PDA

View Full Version : The Best Video, Captured By Any Copter



zenny
07-02-2011, 04:18 PM
After checking so many videos, captured by *-copters, I found the following the best. Check yourself. But that is not my video. I just found it on the net. Enjoy!


http://vimeo.com/18873955

What do you think?

Honk
07-02-2011, 05:10 PM
Truly amazing! Amazing flight skills, amazing place...!

OldGeek
07-03-2011, 04:45 AM
Have to admit that is some night flying and awesome backdrops :>
Thanks for sharing!

pradiptya
07-03-2011, 05:04 AM
this guy, awesome http://shrediquette.blogspot.com/

zenny
07-03-2011, 08:57 AM
William a.k.a shrediquette wrote to me last May that "It was possible to make the Cameroon movie because I have 18 years of experience flying helicopters and four years of experience with copter electronics and software." ;-)

Pretty an extensive experience, huh!! :-)

chris1seto
07-03-2011, 06:54 PM
Very, very nice. Someone's a fan of Moby's music ;)

q6600
07-17-2011, 12:48 AM
...

http://www.youtube.com/watch?v=hU0_tGHlR7g

ridgebackred
07-17-2011, 05:18 PM
Fortunately both William and David both show how they get such amazing results.

http://www.rcexplorer.se/

David's wood tricopter with the KapteinKUK flight controller show just how much can be done without an $8000 super AP setup.

zenny
07-17-2011, 06:55 PM
Fortunately both William and David both show how they get such amazing results.

http://www.rcexplorer.se/

David's wood tricopter with the KapteinKUK flight controller show just how much can be done without an $8000 super AP setup.

Yes, that is amazing to see, but most of the multirotor developers present as if "multirotor-development and flight" is a "privy of a few", and newcomers are strongly discouraged in many a ways.

But this forum, I found as an exception. Thanks to Mikro and all the surfers here! :-)

Nikotine
07-17-2011, 09:32 PM
Damn, I have still a long way to go until I reach that kind of stability... Is it possible at all with the Aeroquad software - Arduino combination?

ziojos
07-18-2011, 12:04 AM
Damn, I have still a long way to go until I reach that kind of stability... Is it possible at all with the Aeroquad software - Arduino combination?

Hi Nikotine ..... smoking kills you ;)

The KK controller is only a gyro based with relatively high main loop speed , A kk Quad fly acro mode like . Every well designed Quad (AQ,MK,XPirates etc) can fly like this

When I say well designed I mean Size, Motor Choice,propellers, balancing, Weight distribution, Air Flow, Tuning, Noise reduction, Vibration Damping, etc. BTW The current release of AQ is a bit (not too much :)) penalized in Acro by its relatively low rate loop.
I think in Acro is the pilot to make the difference and is also a fact that good pilots prefer KK and MultiWii may be bacause other quad are well tuned on stable mode.

But the real challange is building a quad handy and easy to fly in Stable Mode that not need high skills.

Ciao

Josè

aadamson
07-18-2011, 12:29 AM
Hi Nikotine ..... smoking kills you ;)

The KK controller is only a gyro based with relatively high main loop speed , A kk Quad fly acro mode like . Every well designed Quad (AQ,MK,XPirates etc) can fly like this

When I say well designed I mean Size, Motor Choice,propellers, balancing, Weight distribution, Air Flow, Tuning, Noise reduction, Vibration Damping, etc. BTW The current release of AQ is a bit (not too much :)) penalized in Acro by its relatively low rate loop.
I think in Acro is the pilot to make the difference and is also a fact that good pilots prefer KK and MultiWii may be bacause other quad are well tuned on stable mode.

But the real challange is building a quad handy and easy to fly in Stable Mode that not need high skills.

Ciao

Josè

Actually, that used to be the case... KK now can be had in STM32 flavor and all the trimming.s

http://www.kkmulticopter.kr/

Alan

ziojos
07-18-2011, 01:04 AM
Actually, that used to be the case... KK now can be had in STM32 flavor and all the trimming.s

http://www.kkmulticopter.kr/

Alan

good,

everything evolve :), 299 $ and you can add the Navyboard on KK solution and the STM32 is coming. Many solutions are moving on faster processor.In Italy RedFox (He wrote mixer table for AQ) flies Arducopter on STM32 since last May With GPS and Altitude Hold and with effective camera stabilizer.

