PetaPixel

Video Demonstrates the Stabilization Magic of a Brushless Gimbal Rig

If you’ve never seen a 3-axis brushless motor gimbal in action, this video might seem out of this world… or fake. But it’s not fake, just the magic of advanced camera stabilization technology that’ll cost you several thousand dollars to wield for yourself (unless, of course, you have the engineering know-how to build one yourself).

There is literally zero detail about this video in the description, so we’re reaching out to you to see if you can fill in the gaps or correct us if we’re wrong. Obviously the guy in the mirror is using an EOS-M and a brushless gimbal stabilizing rig to demonstrate just how well it performs, beyond that, let’s do some speculating.

Some commenters on both YouTube and Reddit (where the video has exploded) seem to believe it’s one of the amazing MoVi rigs from Freefly Systems — find out more about those here — and it indeed looks an awful lot like the MoVi MR configuration of the M10 (although the M10 handles are still on there).

Here’s a behind the scenes look at what the M10 can do… if you haven’t seen it before you should prepare to have your mind blown:

We don’t fancy ourselves engineers, so if you want to dive into an in-depth technical explanation on what exactly a brushless motor gimbal is and how it works, feel free to have at it in the comments.

Otherwise just watch the video, gently return your jaw to its upright and locked position afterwards, and start doing some black market organ sales math — the M10 costs a whopping $15K.

At that price, it’s no wonder this guy decided to show off what his rig could do when he got it home… we imagine we’d do much the same thing.


 
 
  • Richard Cave

    its a Movi,

  • TWVS

    That’s a very expensive selfie.

  • Larissa

    This is incredible, no doubt. But does anyone know what justifies the price? Is it using crazy expensive materials? all hand assembled?… in the upper east side? I’m genuinely interested!

  • Ralph Hightower

    A Kickstarter project?

  • Michael Andrew Broughton

    the first video is deceptive. the camera seems more stable than it actually is because it’s looking at it’s own reflection. look at the background to see how stable it actually is.

  • Dhaval Panchal

    This is the perfect ad for it. Shame Mercedes beat them to it! http://youtu.be/nLwML2PagbY

  • Photog_

    I believe this model features some form of computer stabilization that would add to the price.

  • http://www.sin3rgy-creative.com/ David Liang

    I get what you’re saying there’s a little momentum the Movi can’t completely account for, but it’s still very incredible. Lining the top of the frame with the edge of the bathroom wall where it meets the ceiling, we can see that while there’s minor movement it’s so incredibly subtle and controlled.

  • Cameron

    The Movi comes with a remote control with which a second operator can monitor the shot and control the pan and tilt of the camera while the first operator can focus solely on the movement of the shot.

  • JudeJackson

    To answer your question, a brushless motor is simply a type of motor that is good for stopping and changing speed and direction quickly. A gimbal just a pivot, and you can combine multiple gimbals to allow rotation (or stabilization) along multiple axes. You can see all three gimbals on this rig pretty clearly at about the 5 second mark when the rig is turned forward: One gimbal is the joint between the handles and the tripod, one is the two joints next to the camera, and the third is the joint directly behind the camera. A brushless motor gimbal rig is a rig with brushless motors in the gimbals so that a computer can control the pivot on-the-fly, and by solving for inverse-kinematics, you can program the motors to keep the camera from rotating. Since most camera jitter is rotation, this works pretty well as a stabilizing system.

  • sam

    since i playing RC.. i thinking i can build one with few gyro, brushless motor and ESC, and some joints… but seems like it will be expensive

  • fisherman10

    I think LG used it for their LG G2 camera with OIS too!

  • Four Letter Nerd

    A perfectly stabilized stedicam shot would be boring, because it would lack the ‘swoop’ feel.

  • Sky

    Test like this is actually quite easy for motorized stabilization systems. Show me how it can compensate for vibrations – than we can talk.
    In that price point stabilization systems are used by heli / airplane photograpers as well as advanced multicopters where compensating vibrations is very important, and this is also while these systems often fail.

  • Hieu Luu

    Not, not even close

  • Hieu Luu

    The price is in the software for stabilization. There’s some cheaper models floating around, but you will notice it wobbles a bit the Movi uses a high end brushless motors built to their specs and they tune it with their software where you can rig up a Red camera and it still stabilizes.

