Hi, hopefully somebody could give me some tips on this one: I have been trying to set up my CS8 with a 3 Axis Gimbal and Radians for the last couple of months now. There are two problems I can't figure out. 1. I have an oscillating tilt movement, that shows jello in the footage. if I turn the image stabilizer of my lens on, it compensates for the movement, and the footage is fine. But it doesn't solve the problem for me. where does the oscillations come from? If I fly with tilt radian off, it's fine - no jitter. The most likely source should be the belt acting as a spring and oscillating with the weight of the camera. And it does look like it, because it occurs with a certain slew speed, both when controlling the tilt angle and when the radian stabilization hits the same speed. But the belt is as tight as I can get it. This problem has been there from the beginning and it doesn't improve, if I use my small 50mm lens. It shows less jello, when I dial down the tilt radian gain to about 50, but still showing some, and that's really low. The output monitor in the radian software is showing the jittering, too. But with no video Tx or something connected. just plain 3x radian chain with Graupner Setup. Power source for radian is coming from gr-16 receiver, 5V. Receiver is powered by 3s lipo and castle BEC. I have swapped both servo and radian with the roll axis, but problem remained on the tilt. This here seems to show the same problem: http://forum.freeflysystems.com/index.php?threads/tilt-jitters.940/#post-11935 but it wasn't solved, I guess. Here is a video of it. It shows as a jello effect. Almost no wind. I am just drifting the CS carefully 2. The good old roll twitch: I also have a play in the change of roll directions which is quite severe. I read a lot about roll twitches as a common problem, so it can't hurt to compare.. The belt is tight. Maybe there is play in the servo? (servo is new) how can I determine if the play of the servo is within its allowed tolerance? It seems to get worse, when I dial up the gain to 135. But I will play with the gain a bit more and see where it takes me. Here's a video, where I flew side to side a bit to enhance the problem. In forward flight, it twitches constantly from side to side, which makes the footage very nervous. The CG of the roll axis is rather below than above the axis, just a tiny bit, to prevent it from going side to side due to it's weight. here's the setup: I use a Canon 5D MK2 with a 24-105 IS Zoom lens. Now, I know this is a heavy lens, and I have read that it's probably not a good idea to use it on the 3 Axis Gimbal, but the problems remain with my small 50mm lens, but that doesn't have an image stabilizer, which is the only thing at the moment, that keeps me from loosing my head. Radian gains are tilt: 95 (doesn't go higher because of the jitter getting worse) roll: 105 Props are balanced in all directions, and motors are balanced with props on until the vibrations were at their minimum. The gimbal is balanced, the camera stays in every angle with belts off. I had to buy the 200mm carbon tubes and the adjustable tilt bar to get the CG right. I am using the wooden plate mod to stiffen up the gimbal mounting plate, as seen here by brads pictures: http://forum.freeflysystems.com/index.php?threads/roll-shake.302/page-2 also using the cable tie mod to reduce the side to side movements of the dampeners, I use 2 black o-rings at the end of each dampener, rest is red. Problems remain with all red o-rings. all my belts are as tight as I can get them. (even the one around my waist) I read this forum a lot since I am working on these issues, and it has been very helpful. thanks guys. But it has also shown me that my first problem at least is not very common, and I should probably be able to solve it. Here, Brad has a similar setup, but really smooth footage without an image stabilizer: http://forum.freeflysystems.com/index.php?threads/roll-shake.302/page-2 that would be my goal. And now please don't say, wait for the Movi, because the appearance of this new brushless system is the most frustrating part of my life. (At the moment) It feels, as if I am working on something, Tabb solved by using another technology, which I can't afford.
Get the MR10! What does your footage look like when your flying with all the Radians OFF? Is the footage nice and smooth looking? No jitter? Josh
Hans you could try to going to all black rings. But that might not get rid of all of it. Other than that you have done the other possible steps, like swapping/replacing servos. Your second video is the one that is a worse case since you are hovering and that is the worst situation since you are sitting in 'dirty, disturbed' air. All that being said simply use something like Warp Stabilizer to get to the final spot. I asked a lot of folks at NAB who have shot with the $100K+ Cineflex and each and everyone of them said they still had to use some type of post stabilization.
