Keyboard Builders' Digest / Advent Calendar
Designing Keycaps for Fun and Profit
A quick overview of Matt3o’s 10-years experience in the mechanical keyboard industry before delving into future projects including a collaboration with GMK.
Published December 6, 2022
It was around summer of 2011 that I started developing a wrist and thumb pain caused by the long coding hours. At the time I was using a low profile Apple keyboard and little I knew about the mechanical keyboards world.
What I knew was that if I wanted to keep doing my job as a developer I had to solve my posture problem before the pain turned chronic, so I started lurking around the rather small at the time MK community. The best stuff came from South Korea, they were years ahead of us, already designing customs and organizing the first group buys. Eventually I ended up buying a tenkeyless Filco Majestouch with Cherry MX red. Investing over $150 in a keyboard was unthinkable to me…
How it all started
I very quickly realized that the Filco wasn’t doing much for my wrist pain even though it helped: I needed to design a keyboard that eased my day work. I’ve always been a tinkerer and I like to DIY my way through problems without having a real qualification in any specific field, a kind of professional generalist.
So in 2012 I joined Geekhack and Deskthority and I started building my first keyboard. It was a kinda 65% but the problem was finding compatible keycaps. While today 65% are the norm, back then even finding a 1.75U shift was an endeavor, let alone a 1.5U Backspace.
If I wanted a custom keyboard I needed also custom keycaps and that’s how DSA Retro was born.
Apart from some South Korean group buys I believe DSA Retro was the first widely available DSA set ever made and it sold what the historians describe as a “fuckton” considering the incredibly smaller market of 10 years ago. The DSA profile was selected because of its versatility, keys are all on the same row so it’s easy to mix and match to cover basically any custom keyboard.
I had the layout and the keycaps, I just needed the keyboard. There was very little documentation so it took me quite some time but ultimately I had my first keyboard hand-wired: it was the glorious Brownfox.
The next logical step was to make it available to the community, so I started talking with Drop first (at the time Massdrop) and Input:Club later and we partnered to make 65% keyboards widely available to the western market.
The collaboration finally resulted in the lovely Whitefox! The first iteration of the keyboard (when it was still a customizable kit) is probably one of the projects I’m most fond of. I know it’s pretty “raw” by today’s standards but it was meant as a very inexpensive kit with interchangeable plates so we had to take some shortcuts with the design.
That was the beginning of my new career. Up to that point it was still a hobby but it was taking more and more of my spare and work time.
Many more projects followed, notably DSA Granite (a less-niche DSA set), but what really changed my life was MT3.
MT3: high profile galore
Around 2014 Topre released The RealForce HiPro, and I knew something was missing in my life. It’s true that SA was already kicking but I’ve always seen that profile as a hack and anyway Signature Plastics production time was already 8+ months.
It took me a while but in 2015 I started working on a new High Profile Spherical Top keycap directly inspired by the best keyboards ever made; of course I’m talking about IBM beamspring. The market was growing fast but we were nowhere near the critical mass we are seeing today. Manufacturing a completely new profile was no easy task, Chinese factories were not very experienced in the process and the road was bumpy. In 2018, after three years from the initial concept and with the benevolent sponsorship of Drop, we finally released MT3! I have a full story about it if you are interested.
The goal was to create a high profile key that had some kind of sense in the modern mechanical keyboard scenario. SA has a flat R3, while MT3 has a row-3 that is (mostly) flat on a 5-9 degree angle, which is a sweet spot for keyboards today. The rest of the rows are designed around R3. That being said, it’s still a pretty tall profile and you gotta like that. :)
Fortunately many shared the love for MT3 and at that point I left the software development world and keyboards and keycaps became my full time job.
The future is MTNU
Don’t get me wrong, I love MT3 but it was done mostly for the look. I wanted to push those nostalgia buttons and I tried my best to make it as ergonomic as a high-profile could be but I understand that MT3 is not for everybody.
I wanted a spherical top keycap that could be a valid alternative to a cylindrical. What would a “cherry profile” look like if it were spherical? That’s how MTNU (pronounced MTNew) began.
I gathered quite some feedback from MT3 users over the years and the main concerns are: 1) too tall, 2) edges too sharp. So I started from there.
