Now that the updated Kinect for Windows SDK is available for download, Engineering Manager Peter Zatloukal and Group Program Manager Bob Heddle sat down to discuss what this significant update means to developers.
Why should developers care about this update to the Kinect for Windows Software Development Kit (SDK)?
Bob: Because they can do more stuff and then deploy that stuff on multiple operating systems!
Peter: In general, developers will like the Kinect for Windows SDK because it gives them what I believe is the best tool out there for building applications with gesture and voice.
In the SDK update, you can do more things than you could before, there’s more documentation, plus there’s a specific sample called Basic Interactions that’s a follow-on to our Human Interface Guidelines (HIG). Human Interface Guidelines are a big investment of ours, and will continue to be. First we gave businesses and developers the HIG in May, and now we have this first sample, demonstrating an implementation of the HIG. With it, the Physical Interaction Zone (PhIZ) is exposed. The PhIZ is a component that maps a motion range to the screen size, allowing users to comfortably control the cursor on the screen.
This sample is a bit hidden in the toolkit browser, but everyone should check it out. It embodies best practices that we described in the HIG and is can be re-purposed by developers easily and quickly.
Bob: First we had the HIG, now we have this first sample. And it’s only going to get better. There will be more to come in the future.
Bob: There’s no downside to upgrading, so everyone should do it today! There are no breaking changes; it’s fully compatible with previous releases of the SDK, it gives you better operating support reach, there are a lot of new features, and it supports distribution in more countries with localized setup and license agreements. And, of course, China is now part of the equation.
Peter: There are four basic reasons to use the Kinect for Windows SDK and to upgrade to the most recent version:
- More sensor data are exposed in this release.
- It’s easier to use than ever (more samples, more documentation).
- There’s more operating system and tool support (including Windows 8, virtual machine support, Microsoft Visual Studio 2012, and Microsoft .NET Framework 4.5).
- It supports distribution in more geographical locations.
What are your top three favorite features in the latest release of the SDK and why?
Peter: If I must limit myself to three, then I’d say the HIG sample (Basic Interactions) is probably my favorite new thing. Secondly, there’s so much more documentation for developers. And last but not least…infrared! I’ve been dying for infrared since the beginning. What do you expect? I’m a developer. Now I can see in the dark!
Bob: My three would be extended-range depth data, color camera settings, and Windows 8 support. Why wouldn’t you want to have the ability to develop for Windows 8? And by giving access to the depth data, we’re giving developers the ability to see beyond 4 meters. Sure, the data out at that range isn’t always pretty, but we’ve taken the guardrails off—we’re letting you go off-roading. Go for it!
New extended-range depth data now provides details beyond 4 meters. These images show the difference between depth data gathered from previous SDKs (left) versus the updated SDK (right).
Peter: Oh yeah, and regarding camera settings, in case it isn’t obvious: this is for those people who want to tune their apps specifically to known environments.
What’s it like working together?
Peter: Bob is one of the most technically capable program managers (PMs) I have had the privilege of working with.
Bob: We have worked together for so long—over a decade and in three different companies—so there is a natural trust in each other and our abilities. When you are lucky to have that, you don’t have to spend energy and time figuring out how to work together. Instead, you can focus on getting things done. This leaves us more time to really think about the customer rather than the division of labor.
Peter: My team is organized by the areas of technical affinity. I have developers focused on:
- SDK runtime
- Computer vision/machine learning
- Drivers and low-level subsystems
- Samples and tools
Bob: We have a unique approach to the way we organize our teams: I take a very scenario-driven approach, while Peter takes a technically focused approach. My team is organized into PMs who look holistically across what end users need, versus what commercial customers need, versus what developers need.
Peter: We organize this way intentionally and we believe it’s a best practice that allows us to iterate quickly and successfully!
What was the process you and your teams went through to determine what this SDK release would include, and who is this SDK for?
Bob: This SDK is for every Kinect for Windows developer and anyone who wants to develop with voice and gesture. Seriously, if you’re already using a previous version, there is really no reason not to upgrade. You might have noticed that we gave developers a first version of the SDK in February, then a significant update in May, and now this release. We have designed Kinect for Windows around rapid updates to the SDK; as we roll out new functionality, we test our backwards compatibility very thoroughly, and we ensure no breaking changes.
We are wholeheartedly dedicated to Kinect for Windows. And we’re invested in continuing to release updated iterations of the SDK rapidly for our business and developer customers. I hope the community recognizes that we’re making the SDK easier and easier to use over time and are really listening to their feedback.
Peter Zatloukal, Engineering Manager
Bob Heddle, Group Program Manager
Kinect for Windows