Hey Hans Looks like you are very close. A couple of things? When you tap the gimbal does it vibrate on the tilt with the radians on? You should be able to go to about 130-150 or more on tilt and it won't oscillate after you tap it. Yes as Josh said, lets see a Radian off clip or at least tell us what its like? It should be absolutely smooth. As far as technology goes... trust me there will be conversion components for a few hundred dollars, some already available. Wait for a few more months and the things is going to blow wide open. The direct drive gimbal that is rock solid is not far off for all. Even big cameras! Just do a search on youtube for Alexmos and his 3rd axis prototype.
Hey Josh, here is the shot without radians, the image stabilizer was off in all the videos I posted so far. I can see a bit of jittering even here, but it's very little. That must be some vibes from the cs. Gary: I had all black o-rings on, (because I couldn't get my hands on blue) and they were very stiff. I had the odd vibration coming through from the CS. I fly with quite a high Gyro D (18) which tends to vibrate a bit, but the micro wobbling is gone. I tried the Final Cut X built in stabilizer and they make the shots look steady. I have a headache cropping and therefore reducing the quality though, so I would like to use as little post stab as possible. I'll try Warp Stabilizer and see, if there is any quality difference. Brad: I just did the tap test with the heavy 24-105 lens and it vibrates a lot on all gains, probably due to belt and gimbal flex. It reaches out far and is heavy. On a gain of 30, the vibrations after the tap stay the same, but I can move the tilt really smooth. But that being said, on 30, the stabilization wont help me much, I guess. I don't understand though, why I should have the same jello effect with the short 50mm lens, if it doesn't vibrate after the tap. The agitated movement, that brings up the jello seems just as bad with the short and light lens on the same low tilt gain of 95. The best alternative would be buying a 24 or 35 mm IS, as most pilots with a 5D here use. Then I would still have vibes, but could iron them out with the image stab. It just seems to me, I haven't found the real source of the jello yet. Here's a video of the tap vibration on gain 95. Maybe you guys can compare. Having in mind, with this heavy lens, the vibration is a lot stronger, because it reaches out far. But as I said, the vibrations in the movement remain with a short 50mm lens. I hope your right about the technology part. I think leaving the belts off, will be a lot easier. BTW I watched the Alexmos video that starts in the bathroom in front of the mirror. I love that small kid (probably his son) that he films running outside, it's funny and full of energy. It made me laugh.
I see the small vibration in the Radian off shot as you mentioned. I also see you have APC's ... have you tried Xoars? APC can be decent and solve some vibes but also can cause them. Try switching to a much more rigid prop and you will see a huge difference if you balance them correctly and your motors bearings are all good. Generally, anything that isn't rigid could be causing problems for you. As far as using a longer and heavier lens... you might try putting some foam under the camera. What is going on is the tuning fork effect and it is because the camera is hard mounted and all the focus is put onto the servo slop and belts which are the weak points. If there is just a slight bit of play between the camera and the mount, because of some foam in between, it will most likely cut some of the resonance out. You'll need to play with it to get it right. I put foam between my cameras and the mount always. Depending on the situation, quite a bit sometimes. Yes I have been doing a project with direct drives for 360's ( or any gimbal ) for awhile now. Bill's thread shows another route and there are many you can purchase online now which will do similar with a little elbow grease. As someone in another forum said " Its a brushless summer"
thanks! I will try the Xoars one day. I am quite curious about the change they will bring to flight characteristics.
Brad, What kind of foam are you talking about? Like foam from a Pelican case? How thick? I just bought some foot gel cushions and was going to insert a piece between the camera and the mount. While I have you here Brad, you are one of the few WKM users here and seem to get good results. I've tried all the tricks except messing with my WKM settings. Would you mind sharing your WKM settings? I'm trying to fly my 5d. Getting closer, but still not satisfied. thanks! Bry
I use stiffer foam like double sided mounting tape and slightly softer and some rubber as well. You just have to experiment My settings are somewhere around 185 185 125 125 75 75.
