Comment by Greg Fletcher on August 24, 2012 at 7:09pm

How did you do that? With a video editor and synced up the vid with the MP displays?

Comment by Rana on August 24, 2012 at 7:54pm

Nice ! You seems to have been flying at an altitude, not more than 50 meters but in the overlay video, it was showing much much higher. Was it set to altitude at sea level ?

Pls. share your video tool info and process.

Comment by Ger Harris on August 24, 2012 at 9:00pm

I guess I did miss sharing the video process:

1. The original footage was taken from a 808 #16 HD camera mounted on the canopy of the Bixler. 
2. The telemetry (top right) screen was recorded directly from my Mission Planner - "Flight Data" ground station with native options. (Right-click in artificial horizon)
3. After the flight the two movies were combined picture-in-picture mode with a Mac and iMovie.  
4. The flight log was replayed in Mission Planner - "Flight Planner" with appropriate geo-fence and way-points enabled, and recorded with Camstudio.  (free)
5. Unfortunately Camstudio didn't record in a codec iMovie could understand so I had to convert using WinFF (free)
6. Flight Planner screen (top right) recombined with main movie created in step #3.

The process was a bit inelegant but I learned a lot about video and the APM2 in the meantime.  The most difficult step was synchronizing all the events together as some video speed had to be adjusted to correspond with the actual length of the video.   (When is a second not really a second?) 

Rana: Thank you!

My default altitudes were set for 50m with my ground altitude around 381m above sea level.  All waypoints were set for "Hold default Alt."  While I am getting fairly comfortable with "trusting" the APM2, I'm slowly letting it get higher and further away from me in auto mode.  My next mission tomorrow may be pushing some of my line-of-sight limitations.

Comment by Rana on August 25, 2012 at 12:52am

Excellent Harris !

Comment by steve F11music.com on August 25, 2012 at 7:30am

What are your main PID settings?

Moderator Comment by Carl Campbell on August 25, 2012 at 11:35am

That is slick.  Good job 

Comment by Matthew Schroyer on August 25, 2012 at 2:39pm

Very well done!

I considered a similar process once with a setup that used a GoPro Hero2 and a small, lower-res camera similar to the 808, but it didn't have the luxury of telemetry or tlogs, so the log files couldn't be replayed and screen-grabbed from the Mission Planner like you've done here. There's programs and websites that could reproduce some of the OSD data through the gpx file, but the results aren't nearly as impressive. Still hoping someone could find the time to code something to simulate all the perks of the tlog replay, but with using the .log files instead. Unless I'm missing something, that is.

Comment by Rana on August 25, 2012 at 4:21pm

Is windows version of iMovie available ?

Comment by Ger Harris on August 25, 2012 at 7:37pm

F11music:

I haven't found the need to modify too much beyond the stock values -

• PTCH2SRV_P = 1.2 (Essential otherwise Pitch is too soft)
• RLL2SRV_P = 0.8 (Essential otherwise Roll is too soft)
• AHRS_YAW_P = 0.2 (Bixler PARAM file suggestion)
• HDNG2RLL_D = 0.02 (Not Sure)
• HDNG2RLL_I = 0.1 (Not Sure)
• AHRS_YAW_P = 0.2
• Future Change: Reduce Circle Radius?

Bixler, Sky Surfer, and EasyStar (Ailerons) really have it easy in the 3DRobotics PID category as so much seems to be based on those airframes.

Matthew:

Thank you!  I'm really just "standing on the shoulders of giants" with this video.   All I really did was combine multiple screens that I thought would be relevant for my own analysis.   I only thing I brought to this table was a smattering knowledge of video editing and scaling.

Rana:

Unfortunately iMovie is a Mac only program.   However, if you have a friend who has one I would highly suggest borrowing it as Apple does a good job making the whole picture-in-picture as simple as it could possible be.

Comment by Ger Harris on August 26, 2012 at 7:04pm

@Rana

Nice - What program did you use?   One of the hardest things in PIP video is correlating multiple events from different perspective.  In general, I recommend some sort of timestamps to ease the process.  Even if you match the timestamps you may find that oddly one video may go faster than another due to encoding or such.

Comment

• ‹ Previous
• 1
• 2
• Next ›
• Page