One of the most difficult user interfaces to get right is video editing. It is complex and fiddly with large amounts of precision required even after four or five hours of straight editing. Seeking to bring some of that interface out into the real world, [Zack Freedman] built a mechanical video editing keyboard.
The keyboard in question features popular shortcuts and keys to breeze through different parts of editing. The biggest feature is, of course, the large scrubbing knob, allowing [Zack] to fly through long video with precision. We’ve seen our fair share of mechanical keyboards that aren’t traditional keyboards on Hackaday before, such as this number pad or this macro pad.
One of the unique constraints of this project was the fact that Zack had a deadline of two days. This self-imposed deadline was to help focus the work and drive it towards completion. This meant that it had to be designed in such a way that roadblocks or troublesome features could be designed around or cut out altogether. At its heart, this project is just 14 mechanical switches, 4 potentiometers, and a Teensy to drive it all. It is the design, prototyping, and thought that went into this project that makes it noteworthy. There are plenty of lessons here about how to manage a project’s timeline and advice about how to actually finish it.
Code, STL’s, diagrams, and instructions are all on his GitHub.
Thanks [Zack] for sending this one in!
There was a popular video editing suit that had a similar hardware solution a few years back. It later went open source. I think it was named “light” something. Interesting.
Lightworks?
It never went open source, they just released a “free” version.
That’s right. Thanks for the correction.
> At its heart, this project is just 14 mechanical switches, 4 potentiometers, and a Teensy to drive it all.
There are no potentiometers, they are quadrature encoders. They are better fit for the job because of no wear. Also they can 360 deg. revolution…using pots for this kind of project would be stupid.
For the author – I really like the build!
There are continuous pots kind of like continuous servos that basically just knock out the end stops so that they can keep spinning and skip from full resistance to zero at the crossover point, or at least you can make them by chopping up a regular rotary pot. But yeah still not quite the right component for this.
Clicking through the link to his firmware on github, it’s easy to see the first non-comment line includes Encoder.h and several lines in are 4 Encoder instances named “topKnob”, “middleKnob”, “lowerKnob”, and “wheel”.
While you are technically correct that such a specific type of potentiometer exists, Ondro was correct in saying the project uses quadrature encoders, not pots.
And I also really like this project. Been contemplating doing something similar with trackballs for Final Cut Pro…
(full disclosure: I’m the author of that Encoder library, and also the guy who designed Teensy…)
Having built something similar… watch the way the jog wheel wobbles when he uses it. The shaft of a rotary encoder isn’t meant to handle the torque that the wide wheel plus a downward-pointing finger put on it. (No offense! This is an awesome weekend project!)
So to be the feature creep, I would say: support the face of the wheel from underneath with (slippery) plastic so that it doesn’t wobble so much. This is what consumer gear does. Or put it on bearings. Mmmm… sexy bearings.
Just so you know: the cheap rotary encoders use mechanical contacts, so the “no wear” thing is not quite true. They should still last longer than pots, but you still have contact bounce, and they CAN get noisy. Optical or hall effect encoders are quite a bit more expensive.
The really cheap ball mice had those. Either the tracks on the PCB discs wore or corroded, or the little metal spring brush contacts did.
Nowadays hi-end optical mice use those for scroll wheel, and its the second most common failure point after shit Chinese made (vs Omrons made in Japan you would get in early 2000 mice) microswitches.
“the cheap rotary encoders use mechanical contacts, so the “no wear” thing is not quite true”
And how! Every piece of musical gear that I have from the late-90s to mid-2000s that has an encoder has needed service. Our car has had a skippy encoder in the GPS since we bought it. Drives me nuts!
I don’t know whether I hate skippy encoders or scratchy pots worse.
I was wondering how he was going to make a shuttle wheel from a potentiometer. Quick answer for anyone who doesn’t want to watch the whole video: he didn’t. Those aren’t pots, they’re rotary encoders.
Reminds me a little of the jog shuttle on my (former) VCR. Center rotation, with an outer, centered knob. step-by-step video increment, with variable fast-forward/reverse in the outer. Wireless too.
Coincidentally there are a number of projects out there to make a jog wheel out of a VCR head.
I’m definitely impressed by the “get it done” ethic. I once needed to do some closed-captioning, and since I can’t type at the speed most people talk, I needed to be able to stop and start a video player without moving my hands from the keyboard. I could have ordered an Arduino Nano clone and put something like AVR-keyboard or Easy-AVR on it to build a keyboard with only two keys, but I didn’t have time for that. So I went down to Goodwill, bought a $5 keyboard, stopped by Radio Shack (this was a few years ago) to pick up a project box and two big pushbutton switches I thought could stand up to being stomped on, and within two hours, had my two-button keyboard. This involved figuring out which pins on the microcontroller connected to the two keys of interest, Dremeling out the microcontroller portion of the board, and wiring the two new switches to the appropriate pins, and cutting a notch in the side of the project box to put the USB cable through, and stuffing the microcontroller into the box, along with a piece of paper showing the whole keyboard matrix, in case I needed to change which key codes it generated, later. I knew when I started that the risk I was taking was that a job that was already taking too long was going to be further delayed by this attempt at accelerating it. But I got lucky, and it did the job, and paid back the total time invested in that first project I used it on.
Yeah.
No, I won’t mention any of the MANY times I thought I was doing a weekend project, but really just created a box full of uncompleted nonsense. So many boxes.
Is possible play with this on TEMPEST and mame?
i like this game but joystick is not good options