I have always wondered if it mad more sence to have dampening on the mounting place versus the center hub. That would get rid of the gain issues with pan jitters and make it easier for the jello to be isolated to the camera. But I think josh is right the brush-less gimbal will solve all of our woes. I had a friend come down this week from Durango Colorado with his Zenmuse and S800 and Nex7. We shot each others birds in the air at the same time. We had had really gusting winds that were racking both our aircraft. His Zen was able get some awesome footage of my bird, unfortunately I was not able to return the favor. All of my footage of his bird turned out to be shyte. Albeit I was using my Vixia instead of the 5D because of the 10knot gusts upto 20 I was a little nervous if I would be able to hold her without it being over controlledso I chose a lighter camera. My vixia has been racked before when I crashed my 550, but still that zen even w/o any post stab was almost spot on. My 5D I would have had to still use warp stabilizer and maybe would have got some good footage. I find that if you get can the best stable footage from the aircraft w/o post stab work then post stab such as warp stabilization looks even that much better. One of the limitations warp stabilization has, is if the jitter is too aggressive, then you get a lot of motion blur artifacts and it looks dirty. @ Brad I will love this new brushless summer. In a few years it seems that this "little hobby" has graduated from the barn storm era to the Jet age
Thanks Brad. I've been toying with the idea of stuffing foam inside the tubes of the gimbal to help deaden vibrations. Or sorbathane gel pads stuck on the arms and down tube. Drummers use them in their drum kits to get rid of unwanted vibration.
Bryan, I was thinking of the same thing, but instead of foam I was planning to use dynamat. It's used to kill vibrations in extreme car audio setups. It's a sticky-backed rubber that comes in a few different thicknesses. If you use foam under the camera, only memory foam is the way to go IMO. Everything else including gel is bouncy. If vibrations are getting through the CS vibe isolators a then you have bigger problems. At far as the roll twitch goes, I took my own advice and cut a small foam "donut" and put it under the roll servo pulley. I was able to take the roll gain up from 95 to 170 without twitching and the roll stabilization is a lot better.
Tim: Regarding the foam "donut." I not sure I quite understand when you say you put it "under" the roll servo pulley -- do you mean the donut is interposed between the belt and the pulley? Or do you mean the donut is around the splined shaft that drives the roll servo pulley? If you have time, could you post an image please? Thanks Andy.
Tim, Do you have a source for the memory foam? I'd assume it would have to be a relatively thin layer to put between the camera and the mount? How thick? What's the theory behind adding resistance to the pulley? Why might that improve performance?
The foam is between the servo pulley and the top of the servo case, so the resistance reduces the effective backlash as much as possible. The piece I cut was from egg-foam used in sound studio walls. It's not rubbery, but it has a nice firmness to it. I've also used felt in the past. The piece is about 3/8 to a 1/2" thick and rounded like a donut. Much thicker and I would risk over-amping the servo/ BEC, I think. Bryan, I don't really have a source for memory foam besides stealing it from the end of the bed pad. Flying-cam used to surround their cameras in foam, but I've never had success trying to damp vibes at the camera. I would look up the food chain instead.
This is where I am in 5-8mph wind. Still not great. The first part of the flight the Radians are off so you can see how she flies raw, then you can clearly see at 33 seconds when I switch them on and they jump to attention. standard C8 3 axis gimbal tilt 85 roll 145 APC 12x3.8 props WKM at 170, 170, 150,170 75, 75 Canon 5dmII 20mm lens (no IS) perfectly balanced all axis. props, motors balanced. doubled gimbal plate for stiffness, blue O-rings, zip tie mod to vibe isolators, some foam inside the O-rings lessen deformation, felt behind servo pulleys, and days, and days, and days reading every post on this topic several times over I feel I've methodically tested every viable gain setting. I've tested WKM settings up in the 200 range. Not sure what else to do at this point. Wait for affordable brushless?
Bryan I don't see anything wrong with your video. Dialed in very well. I'm watching on a ipad but seriously don't see any jitters on any axis. Even the un Radian beginning wasn't bad.
Thanks Gary, but this is in almost no wind. So I guess you're saying post stab can handle this much errant movements and more? I dont have Warp Stabilizer at the moment to test. What wind speed do you consider too much to get usable shots from something like a 5d? I know there are some variables with gusty-ness etc...but very curious to know what people who are working with a 5d consider too much wind.
If you look closely at your video you can see the lens moving more than the tilt axis itself. You need to support the lens more since the only point of contact for the camera is in the middle of the body. I have the upgraded sliding plate and use some foam under the lens to keep it from bouncing like that.