BTW my idea is that faster is better (obvious) but from the old "Arduino" we can have more than we have now :), however,even Arduino seems moving to faster processor ;)

Ciao

ziisco
07-18-2011, 05:13 PM
wow this 2 videos are just amazing! I really need to get my FPV stuff up and running!!! that's just fantastic.

About stability, I never had an AQ which was not that stable. In rate mode you normally don't have any problems at all, but my stable mode is also werking perfectly (besides the leans issue of course :D). A friend told me that stable (auto level) is bad for videos anyway because each time you center sticks it tilts very fast to level and that just looks unnatural und ugly in the video. so what I've learned is to not stop forward flight at all and then your videos will look natural...

ziojos
07-18-2011, 06:03 PM
wow this 2 videos are just amazing! I really need to get my FPV stuff up and running!!! that's just fantastic.

About stability, I never had an AQ which was not that stable. In rate mode you normally don't have any problems at all, but my stable mode is also werking perfectly (besides the leans issue of course :D). A friend told me that stable (auto level) is bad for videos anyway because each time you center sticks it tilts very fast to level and that just looks unnatural und ugly in the video. so what I've learned is to not stop forward flight at all and then your videos will look natural...

Hi, I agree, a well tuned Stable Mode is good to take good video, and is easier ;). Good pilots do better in Acro. Acro may be sometime smoother. I think the big advantage of quads on the Heli is the Control Routine :) and the ability to have helps from it.

What's the leans issue ?

Ciao

ziisco
07-18-2011, 06:06 PM
you don't know a about the "leans"? you're lucky =) accel calibration (or something like that) issue in 2.4 AQ. the flightangle() gives a level offset after hard turns or long forward flight and stuff like that. means your quad starts to tilt to one side after a few mins flight time (in my case). that's the one bug / issue AQ has atm that stops us from going ahead with other cools stuff :(

ziojos
07-18-2011, 06:32 PM
you don't know a about the "leans"? you're lucky =) accel calibration (or something like that) issue in 2.4 AQ. the flightangle() gives a level offset after hard turns or long forward flight and stuff like that. means your quad starts to tilt to one side after a few mins flight time (in my case). that's the one bug / issue AQ has atm that stops us from going ahead with other cools stuff :(

I never experienced this issue :) , my nephew (the pilot) don't fly hard like loop, tonneu, flip, but after a full battery the quad fly just like at the takeoff.

In the past I observed issue with overtuned DCM (mainly in the integral term) that gave problem on YAW control that lost the sync but actualy I haven't

BTW my Stable Mode Fligth Control Routine is totaly different from standard AQ.
Ciao

ziisco
07-18-2011, 10:53 PM
i know, you have your own AQ code, working stable mode, altitude hold without baro and GPS hold :D amazing! and we can just wait for the leans bug to be fixed :P

aadamson
07-19-2011, 03:06 AM
If you are just flying in stable mode and not prolonged attitudes that create any lateral movement while the quad is level, the you will not see the *leans*... you may have vibration induced attitude issues, but while related, those are not the same as the leans issue. The leans only shows itself from prolonged aggressive flight in attitude mode or prolonged turned while the quad is level. If you are running DCM and you have the low gains, the *leans* are always waiting for you, if you are just hovering or just in attitude/stable mode with no aggressive flight *and* you have a smooth mostly vibration free quad, you won't exhibit the issue.

*all platforms* that use DCM (or M/ARG - yes it effect it too) have this issue, if they run low gains to mask it *and* they don't use any form of accel weighting on how the Accel is used in the *leveling* of the AHRS.

