We are switching the status of the Believotron project from active development to pause. It is high in our hearts, but the economic realities force a refactor and restructuring of the design philosophy. And that takes time, hardware iteration, and software development.
We currently need to put the project on hold and take care of some spinning plates. We saved all our nickles and were able to fund over 2 years of full time research and development, and many many more years of part time help. That money ran out in 2017 and we’ve been scrambling since then. Side work has become full time work. Fulltime work has become secondjob and thirdjob. We’ve been doing a very poor job of community engagement. It’s time to pause the project and take care of a few things so we can get back to making great progress.
Hardware development means low volume and slow changes
Code can be deployed almost instantly. Hardware takes at least a week, if not several to fabricate a new circuit board, install all the components, and test out the design. Over the course of several years, this has put development out of reach for the short-term.
Low-cost synthesizers don’t cover the cost of R&D
Although the project is very rewarding, the sale of synthesizers and accessories does not even come close to the cost of developing prototypes. Our goal was to create one of the lowest cost, highest featured synthesizers, and the focus on a low beta price meant that we would be subsidizing the cost of hardware. It seemed fair. Hardware iterates slowly and you are taking a risk by getting something that is mid-feature.
What this approach did was slow down development. Instead of learning from our mistakes on a given revision and then changing the hardware and software, we wound up slowing down the feature change to match the group of beta users on a static piece of hardware.
Turnkey raises the price
We believe that a great design that only highly skilled technologists can build or use isn’t as good as an okay design that spreads far and wide. There is a large technology gap from buying a consumer product and soldering one from parts. We thought that the best users and best feedback would come from people who had very low experience with hardware, programming, and synthesizers. Using this design goal, we aimed for a turnkey system. Something that is incredibly low cost and powerful for a super-user, but assembled and ready to go.
There is a disconnect in this goal. There is a lot of additional overhead in some of the components. Traditionally, component costs are sold at 3X the purchase price. Things like the Axoloti core, the enclosure, and the daughter board become prohibitively expensive. Some of the designs we’ve shipped would have cost $800 if we did this.
We need to change the strategy of the design to match the low-cost goal.
Recreating the wheel is expensive
It’s a lot of fun to source your own parts and develop the whole tip-to-toenail software chain. It’s another thing to compete with every synthesizer company’s low-cost offerings. There are amazing $50 MIDI controllers that would cost $200-$400 in a low volume development environment. Recreating a MIDI piano keyboard would cost several hundred thousand dollars in plastic mold costs, not to mention the circuit boards. Imagine wanting to change anything. A lot of off the shelf MIDI controllers are better than anything open source is going to produce in the next few years. Competing with this in a development environment is a losing struggle.
We are going to do a few things in the next phase of the project. Here are our goals:
Raise the technology level expectations to someone who can through-hole solder
Possibly raise the technology level expectations to someone who can reflow surface mount components
Raise the technology level to someone who can write C/C++
Create a hardware accessory that processes physical user input and graphical user interface to offload processing from the Axoloti core
Stop selling high cost sub-assemblies and sell things like the Axoloti core as a pass-through accessory
Provide excellent assembly documentation so we don’t have to do it for you
Create prototype loaner units instead of selling a finished beta
Changing the small black and white OLED to larger, color touchscreens
Adding an offloading microprocessor, probably a Raspberry Pi
Switch from high-cost prototype circuit boards to off-the-shelf MIDI controllers
Where to go from here
If you are a beta customer and are unsatisfied, please email us. We’re slooooowly going through our emails, but we will help you out between plate spinning.
In the meantime, if you’re interested in learning when the project unpauses, please sign up for our email newsletter: http://believotron.com/find-out-more