Leap Motion: First Impressions

My Leap Motion dev kit arrived in the week, yey! I’ve only had a couple of evenings with it but first impressions are good.

The device itself is tiny (about lighter sized), lightweight and very well made. It has a nice grippy base with the logo etched into it which stops it being dragged around by a stiff USB cable, very thoughtful!

The micro USB connector is on one end with a indicator LED on the other and that’s about it. The only other thing of note is a second set of connection pins alongside the standard USB ones, maybe indicating that the hardware is set-up with some future-proofing in mind?! The dev board I have is v.06.5.

There is loads of information on the net about the device and what it can do so I won’t bore you with a sales pitch here, let’s just dive right in…

Getting stuck in

Set-up was quick and easy. Install the SDK, plug the device in, start the app and sign-in. The first time I connected it it auto-updated the firmware, which took about 20 seconds, then all the examples with the SDK just worked. Sweet!

Most of the development I will be doing with the Leap Motion will initially be browser based (JavaScript). There are some good opportunities for packaged apps/extensions with the Leap that will be quick and easy to install and use. I put together a couple of simple demo’s using the D3.js and three.js JavaScript libraries to see what the accuracy and latency is like, and to get an idea of what it would feel like to actually use the thing!

Demo’s

You can get both the demo’s on my GitHub, take a look around and fire them up. You will need an actual device to play with them though, naturally.

I was told that too much ambient light can badly effect some of the development units so I let the light in during the three.js demo. There were a few stutters every now and then, but nothing major, and these went away when the sun was not directly on the device.

The Leap app sent up a handy flare (pop-up notification!) to tell me there was too much light at one point, which was a nice touch, but I really didn’t notice too many issues. This may be due to the fact that it is v.06.5 board, and meant to have much better handling of these conditions compared to previous versions.

It also gets quite hot during normal operation too (on par with a charging mobile), so too much direct sunlight would probably be a bad idea anyway.

In a nutshell

As you can see from the demo’s the device’s output is very accurate indeed, especially considering how simple the examples are. There is no extra smoothing or error checking in the JS but the output is still pretty much spot on.

In regards to using the Leap on a day to day basis I will have to get back to you on that one, it is such an alien way of interfacing with a computer I have not got used to it yet! Although I am pretty sure it will not be replacing my mouse and keyboard just yet, I imagine it being a great addition to my desk for more bespoke tasks when needed, such as moving windows around multiple screens or scrolling through a datasheet while making notes – or when showing off to friends and family!

A really cannot wait to start properly diving into more concepts with this fantastic little device. I have been playing around with so called “gesture interfaces” for a few years and nothing I have come up with has been anywhere near as accurate as this, and with the near-zero latency the possibilities are endless.

Watch this space…

Update: Leap have just announced the release date for the final production model – May 13th 2013 – at less than £100 (with shipping, taxes and all that malarki) it really would be rude not to!

One response to “Leap Motion: First Impressions”

  1. Wilson says:

    This is so exciting. I don’t think hand gestures are better than using a mouse or keyboard for getting work done (in the same way I find touch screens a let down). But the ability to present data back to people using gestures (like in Minority Report) is very exciting.