Thanks for the posting the Deshaker setup, I'll try it out. Keep up the good work. One of the things I tried was sharpening the video. Because your videos were captured in progressive mode, sharpening helps your video quite a bit. I tried it with your 794 video. Here is the result:
Posts: 267
Joined: 4/21/2007 From: Huntertown, IN, USA Status: offline
Now that I can accurately measure height with the on board altimeter, I've been trying to see how high I can take video and stills. Here's the best so far at a peak of 270 feet:
Still using the green foam custom mount and the Panaxonic DMC-FX07 camera.
Dirk, I agree that QuickTime Pro is good if you keep everything in MOV mode. I have been using Sony's Vegas Movie Platinum 8.0 for editing and conversions.
Posts: 1511
Joined: 12/3/2005 From: Atlanta,
GA, USA Status: offline
That video was great Old Man Mike! It was really stabile too. It's good to see what almost 300 ft looks like. How high is your radio tower? Can you post a version of the video without processing? Thanks.
_____________________________
Up there, you don't have time to think, if you think, you're dead! Never mess with a pilot having more rotors than you!
Posts: 267
Joined: 4/21/2007 From: Huntertown, IN, USA Status: offline
quote:
ORIGINAL: Sky High
That video was great Old Man Mike! It was really stabile too. It's good to see what almost 300 ft looks like. How high is your radio tower? Can you post a version of the video without processing? Thanks.
Thanks! The radio tower is 90 foot. The unprocessed video file is around 500mB for a 5min flight. Here is what the video looked like before the deshake processing:
I've also include the altimeter log graph. It is amazing to me that it even shows the change in height when I picked it up for the TI arming at the beginning and when I picked it up to cut off power at the end.
< Message edited by Old Man Mike -- 7/28/2007 4:48:50 AM >
Posts: 267
Joined: 4/21/2007 From: Huntertown, IN, USA Status: offline
quote:
ORIGINAL: Sky High
I meant uprocessed for stabilization. Could you post the video unstabilized but still compressed? Thanks.
Sorry it took me a while. I had not kept that intermediate step and had to reproduce it from the original mov file. I've edited my post so that it will now show the same period of time without the deshake processing. It did not look as bad as what I thought it would without the video deshake. But I still like the deshake version better.
Posts: 1511
Joined: 12/3/2005 From: Atlanta,
GA, USA Status: offline
Yes, I agree. The picture looks alot better than the SAVS camera and it doesn't bounce! The processing makes it even better. So, is there a way to know how accurate that altimeter is? You could fly level with the top of your tower which you know is 90 ft and see if the altimeter shows 90 ft.
_____________________________
Up there, you don't have time to think, if you think, you're dead! Never mess with a pilot having more rotors than you!
Posts: 267
Joined: 4/21/2007 From: Huntertown, IN, USA Status: offline
quote:
ORIGINAL: Sky High
Yes, I agree. The picture looks alot better than the SAVS camera and it doesn't bounce! The processing makes it even better. So, is there a way to know how accurate that altimeter is? You could fly level with the top of your tower which you know is 90 ft and see if the altimeter shows 90 ft.
I've done exactly that for not only my tower but at a friend's tower which is at 120'. Both times it was within a few feet of the actual height. At the beginning of the recording, I have it set for an auto zero. At the end of a 10 min flight it is sometimes as much as 8 feet in error. With the more typical 5 min flights it is almost always within a few feet of zero. I am quite amazed that it works so well and thought I might eventually use it to implement an altitude level hold.
I tried the image measurement technique by calibrating the number of pixels for a known reference size object at a known straight line distance. But unless your camera is pointed straight down at an object of known size, calculating the straight line distance based on estimated angle makes for a very inaccurate measurement. For example, I had use the image measurement technique on that crash video I posted and thought it was close to 500' high. Based on the flights I've made with accurate measurements from the altimeter, I think the crash video was closer to 250 feet in height.
Posts: 267
Joined: 4/21/2007 From: Huntertown, IN, USA Status: offline
I've also been experimenting with different settings on the camera for stills. Here's one taken at 200 feet while flying with the TI OFF (Cloudy conditions seemed to make TI mode fly erratic):
Posts: 273
Joined: 4/19/2007 From: Eagle, ID, USA Status: offline
quote:
ORIGINAL: Old Man Mike
I've also been experimenting with different settings on the camera for stills. Here's one taken at 200 feet while flying with the TI OFF (Cloudy conditions seemed to make TI mode fly erratic):
Hi Mike, I picked up the Panasonic FX07 today. How do you trigger it for stills? Is there a way to have it shoot continuously, or do you have to use the timer (10 seconds is not enough time to spool up the T-Rex and climb into position).
Posts: 267
Joined: 4/21/2007 From: Huntertown, IN, USA Status: offline
quote:
ORIGINAL: BB_DF
Hi Mike, I picked up the Panasonic FX07 today. How do you trigger it for stills? Is there a way to have it shoot continuously, or do you have to use the timer (10 seconds is not enough time to spool up the T-Rex and climb into position).
Thanks, Bruce
After so many hours of research and development, I'm not sure that I should release the final design. Final cost might be too high for some. Anyway, here it is:
Posts: 273
Joined: 4/19/2007 From: Eagle, ID, USA Status: offline
Ah yes, simple and elegant! Great little camera - thanks for the recommendation. I just ran a test and it will shoot high-res pictures for over 12 minutes with a 2GB SD card, which was only $24 at Best Buy.
Here's my camera mount for the T-Rex. Since conventionals tilt a little in a hover due to the side thrust of the tail rotor, you need to have a complicated pendulum mount if you point the camera forwards, to keep the horizon level. I figure it's just as easy (maybe easier) to shoot from the side, and all that complexity just goes away! The camera and 1/16" aluminum angle are so light I can't tell any change in the CG when I pick it up by the head. Any side tilt due to translating tendency in a hover will just change the downward angle, and not affect the horizon (in theory). Flight tests coming soon.
BTW it's such a joy to fly the 600 without that boat anchor hanging out underneath! That might have to wait for the Maxi-Joker.
Cheers, Bruce
< Message edited by BB_DF -- 8/1/2007 12:00:25 AM >
Printable Version
All Forums >> RC Helicopters >> Draganfly Innovations Products >> RE: Another Custom Camera Mount 
RE: Another Custom Camera Mount - 