The hitting area has to be very comfortable and should not cause any harm to your feeble fingers. Back to the drawing board I ended up with a soft spherical scoop with a round top edge profile like so:
The finger happily snugs into the keycap and doesn’t find any hard edge on the front and back, much like on a cylindrical keycap.
At the beginning I had no idea who was going to manufacture MTNU and it was only because of an unpredictable series of fortunate events that I ended up collaborating with GMK.
It turned out that GMK was developing a similar profile and seemed tragically unfruitful to compete, so we joined forces. I’ll be honest with ya: It wasn’t an easy decision to make. GMK reputation was slowly declining mostly because of the very long production times and delays, but I always appreciated their work and after many meetings and emails I ultimately agreed to collaborate and honestly I’m so glad I did because working with them is fantastic and a true honor.
I worked with many Chinese, Taiwanese and American manufacturers but nowhere I’ve seen the level of professionalism that I experienced at GMK. It’s true that they went through a series of hiccups due to various “perfect storm”-like events but they are getting back on track and their new production line is ready to deal with the increased demand.
It’s thanks to the collaboration with GMK that MTNU went from a sweet dream to a working product. It took just a few weeks to get the first double-shot prototypes and I was blown away.
When you go from seeing your creation on screen to holding it in your hands, it’s always a tear inducing moment for me. The texture is just absolutely perfect. As much as I’d love a super smooth surface, a very slight texture helps hide injection lines, so we went for the lightest texture possible that still ensures the best visual result.
Since MTNU is a no compromise profile it will be full blown doubleshot PBT. No wearing ABS, no compromising PBT-blend, just plain old durable 100% PBT. This might scare some of the designers because color matching will be more complicated, but c’mon guys, you can do it. :)
So what’s the deal with MTNU?
The profile is meant as a medium height, general purpose, very pleasing to type on alternative to cherry. The spherical scoop is inviting but not too deep to cause a sharp edge and the curved top outline softens the hitting area even more. Your fingers don’t hit on anything when floating over the keyboard; that makes MTNU a kind of “advanced” profile as your only reference when touch typing are the homing keys (F and J). The cherry profile, for example, has a very strong step from R1 to R2 and that makes it easier to locate the rows.
MTNU – despide being a modern profile – still borrows the look of vintage keycaps and for that reason we decided to go with a “retro” font. The typeface is clearly inspired by old IBM and Radioshack keyboards but it has been designed from scratch to be a bit less… archaic.
At first sight it might look like a DIN font, but then you start noticing a few defining characteristics, like the barred 0 (zero) or the old school A.
At the time of this writing we are producing the second round of prototypes. I don’t expect any surprise as the first batch was already great. If all goes according to plan I should have them in a couple of weeks and so final production should start in January. I believe from there it’s another 4 months, so crossing all fingers we should have our first set around summer 2023.
I could go on forever talking about MTNU but I have more things to cover so if you want to know more I invite you to follow my blog or join my discord.
Penguin Belly Slide
I feel that uniform profiles have been neglected, but I always appreciated the custom-keyboard friendliness of profiles like DSA even though —I’ll admit— it’s not the best keycap shape in the world.
We need to fix that and here comes Penguin Belly Slide (PBS, working title).
PBS is a weird cookie. It has a spherical scoop but then the top is cut by a cylinder. So the front/back edges are round while the left/right are straight like in cherry profile, but still the hitting area is deep and round like a spherical keycap.
The result is a top surface that looks shaped over the belly of a penguin, hence the name “Penguin Belly”. “Slide” comes from... Well, because typing on them is just fun!
As mentioned the purpose of PBS is to make custom keyboards and weird layouts easy to cover and for the same reason it’s also very important for me to make this profile as inexpensive as possible. PBS doesn’t have rows, all 1U keys are the same and that helps keep the cost down but to further limit expenses I’m thinking of going dye-sublimation. Although not set in stone yet (double-shot can still happen) a dye-sub version would allow a simpler production process and simpler generally means cheaper.
I’ve already sent a few 3D printed prototypes around and the feedback has been very good so far.
