This week we were tasked with fleshing out the less technical aspects of our final projects, answering the following questions:
What will it do?
Who's done what beforehand?
What will you design?
What materials and components will be used?
Where will they come from?
How much will they cost?
What parts and systems will be made?
What processes will be used?
What questions need to be answered?
How will it be evaluated?
What will it do?
Very simply, my final project will be a musical synthesizer, incorporating some common practices of these machines (sliders, potentiometers), with some less commonly seen aspects (acoustic elements, capacitive touch plates).
It will allow musicians (I don't particularly like the term user, as it has some negative connotations) to use this instrument as a tool in order to create music, in a hopefully more organic way than is common for synthesizers.
Who's done what beforehand?
I will not go too far into details about the history of sound synthesis, although it is a fascinating topic, but let me resume a little bit of what I know.
A lot of this information comes from my various experiences as well as Laurent de Wilde's fantastic book Les Fous du Son.
In the 1920s, at similar times, Soviet inventor Leon Theremin invented the Theremin, which is often credited as the first modern electronic instrument. In fact, a few preceded him, as early as the 1890s with the dynamophone by Thaddeus Cahill, but for our purposes this is not very important information.
The theremin was rudimentary and did not allow for very dynamic playing, and was hard to control, but it was extraordinary in that it provided a completely different way to control sound, through disturbing electromagnetic fields created by two antennas with the hands.
Some similarly incredible instruments in terms of musician control are the Buchla Music Easel in that it used capacitive touch instead of a traditional full keyboard, the Expressive E Osmose (to catapult us back into contemporary times) in that it uses a traditional looking keyboard with each key being independently actionnable, twistable, pushable to create possibly the most expressive keyboard instrument to date.
We could also add the Computone Lyricon to this list, due to its use of breath as the primary method of expression.
There are many more synthesizers which deserve to be mentioned, but I do not really have the space here! Maybe one day I will make a full report...
I do not mention the plethora of modular and semi-modular synthesizers not requiring user input at all and playing generatively, or through patch point connections, as while these require a skilled musician to operate, they generally do not take into account the human interaction.
My synthesizer tries to take into account all this, while being (for the moment) extremely simple to operate and make music with.
You can find various inspirations on this moodboard I made for my project.
What will you design?
So far in my prototyping, I've designed everything about the project, from the signal flow :
To the box (which is very much a work in progress):
The side panels:
To the different electronic boards:
What materials and components will be used?
For the boards:
Double sided copper plates x4
SAMD21 x2 2,79€
10 pin headers x7
4 pin headers x3
Mini USB connectors x2 0,76€
Power jack x1 0,57€
Voltage regulators 3.3V 1A x2 0,35€
Sliders 10k ohm x6 8,01€
Potentiometer 10k ohm x1 2,42€
TRS stereo jack x1 0,75€
LED Green x4 0,21€
Capacitors:
1uF x6 0,1€
Resistors:
5K x2 0,09€
1K x2 0,09€
100 x2 0,09€
220 x2 0,09€
Schottky diode x1 0,12€
Cables:
2x 10pin female to female cables
4x female to female single cables
For the box:
4mm plywood, laser cut
4mm acrylic, clear, laser cut
3D printed side panels, PLA
3D printed PCB holders, PLA
Sandpaper
Where will they come from? How much will they cost?
All components except the sliders and potentiometer were already in the lab.
The sliders were bought by 10, and cost 80€ (8€ per unit for very high-quality units).
The potentiometer was part of a kit I had purchased, so the individual cost is hard to establish.
I've put links to most components in the BOM above, with some things like cables and the materials (plywood, PLA, acrylic and sandpaper) left for makers to acquire themselves.
I imagine most of the electronic components are made in Taiwan or China, and were purchased through Digikey or Mouser.
What parts and systems will be made?
All parts of the syntheiszer will be made, as explained above. If I choose to add parts such as an audio mixer, a speaker, or others, I will design and make the amplifying circuits for these.
I will also make potentiometer and slider covers out of 3D printed and molded parts.
What processes will be used?
Processes used:
3D modelling using Rhino, Grasshopper
2D modelling using Rhino, Inkscape
2D and 3D milling using Roland SRM-20
Laser-cutting using Trotec Speedy 400, Epilog laser cutter
3D printing using Prusa MK2, Bambu Labs A1
Molding using clear epoxy
Programming using Arduino IDE
Electronics design using Kicad
What questions need to be answered?
