed and I now know that I can include heaps of amazing features that I thought would be great but I dismissed as not being acheiveable. Very excited to keep learning more and experimenting!

[diy-e-percussion]

Inspiring instrument examples

Found out about these really innovative new instruments today! Super inspiring to see what people have come up with. Particularly love the radial sensitivity and pressure change of the Boppad, and the easy approach for making music and loops of the ORBA, as well as the use after-touch and the sensors to modify all the sounds!

BopPad

Artiphon ORBA

[diy-e-percussion]

Marimba Key Sensitivity Test

Testing out the sensitivity of the marimba keys when running the piezo pickups through the drum brain ( legagcy / medelli dd506 ).

This is still with the different attachment techniques I was testing out, some of which definitely aren’t working, since some of the keys are coming loose!

Gets a gauge on the sensitivity level at the moment though.

[diy-e-percussion]

Electronic Bata Find

I was listening to the soundtrack to one of my favourite 90s games, Rayman, which I realised is full of Latin percussion, and I came across these tracks which sound to me quite inspired by the bata! 

The soundtrack was mostly composed by Rémi Gazel.

[diy-e-percussion]

Samples on Freesound.org

I’ve now uploaded some of the samples I’ve been using to freesound.org - a great website for finding and sharing sounds! *Always check the Creative Commons licences though to find out what you can do with them though*

You can find my sounds here:

https://freesound.org/people/Sassaby/

[diy-e-percussion]

Found another electro-bata!

This is so far the only other example of an electronic-bata set I’ve found, and it’s from about 12 years ago! Funnily enough it’s in a marimba setup :P Good inspiration for both my projects!

[diy-e-percussion]

808 Bata Samples

I played around with some samples for the Bata today using 808 toms, snares, claps and cymbals.

From the Wikipedia on the 808:

The Roland TR-808 Rhythm Composer, commonly known as the 808, is a drum machine manufactured by the Roland Corporation between 1980 and 1983. It was one of the first drum machines to allow users to program rhythms instead of using preset patterns. Over the course of the 1980s, the 808 attracted a cult following among underground musicians for its affordability on the used market, ease of use, and idiosyncratic sounds, particularly its deep, "booming" bass drum. It became a cornerstone of the emerging electronic, dance, and hip hop genres, popularized by early hits such as "Sexual Healing" by Marvin Gaye and "Planet Rock" by Afrika Bambaataa and the Soulsonic Force.The 808 was eventually used on more hit records than any other drum machine. Its particular popularity in hip hop has made it one of the most influential inventions in popular music, comparable to the Fender Stratocaster's impact on rock. Its sounds are included with music software and modern drum machines, and it has inspired numerous clones.

I sourced these samples mostly from Roland’s SPD-SX samples, which you can find here  + a few other electronic drum samples I’ve been collecting.

Test Run

I found it quite hard to improvise a rhythm using all six heads on the fly, but here’s a little test to show off the samples.

More on the 808

[diy-e-percussion]

Electro Batá Set Test

I recorded a quick test with the electro-batá set I’ve been working on, over some live drums I recorded. It’s all pretty rough playing wise but it does the job - the main thing I wanted to check is whether I could get the latency down enough to track parts with the Batá in time over a recording - and I’m pretty happy with the results!

I’ve been imagining this 6/8 pattern called Ñongo over a heavy backbeat groove, so I decided to try it out - it was a pretty hard one to play at the speed I choose so it was a challenge but a lot of fun as well!

For those playing at home, the drum track was required with just one mic, using the Zoom H2 in front of the kit, sitting on some boxes facing in towards the snare / hi-hat. The Bata was recording MIDI only, which then translates to samples.

Latency

So I was very happy with the latency - I played the part live over the drum track, using my speakers to hear the recording and the batá (an advantage of MIDI-recording since there’s no actual sound from the room) with a metronome, and it felt perfectly in time for me apart from my actual playing mistakes and my own time drifting.

The setup I was using was the ASIO4ALL driver, with a Sample Rate of 96kHz, and 64 samples, which Reaper estimated to have a delay of between 5.6-7.6ms. It wasn’t enough for me to notice!  All of this was still running through my legacy dd506 drum brain.

I don’t fully understand why the higher sample rate settings etc. made it better, but I had read to try this in several articles on reducing latency, and it certainly was much better than when I was trying at 44100 or 48kHz.

Set Design

In terms of the actual play-ability of the set, it was not bad. I could reach all of the drum heads which was pretty important since the pattern was so fast, it was at a good height that I wasn’t leaning down overly, and the width seemed to be good - shorter than what an acoustic set would be but quite comfortable. My main issues were still hitting the plastic rims of the drum pads, since they weren’t designed for playing with hands, and that the pads can sometimes move if they become loose on their mounting rods.

I could fix both of these issues by making my own pads, with no rims and a different mounting system, which is something I’m considering, but I’ll explore after some more testing.

I was also able to track the cowbell and shekere parts for the end of the video quite nicely, just setting up another track and assigning those samples to the right pads.