Alan

ziojos
07-19-2011, 09:38 AM
If you are just flying in stable mode and not prolonged attitudes that create any lateral movement while the quad is level, the you will not see the *leans*... you may have vibration induced attitude issues, but while related, those are not the same as the leans issue. The leans only shows itself from prolonged aggressive flight in attitude mode or prolonged turned while the quad is level. If you are running DCM and you have the low gains, the *leans* are always waiting for you, if you are just hovering or just in attitude/stable mode with no aggressive flight *and* you have a smooth mostly vibration free quad, you won't exhibit the issue.

*all platforms* that use DCM (or M/ARG - yes it effect it too) have this issue, if they run low gains to mask it *and* they don't use any form of accel weighting on how the Accel is used in the *leveling* of the AHRS.

Alan

Alan,

I said that I never flight hard but that I use the quad normaly, not only Hovering, however in general the stable mode by definition is a bit limited in hard hanouver.
After last zisco posts I tried to simulate leans simulating by hands Huge, multiple and long attitude then leaving the quad leveled on the floor and no Artificial Horizont, Commands, Sensors were affected.
Now I'm working with quite high DCM Kp=0.05 and Ki=0.000003125 to gain better control and smoothness in altitude control but usualy I had standard values like Kp=0.02 and Ki=0.000002;

As you know from many of my past posts I use:

unbiased Gyro to compensate Gyro Drift
accel_weight to cancell unwanted acceleration effect when Acc is different from 1 g
JJ-Stable Mode is COG compensated (this more to fight wind gusts)

I'm happy the most of Unbiased Gyro that give me many advantages (In Acro and in Stable) . If I remember this was inserted in the SW and after erased. May be this help to correct "leans".

More, in many test I observed that the YAW and PITCH and ROLL DCM tuning are mutual influenced.

Last but not least the Quad version of DCM (I think also ARG/MARG) is not Coriolis Compensated

BTW I will test more specific flight but seem that al is OK

Ciao

aadamson
07-19-2011, 01:38 PM
With your accel weighting that most likely will help. Hard to judge your gains as I don't think you are in SI units while AQ is. Gyro has little to do with this, it's all from the noise on the accels, the low gains, and lateral movement with no bank.

A good way to test for it is to take your quad, leave it level and rapidly move it back and forth in one dimension and then watch the attitude indicator, if it shows a bank or pitch, it has the *leans*, if when you stop, it takes a few minutes/seconds for it it dissipate (duration will be gain dependent), that is a best case duration of time for the leans, any vibration in the quad will extend that time

ziojos
07-19-2011, 03:26 PM
With your accel weighting that most likely will help. Hard to judge your gains as I don't think you are in SI units while AQ is. Gyro has little to do with this, it's all from the noise on the accels, the low gains, and lateral movement with no bank.

A good way to test for it is to take your quad, leave it level and rapidly move it back and forth in one dimension and then watch the attitude indicator, if it shows a bank or pitch, it has the *leans*, if when you stop, it takes a few minutes/seconds for it it dissipate (duration will be gain dependent), that is a best case duration of time for the leans, any vibration in the quad will extend that time

Alan,
To have an idea my actual gains are 2.5 times the default Arducopter DCM (No SI), I'm sure that Accel Weghting is one of the key of JJ-copter stability, I adopted this feature just to better handle quick change of direction (FYI :) I posted this since last year on the forum and someone fired my Ideas). It is natural a bit of accumulation when rapidly move it back and forth in one dimension but my quad dissipate this in less than half a second ( I retested after your post, however this is one of my standard flight control quality test, I do this with many other tests after every important change). To solve other questions related to altitude stability I observed that DCM is prone to over-evaluate angle on quick attitude changes and this is quite impossible to avoid but also this is compensated very quickly.
My conclusion is that a bit of drift is physiologic on accelerations quick and/or huge, but my conclusion is that seem that I control this quite well.
When I did the test that you suggested after more than 20 strong shake JJ-copter accumulate 12-15 degree but when I stop it instantly autolevel (less than half sec)I never seeen accumulation on less than 4-5 strokes may be becose it is dissipated very fast. In real life I'm sure that never happen that my quad has the opportunity of accumulate .
On vibration I've balanced all very accurately but this is a must to do a good quad ;) not only for the "leans"

Hope this help to solve "leans".

Ciao