As far as questions, the main one is: "How far along can I go? Can I add many parts that work well together? Can I get as far as adding analog electronics? Acoustic parts?"
I suppose I could also ask: "How usable will the system really be for music-making?"
How will it be evaluated?
I suppose the synthesizer will be evaluated first for its design and its use of different processes and second for its actual use as a music instrument.
This could be somewhat of a shame as I have seen many ugly instruments produce fantastic music in the right hands, and I am not very design-focused (although I am hoping to change that!)
Final Project
Synth Project Development
After much deliberation and working on a phone design followed by a musical accompaniment tool, I finally settled for making a synthesizer. Not just any old synth, and not a purely digital synth either, but rather something that combines different parts, digital synthesis and control (with capacitive touch plates rather than keys), analog sound processing and mixing (with an analog filter and mixer unit, as well as a microphone and preamp), as well as some acoustic parts (a string module or perhaps a reverberating plate).
By combining different elements from these audio fields, I hoped to create something new and different.
The initial idea was to create a double-sided instrument, allowing multiple musicians to play together and collaborate on music-making, while facing each other.
I evolved that idea into making the entire thing one-sided as I realised a double-sided instrument restricted the usage to at least 2 musicians at all times, while a single sided instrument could be played by 1 or more people.
In order to think a bit more about it, I made a signal flow diagram which was meant to show the way all the parts would fit together into a whole.
After this, I began thinking about the way the front panel would look, and sketched out my idea.
This was done quickly and changed a few times (as you can see from the eraser marks on the page...)
This reflection led to the first working prototype, using FM synthesis running through the Mozzi library.
After creating the initial prototype above, I went back to the drawing board to make a mood board.
There were a few issues with my prototype:
1. No markings or indications of any kind to show the function of the sliders and knob
2. No filtering or effects or modulators (outside the built-in modulation of FM synthesis), leading to a rather barebones experience
3. No function yet for the two pad-attached sliders
4. No real design language outside of the lasercut wood
5. No analog or acoustic functions as initially imagined
6. The positioning made it quite uncomfortable to use for prolonged periods of time
7. The internal electronics were not attached and were dangling around inside the box
Other Final project Idea: Final Project Initial Draft Week 1: Simple Phone
This first week we were tasked with coming up with an initial idea for a final project.
I looked first towards projects like the FairPhone and the Light Phone.
The FairPhone is an easily-repairable Android smartphone trying to use as many recycled and fair materials as possible.
The Light Phone is an Android-based "dumb-phone" with an e-ink screen that allows you to access simple features such as calling, messaging, a calculator, calendar synced with google, and a few others.
Both are very interesting projects but I felt that the FairPhone was too traditional in using Android and the Light Phone did not go far enough in open-source development and providing value for the users.
While researching, I came across Paxo, a fully open-source smartphone using a custom-OS.
Wanting to use some ideas from each of these projects, I thought of a simple phone to give people real access to their tech, instead of being locked behind the black box of glass, plastic and metal that we are now confronted with.
I thought of a square 1 to 1 aspect ratio that differentiates it fromm the more traditional 16 by 9 and 18 by 9 aspect ratios of current smartphones, and makes it seem more like a tool than a content-consumption machine.
With an e-ink display and simple calling and messaging capabilities, it would provide all the basic necessities of a phone, without the many distractions and attention-stealing of a traditional light emitting display.
The Simple Phone also would have some modular components, much like Arduino Shield modules.
Some Ideas for Going Further
These modular kits come in the form of DIY kits with build instructions and explanations as to their functions.
Each kit would have its own app on the phone to control it. My first thought was a simple analog synthesizer, due to my background in music. Tony gave me the idea of being able to route phone calls' outgoing voice through the synthesizer for some vocoder-style effects.
After further reflection, this led to a camera kit for taking photos, which could either be a DSLR-style camera with integrated viewfinder or a simpler digital camera (although there would be the question of how to correctly display this on an e-ink screen?)
These kits could also include a DAC (digital to analog converter) for playing high-quality audio files, or a simple recording kit with an omni-directional microphone and higher quality speaker.
Not drawn in the image above could be a solar panel and extra battery kit for charging the phone and other devices while outdoors, or even some mechanical devices such as a clock.