Designing a “flat” profile might seem trivial but it is actually taking more time than I anticipated. With sculpted keycaps it’s the key angle that helps with ergonomics but when you don’t have that all you can do is play with the flat key shape. That’s why I’m running so many prototypes for PBS, far more than I usually do. If you are interested in testing the profile please join my Discord server and check the #PBS channel. I’m always looking forward to receiving feedback but bear in mind that so far I have only 3D printed samples.
Now the bad news. PBS doesn’t have a manufacturer yet. I have to sponsor myself the tooling and it’s not a small sum to put together and with all the projects that I have cooking It's not easy to find a spot for the Penguin in my timeline. I understand it’s a niche keycap but I do hope we can start the pre-order phase in 2023 and if all goes according to plan the uniform revolution will begin by the end of next year.
Open Programmatic Keycap (OPK)
In the many years I’ve been designing keycaps I’ve always found the 3D modeling process a bit tedious. I’m a coder at heart so when I find myself doing the same thing twice I ask myself if there’s a better way and if the process can be streamlined. It’s true that you generally work with parametric CAD like Solidwork or Fusion360 but still the process is never 100% automated.
Wouldn’t it be great if you could just set a few parameters and the software created all the keys you needed, including ISO enter, stepped capslock, spacebars,...?
That’s why I started working at OPK.
OPK is an open source library to generate spherical top keycaps written in python and powered by CadQuery. It’s not the most stable option (OpenSCAD is definitely more mature) but it lets you export in STEP format which is exactly what I needed.
The library is still in its infancy but it can already generate a wide array of keycaps of any size and height including carved or embossed legends. I believe it’s a great tool for DIYers who need to 3D print a custom set and for 3D artists who can export a whole set ready for rendering.
OPK can also do legends but of course we can’t double-shoot at home (well yeah technically we can but it’s complicated and time consuming). With an FDM printer you could try multi color printing but the end result is not great, for higher quality we need to go resin. I made a video recently where I show that decent results can be achieved very easily using resin or putty infill; it won’t be perfect but good enough and it’s a lot of fun anyway.
If you like the idea please check out OPK and send feedback, if you know CadQuery maybe you can also help with the development.
2022 has been a busy year, we are finally leaving the pandemic behind (even though don’t forget that China still has a zero-covid policy and that impacts production) and the keyboard community literally exploded. Building customs is extremely easy compared to a few years ago, keycaps are often in stock and you don’t need to join a years-long group buy (if you don’t want to), there are so many switches that we ran out of stem colors and names (Holy Pony Giraffe Lubed Ghetto Violet Retooled anyone?).
Things are overall better, the market has matured and reached a more casual audience and that caused a kind of short circuit. 10 years ago a group buy that sold 100 kits was considered a huge success, today the figures are 10-fold and nobody is willing to add that weird kit that sells only 10 copies because over 1000 it’s nothing, just wasted resources. So on one side we have a lot of options, better prices, wider audience, resources, youtubers, ... ; on the other we lost that camaraderie and pioneerism that usually distinguish the early adopters. The market is steered by the trend of the moment more than the passion of true passionate users.
The mechanical keyboard phenomenon inflated too much and too fast. The question is: is this a bubble? Will it explode or normalize? And at what stage of the curve are we? Are we rising or descending? These are all questions I asked to many people in the business and nobody really knows. Personally I believe we are still riding the high tide but to me it doesn’t matter that much.
When I tell people that mechanical keyboards are my day-job they can hardly believe me, and honestly it’s kind of a dream for me too. It’s not that I’m better or smarter than others in the business, my only merit is being… persistent. I have been here since the beginning and the flame is still burning. I don’t care if mechanical keyboards won’t be trendy anymore in a few years, I will still be here, this is just what I like to do.
|Fav. switch||anything that clicks or clacks|
|Other hobbies||3D printing, retro computing, scifi, painting, kitbashing, prop making… sorry guys, I’m a real nerd|
|Links||Blog: https://matt3o.com, Video: https://www.youtube.com/@matt3o/videos, Discord: https://discord.gg/xBVuZnjDM2|
Published on Tue 6th Dec 2022. Featured in KBD #106.
Did you like reading this post?