Radian stabilizers
- Thread starter 3dheliguy
- Start date
ZAxis
Member
We've just added a Radian to our PH AV130 gimble and experience so far has been pretty positive. We're controlling roll and tilt only. The roll unit is fixed to the front of the roll cage and the tilt is on the camera platform. Compared to MK gimble roll stabilisation it is just so much better, we've just been flying in 15+ mph winds and the horizon was locked. The tilt took a bit of tuning but now works well. Just make sure your camera is fitted so as to keep the gimble nicely balanced.
The changeover from the MK proportional control to the Radian slew control of the tilt is a bit of a culture change but easily adapted to.
No going back to MK !
andy
The changeover from the MK proportional control to the Radian slew control of the tilt is a bit of a culture change but easily adapted to.
No going back to MK !
andy
seattle_helo
Member
ZAxis
Member
Quick example filmed this last weekend. 20 mph winds, look at the trees !
ADX3H / MK / AV130 / Radian / Sony CX730
Recorded 50p 1080, uploaded as 25p 720.
No post anything. A little jitter when descending.
andy
Last edited by a moderator:
Tabb Firchau
Member
Looking good Andy!
DucktileMedia
Drone Enthusiast
That video is amazing for 20mph. Thats a Y6 setup right? I have to say though, MK has always been king in wind and all these videos using the Sony CX cameras are already going to look much better. I think gimbal tests from here out should not use a stabilized camera to show real raw results one might expect when using a dslr or non stabilized camera.
Agreed - it should be a law
- and compulsory also to fly at a beach looking out to sea so we have a nice flat horizon and can see if it drifts on the roll axis
ZAxis
Member
Iris ... yes Y6. It was flown unaided by GPS and just about at its limit. The CX730 was bought to overcome the rattles inherent in any servo driven gimble, the only drawback was that it does nothing to help the roll axis. The MK just failed miserably to keep the horizon level once the craft was moving so the Radian looked like the way to go. It definitely was.
I always look at the overall system rather than fretting about perfection in individual items. The final outcome is always the sum of the parts, remember that the CineFlex is a system within a system within a system not a single one.
andy
I always look at the overall system rather than fretting about perfection in individual items. The final outcome is always the sum of the parts, remember that the CineFlex is a system within a system within a system not a single one.
andy
Bartman
Welcome to MultiRotorForums.com!!
there's a setting with MK that reduces its tendency to roll the camera off-horizon when flying sideways. making the roll control less dependent on accelerometer values is the trick, not perfect but much better.
http://www.multirotorforums.com/showthread.php?4709-MK-gimbal-control-horizon-roll-tweak
http://www.multirotorforums.com/showthread.php?4709-MK-gimbal-control-horizon-roll-tweak
Agreed, Andy, that we can't expect Cineflex level of behaviour. The root cause of the horizon issue is the level of MEMS sensors used in these stabilzation systems - they are, by nature, affected by linear acceleration. The Cineflexes of this world use a different type (and price) of gyro that doesn't suffer that problem. But it's the Zenmuse that has me believing there is a solution at this level. Expensive though the Zenmuse is, it certainly isn't using the big-boy gyros, so they have found a mathematics-based solution.
This is the justifications, to my mind, for a dedicated gimbal controller - the fact that you can run completely different settings for the camera versus the aircraft. But the residual roll you refered to in that thread is what I was referring to above - the MEMS gyro is itself sensitive to lateral acceleration. All the "sensor fusion experts" I've spoken to about this tell me it's a "difficult" issue - but none of them has yet come out and said it's an "impossible" one, so there's hope yet!
ZAxis
Member
A major difference between the Zenmuse and other stabilisers seems to be that it performs a more complicated startup routine which I'd guess maps the movement of the gimble+camera as a whole. This makes predicting operating behaviour simpler and perhaps compensates for the use of cheaper sensors. Everyone else seems to search for level and do no more. I wonder if there's scope for a deeper mathematical pre mapped approach for these guys, one that they haven't tried ( or been too afraid to ) yet. The Radian with its three box approach should seem more attractive as a starting point than a single integrated unit. All we need now is a mathematician ....
ZAxis
Member
We did try a bit of tweaking but the roll never came good. I'm with jes1111 over this, we can't mix gimble and aircraft control in a single unit. They are basically two independent systems and need controlling as such.
andy
I'm sceptical about the Zenmuse's start-up dance - Denny thinks it is an auto-tuning routine but I don't buy it: if so, how come all the "you can't use anything but this camera with this lens" nonsense?
I'm not expecting Tabb to give away any trade secrets about the Radian, but I'm hoping he will be his usual open and straightforward self about this thorny problem
I'm not expecting Tabb to give away any trade secrets about the Radian, but I'm hoping he will be his usual open and straightforward self about this thorny problem
anaka
Member
simpy impressive...have just a question, that sony cam is famous to be extremely stabilized...was the optical stabilizer on or off?
don't want to be plemic but just want to be sure to understand well the skills of Radian
one more time great stuff!
Last edited by a moderator:
Bartman
Welcome to MultiRotorForums.com!!
I'm sceptical about the Zenmuse's start-up dance - Denny thinks it is an auto-tuning routine but I don't buy it: if so, how come all the "you can't use anything but this camera with this lens" nonsense?
I'm not expecting Tabb to give away any trade secrets about the Radian, but I'm hoping he will be his usual open and straightforward self about this thorny problem
i hope so too but he's gone and started a forum of his own. not sure why he'd do that but openness hasn't been a strong point of vendor forums, hence the existence of multirotorforums.com so we'll see.