Another Final Project Idea
While discussing with some of my classmates, I thought it would be a good idea to simplify and refine my project a little more.
I wanted to keep the square form factor and the e-ink touch screen, but also wanted to apply the project more towards my passion and field of expertise, music.
The Simple Phone became the "Simple Music Companion" (name to be confirmed at a later date...), a standalone basic accompaniment tool for musicians, which uses a microphone with pitch detection to pick up the basic chord structure of someone's playing, and accompanies the musician with some simple piano, with different stylistic and rhythmic options to be selected on the screen. It then is able to save the MIDI track of what it has performed and that can be used as the basis of a track in a DAW software later.
With this idea in mind, I headed for week 2 and the CAD/3D modelling software portion of Fab Academy.
Final Project Modelling Week 2: Simple Music Companion
You can find the 3D model here.
Here is what it looks like from a back angle showing the kickstand and some of the features.
I started by designing the 2D cube, wanting to keep my design simple, and immediately ran into the very simple issue of not knowing how to round cube edges automatically...
I made a relatively complex design, as shown below, using circles to round out the corners, not great...
I then trimmed the points of the rectangle using the circles, leaving me with the following rounded square.
After playing around with Rhino for a bit of time, I figured out I could build a rectangle with the Rounded attribute and select how rounded I wanted my corners! Much better.
I then offset the resulting square to get the inner dimensions of the screen, and extruded both to get two objects, the screen and the outer shell of the project.
Backplate Render Issues
I made the outer shell protrude away from the screen in both directions, and then closed the back of it with another object, the backplate.
It took me some time to get the backplate perfect, with a lot of different curves and modifications, so by the time I was happy it looked like the image below.
Because of all the different intersecting curves, this was causing a rendering issue where some shadows appeared seemingly out of nowhere on the backplate.
Looking at the Rhino documentation, I was able to fix the issue with the "MergeAllCoplanarFaces" command, which then resulted in a much less crowded backplate.
Kickstand
As you can see in the image above, I had been trying to work on a shape for a kickstand, and I wanted it to have a graceful curve.
Using some ideas from the videos referenced above, I tried to edit the surface control points to make a curved kickstand, resulting in the following:
Putting some time into it made me think a curved kickstand was neither natural nor a good idea, so I simply made a straight kickstand, with a little cylinder attached to make it look rotative, as below.
Adding the Ports, Speaker, Microphone, and Volume Button
I decided that for a realistic looking model, I needed to add a few ports (USB type C and 3.5mm audio output), a speaker and microphone, and a volume rocker button.
I found the USB type C dimensions from Mouser Electronics and replicated that into my project.
The speakers and microphone were simple radial arays that I cut away from the main shape, leaving a hole for the speaker and making the microphone a different material (metal).
Here are some artistic renditions of the two.
As for the volume rocker, I wanted it to curve along with the device's curve, so I created the basic shape and then projected it onto the outer shell of my device, before extruding the shape.
And here is the render of the end result!
With a little video to go along with it
Final Project Schedule and Planning
For the scheduling, I asked ChatGPT to write up a schedule for a synthesizer project, as I will be working on a modular synthesizer project.
Week 1: Electronics Design and Initial Planning
Day 1-2: Research electronic components needed for the synthesizer project.
Day 3-4: Design the electronic circuitry and layout using appropriate software.
Day 5: Review and finalize the electronics design, ensuring compatibility with other components.
Week 2-3: Prototyping
Week 2:
Day 1-2: Start 3D printing prototype components.
Day 3-4: Laser cut any necessary parts for the prototype.
Day 5-6: Begin CNC milling for parts that cannot be 3D printed.
Week 3:
Day 1-2: Assemble initial prototype, integrating electronic components, and test basic functionality.
Day 3-4: Identify any issues with the prototype and make necessary design revisions.
Day 5-6: Continue refining and testing the prototype.
Week 4-5: Building and Testing
Week 4:
Day 1-2: Begin building the final synthesizer based on the refined prototype design.
Day 3-4: Test each component of the synthesizer individually.
Day 5: Assemble the synthesizer and conduct overall functionality tests.
Week 5:
Day 1-2: Conduct thorough testing of the synthesizer, including electronics, sound output, and user interface.
Day 3-4: Identify any issues or performance improvements needed and implement necessary changes.
Day 5: Finalize the build and ensure all components are functioning correctly.
