InsideMicrosoft

part of the Blog News Channel

Microsoft Research: Great Ideas, Where Are The Products?

Microsoft spends over six and a half billion dollars a year on research and development, resulting in many new technologies that go into their products, and many others that conclude, successfully, and are never used. It is at the point where, when journalists hear about something really cool being researched at Microsoft, it is automatically assumed that the technology will never be productized and made available to the public.

For example, this week Technology Review wrote about a Microsoft project working on a sort of virtual headphones technology. The system would target sound coming out of speakers at a specific spot where a person was located, adjusting the timing of sound waves coming out of each speaker in array so that sound is cancelled out in some areas of the room and amplified in others.

The idea would be an experience similar to listening to headphones privately, without wearing anything on your head. A person could even lie in bed, watching TV with the sound on, but the person lying in the bed next to them would not hear anything at all!

This virtual headphone technology is exciting, but will it ever be “productized”, that is will it ever find its way into software and hardware in the future? Anything’s possible, but doubtful. I was shocked that Windows Vista shipped with the technology to use microphone arrays embedded in a laptop or monitor to isolate sound sources, and this could join that as a companion feature in the next version of Windows, but the hardware just isn’t there to support it.

According to Richard Sprague, 25% of laptops come with array microphones built in. Frankly, that is a shocking statistic. Where are these laptops? No laptop I’ve ever owned has had a built-in microphone, let alone a fancy array mic. I’ve seen some business laptops with built-in mics, but I’m always buying consumer models that never have that (relatively cheap) feature. It’s probably Dell’s fault for being cheap that I don’t have one, but I had no idea this was a common feature.

Chris Pirillo seems just plain pissed off that stuff like this isn’t being released, at least not for the moment, and he’s uploaded this video:

That tech is talking about the culmination of six years of work, resulting in an amazing noise cancelling solution done completely in software (noise cancelling is usually a hardware problem). The tech actually says that the advantage for him demoing this at TechFest is so that product teams at Microsoft will notice his work and build it into their products. Yes, a Microsoft employee is hoping to sell his product to other teams at Microsoft!

And you wonder why some of this stuff is never productized.

Then there’s Roundtable, which is finally making its way to manufacturing, a decent success story. I’ve talked about Roundtable here before, and it’s a system where a multi-camera Ring Cam is placed in the center of a conference table and records all the participants at the table, as well as tying in external video conference participants. RoundTable is finally going to hit the market as a product you can buy, albeit for $3,000. If there’s any product I want to demo badly this year, it’s RoundTable.

Also, one strategy for getting MS Research products turned into real products seems to be turning the development team into a seperate spin-off company. This week, Microsoft announced it was spinning off one Research project as ZenZui, which will provide a cool widgets interface for mobile devices. Take a look at their cool interface here, which will make Windows Mobile users a little less jealous of the iPhone, methinks:

March 30th, 2007 Posted by | Corporate | one comment



Hosting sponsored by GoDaddy

1 Comment »

  1. A bunch of the headphone stuff you talk about above IS in Vista, in the form of the room correction wizard (available with the Microsoft HDAudio class drivers, OEMs can choose not to provide these).

    Some of the DSP didn’t make the cut for Vista – there simply wasn’t enough time to turn the research prototype into a product – there’s a long way from what you see at TechFest to a product, MSR can take shortcuts that won’t cut it in the final version.

    I also strongly suspect that MSR’s budget isn’t 6.5 billion, btw. I bet that a huge chunk of that 6.5 billion goes to pay the salaries of people like me – the folks who do the day-to-day work to build the products that people buy.

    So IMHO, the vast majority of that 6.5bbn R&D budget directly appears on the shelves.

    Comment by Larry Osterman | March 31, 2007

Leave a comment