Sound

I was pretty happy with the sound - I think it blends well with acoustic drum sounds and sounds good, although there weren’t many dynamics because of how fast and roughly I was playing. It would be good to try a slower pace and try out more dynamics and some of the tapado notes.

Here’s a test just showing the range of samples I’m using - after doing this, I think I’m going to have to adjust some of the sensitivities.

Let me know what you thought!

[diy-e-percussion]

Different Attachment Methods

After my sound tests with the marimba prototype, I was trying to find ways to reduce vibrations from each key travelling to the other keys, after getting some unwanted triggering where the pickups on the surrounding keys would sound when I was hitting the desired key.

The advice from the hive mind was that the screws I was using to connect the keys and rubber to the frame, were likely to be the biggest culprits, transferring the vibrations from the hit down into the frame and subsequently up into the other keys. 

A few alternative methods were suggested - using glue to attach the pieces to the frame and key, using Hook & Loop picture hanging hooks ( like a very strong velcro ) or some combination. OR using a piece of rubber tubing to shield the screw so that it wasn’t actually touching the key and shock absorber. The other angle to work on was the suggestion that using a softer material for shock absorbing than the ugg boot rubber I was using would be more effective - the more the key could actually move and vibrate, the less vibration would be transferred to the other parts. I tried out several of these methods and took some slow motion videos without plugging the marimba in, just to see the different responses to pressure, mallet hits and how much rebound ( or bounce ) I could get for each of the different attachment methods. From LEFT to RIGHT I’m using: A) Very soft, springy foam, glued to the frame, with they key attached with HOOK and LOOP picture hanging tabs

B) Screw going all the way, shielded by rubber tubing on top of the frame, all on ugg boot rubber

C) This is very similar but  I made this one slightly differently and the rubber actually goes down into the frame a bit as well. This seems to work much better, there is a lot of room for the key to move up and down, but it is still firmly attached. 

The key is only touching the rubber tubing at first, which in turn is supported by the ugg boot rubber. The ugg boot rubber is also slightly elevated from the frame, it’s not entirely touching it on the bottom when the key is not depressed. The key has a lot of movement around the screw for this one.

D) Uggboot rubber glued to the frame, with the key attached with HOOK and LOOP picture hanging tabs

Pressure Test

Really interesting to see the different ways the keys move. A moves far too much, B not enough. C and D are the best - although D had a tendency for the key to detach entirely from the rubber 

Hitting Test

Hitting A feels very strange because of how much it moves... B, C and D all feel pretty normal although D feels the best because it wobbles the least.  

Rebound Test

Big difference here - D has easily the most rebound because it is the firmest. A has almost no rebound because the key moves so much, absorbing all the force. B and C feel alright although not as good as D.

RESULTS - ( not plugged in )

The springy foam (A) is not good for both the feel of hitting it and the rebound. The D setup felt the best although came off entirely during all my tests which is definitely not usable. I’ll have to play around with this further because the A key also used the hook and loop tabs and never came loose.

Keys B and C probably worked the best, so now we’ll have to see how they perform plugged in in terms of actually reducing triggering to the peizos!

PLUGGED IN

[diy-e-percussion]

Latin American Marimbas

Doing a bit more research about Latin American marimbas and marimba music!

https://www.youtube.com/watch?v=2hlCPu7imyw  

https://youtu.be/ogsxFt4YKzY 

And a more pop twist...

[diy-e-percussion]

Researching Virtual Marimba Instruments

I did some research into commercial Virtual Instruments available for Marimbas and the like, to see what kind of sounds and settings were available - lots of interesting finds! All of these would be compatible with the marimba I’m building, but I’m also planning to make my own Virtual Instruments from samples i source or record. Here are some different examples!

They couldn’t have gone for a more dramatic voice-over...

http://www.soniccouture.com/en/products/26-percussion/g43-grand-marimba/

http://www.soniccouture.com/en/products/26-percussion/g46-vibraphone/

http://www.soniccouture.com/en/products/18-exclusive-free-content/p201-wassolou-balafon/

http://www.soniccouture.com/en/products/26-percussion/g25-morpheus/

STONE MARIMBA - this thing blew my mind!

[diy-e-percussion]

Very cool performance using a Marimba and a Xylosynth! 

[diy-e-percussion]

Sample Testing - Marimba Prototype

I did another test run with the MIDI Marimba prototype - this time with some foam in between the piezo pickups and the marimba keys - it performed a lot better and I barely noticed any accidental key triggering at all!

In this video I’m mainly demonstrating using it to trigger different types of samples to show the possibilities of the instrument and see how playing it works.

I try out some marimba samples, a vibes, a zither as well as some drum kit and Afro-Latin percussion instruments and some bells - it’s really very flexible!

The sound you were hearing is straight out of the marimba, recorded from the stereo mix on my computer.

Some of these samples are single velocity layered samples and some are multi-velocity layered samples - something I go more into back in this post.

The basics explanation though is that you can either have samples that get louder and softer according to how hard you hit, but is the same sound file being played - OR you can have it setup so that different sound files are played for different hit strengths - this allows you to add lots of detail to the virtual instrument, as all physical instruments respond this way - their sound changes a lot according to how you are playing them, not just getting softer and louder.