Week 6: Documentation and Final Touches
Day 1-2: Document the entire process including design decisions, prototyping challenges, and testing results.
Day 3-4: Create user manuals and assembly instructions for the synthesizer.
Day 5: Finalize any remaining adjustments to the synthesizer based on testing feedback.
This schedule ensures that electronics design and testing are incorporated at the beginning of the project, allowing for seamless integration with the mechanical and software components during the prototyping phase.
I would have liked to also incorporate an analog filter, and am thinking of using this basic equaliser from Sonelec Musique.
The end result of my final project, La Chose FM synthesizer, is explained in the slide and video below!
To the box (which is very much a work in progress):
The side panels:
To the different electronic boards:
The initial idea was to create a double-sided instrument, allowing multiple musicians to play together and collaborate on music-making, while facing each other.
I evolved that idea into making the entire thing one-sided as I realised a double-sided instrument restricted the usage to at least 2 musicians at all times, while a single sided instrument could be played by 1 or more people.
In order to think a bit more about it, I made a signal flow diagram which was meant to show the way all the parts would fit together into a whole.
This was done quickly and changed a few times (as you can see from the eraser marks on the page...)
This reflection led to the first working prototype, using FM synthesis running through the Mozzi library.
After creating the initial prototype above, I went back to the drawing board to make a mood board.
There were a few issues with my prototype:
1. No markings or indications of any kind to show the function of the sliders and knob
2. No filtering or effects or modulators (outside the built-in modulation of FM synthesis), leading to a rather barebones experience
3. No function yet for the two pad-attached sliders
4. No real design language outside of the lasercut wood
5. No analog or acoustic functions as initially imagined
6. The positioning made it quite uncomfortable to use for prolonged periods of time
7. The internal electronics were not attached and were dangling around inside the box
This first week we were tasked with coming up with an initial idea for a final project.
I looked first towards projects like the FairPhone and the Light Phone.
The FairPhone is an easily-repairable Android smartphone trying to use as many recycled and fair materials as possible.
The Light Phone is an Android-based "dumb-phone" with an e-ink screen that allows you to access simple features such as calling, messaging, a calculator, calendar synced with google, and a few others.
Both are very interesting projects but I felt that the FairPhone was too traditional in using Android and the Light Phone did not go far enough in open-source development and providing value for the users.
While researching, I came across Paxo, a fully open-source smartphone using a custom-OS.
Wanting to use some ideas from each of these projects, I thought of a simple phone to give people real access to their tech, instead of being locked behind the black box of glass, plastic and metal that we are now confronted with.
I thought of a square 1 to 1 aspect ratio that differentiates it fromm the more traditional 16 by 9 and 18 by 9 aspect ratios of current smartphones, and makes it seem more like a tool than a content-consumption machine.
With an e-ink display and simple calling and messaging capabilities, it would provide all the basic necessities of a phone, without the many distractions and attention-stealing of a traditional light emitting display.
The Simple Phone also would have some modular components, much like Arduino Shield modules.
I started by designing the 2D cube, wanting to keep my design simple, and immediately ran into the very simple issue of not knowing how to round cube edges automatically...
I made a relatively complex design, as shown below, using circles to round out the corners, not great...
I then trimmed the points of the rectangle using the circles, leaving me with the following rounded square.
After playing around with Rhino for a bit of time, I figured out I could build a rectangle with the Rounded attribute and select how rounded I wanted my corners! Much better.
I then offset the resulting square to get the inner dimensions of the screen, and extruded both to get two objects, the screen and the outer shell of the project.
Because of all the different intersecting curves, this was causing a rendering issue where some shadows appeared seemingly out of nowhere on the backplate.
Looking at the Rhino documentation, I was able to fix the issue with the "MergeAllCoplanarFaces" command, which then resulted in a much less crowded backplate.
Putting some time into it made me think a curved kickstand was neither natural nor a good idea, so I simply made a straight kickstand, with a little cylinder attached to make it look rotative, as below.
The speakers and microphone were simple radial arays that I cut away from the main shape, leaving a hole for the speaker and making the microphone a different material (metal).
Here are some artistic renditions of the two.
As for the volume rocker, I wanted it to curve along with the device's curve, so I created the basic shape and then projected it onto the outer shell of my device, before extruding the shape.
And here is the render of the end result!
With a little video to go along with it
