For those that live in San Francisco (and the surrounding area), you know what Fleet Week sounds like. For those that don’t live around here, it sounds like incredibly loud low flying fighter jets buzzing the city for a couple hours each day for almost a week. Specifically it’s usually the Blue Angels practicing for their air show. Every year, Twitter and Facebook light up with either love or hate for these incredible technological marvels of human achievement built in the pursuit of an even greater military killing force. Regardless of your politics, they’re loud.
I’ve been working at Splunk on our cloud-based product, Splunk Storm, and have been finding new and different things to log and analyze with the service we’ve built. Shortly before Fleet Week 2012, Greg Albrecht came up with the idea of monitoring the roar of jet engines. He’s got a great write up of the whole side-project: Blue Angels Flyover Detection Using Splunk.
I used a Raspberry Pi with a Zoom H2 Recorder that can be used as a USB microphone. I left it setup by the window with it streaming data to our shared free project on Splunk Storm. We wondered at first whether we’d be able to reliably detect the jets since we had no way of testing before Fleet Week. Surprised that none of my friends owned fighter jets? Yeah, me too. But as you can see from the screenshots on Greg’s page of reports generated with Splunk, they weres pretty easy to detect.
I know the holidays are almost over (we still have my birthday!) but I had to share this video of Chris Marion’s latest revision of Fire Hero 3. I’ve been chatting with Chris off an on over the last 2 years about fire art, and how Interpretive Arson/Ardent Heavy Industries has gone about building and running our stuff safely and legally at events.
He goes into great detail on how he built a lot of the system which includes six flame effects that react to guitar playing and two 100ft flame cannons for even more fire. All the plumbing and accumulators are controlled by an Arduino Mega and a RaspberryPi with a custom built server and communication protocol (written in python).
The Python host software is made up of a series of algorithms designed to perform intelligent chord-detection, as well as detect single notes and place them appropriately. The software memorizes all the chord progressions of a particular song as the guitarist plays in order to ensure consistent output for each chord.
The system is designed specifically for live music performances. I’m hoping it gets booked for a show nearby so I can check it out in person.