  • Hieu Luu

    Of ALL the 3axis gimbal videos out there, this is the one Peta picked? Peta really need to do their research.

  • Rob Dunlop

    Watch the first 2 seconds of the video again a few times, and ask yourself how the thing stays level when HE TAKES ONE HAND OFF IT. It’s at least partially fake. Which is all the more absurd given that there are plenty of devices out there, like the Movi and Besteady One that can do what this thing purports to do. Another big clue as to the authenticity of this clip is the lack of any details about the product, which is apparently a DIY brushless gimbal. You don’t invest that much time, money and effort in creating such a device and then neglect to give any details when you reveal the prototype. This is the eagle snatching the toddler all over again.

  • Rob Dunlop

    Some youtubers are saying the fromt of the camera is stuck to a pane of glass, which makes sense.

  • Annoyed

    Awesome! I bet it only costs 30k too!

  • JudeJackson

    As demonstrated in the second video, this type of rig doesn’t require two hands to hold the system at all times to work; it’s only important that the operator holds one of the upper handles. This allows complicated hand-off shots that aren’t easy to reproduce with other types of rigs. It is certainly possible for this video to have been faked, but the technology exists. Most YouTube videos don’t come with descriptions and I certainly am not going to start thinking that real camera products don’t exist just because one video didn’t have a description.

  • Icon

    I’m not too sure but there are literally thousands of such gimbals from chinese makers(go to goodluckbuy dot com and search brushless gimbal)…you can build a full rig for under 300 use) and they almost all look the same. If you’re into RC planes this stuff is old news…the technology has been out there since Aleksey Moskalenko designed the AlexMos board to drive brushless motors(instead of servos) to drive the gimbal stabilization. What Movi and others have done is 1) make this stuff into a ready to use product(no engineering required) and 2) made a very high quality design of the actual mechanical components so they are reliable and most importantly easy to balance. I’ve built mine at home for less than 200 dollars 3 months before the Movi was even announced…It wasn’t easy to get it work reliably though…you get what you pay.
    The landing gears shown in the above video look like the Cinestar old servo gimbal gears(and there are hundreds of knock-off that copy that same design) as for the gimbal, from the look of the motors this is obviously not a motor and looks to be one the DYS or iFlight motor based gimbals….I could be wrong of course but while this is a good demonstration of these gimbals work, it is nothing special.

  • Bogdan

    You can help stabilize vibrations with a gyro. In fact I cannot expalain why Movi does not use one. The brushless motor design is great not that much for the stabilisation but for the fact that a camera operator can turn or tilt the camera while let’s say somebody runs with it or it travels across a wire system.

  • LeChatNoir

    Not a particularly impressive 3-axis camera gimbal, the video appears to be more impressive than it is.
    Since a video via a mirror is by definition aligned to the distance the camera can move in the time it takes for light to bounce off the camera reflect of the mirror and hit the sensor.
    All you have left is the shutter speed to cause blur between the motion.

  • Rob Dunlop

    Watch it again. You can only hold it one handed with the top bar, unless the device is INCREDIBLY light. If you don’t believe me, try lifting up a chair holding it by two of its legs, then take one hand off and see what happens. In this faked video there is NO DROP AT ALL. so I call b.s.even though there are hundreds of working brushless gimbals out there which could actually do what this thing does.

  • JudeJackson

    There is clearly instability as he does the handoff, though with a combination of muscles and impressive controllers the camera is well-stabilized. You can still see clearly in the second video that these rigs are designed for one-man handoffs, which means at some point there has to be one hand holding a side handle. The fact that it’s more probable to you that someone built an elaborate fake rig, installed a pane of glass to the front of his camera, had two buddies hold the glass, simulated the sound of motors, and then filmed thirty seconds of himself goofing around in a bathroom mirror, and then uploaded it to the internet to try and make himself an internet sensation (despite the real thing actually existing), instead of just buying a real rig for anywhere between $300-15000 and making a little test shot that he thought was cool enough to show his friends on YouTube, illustrates some kind of intense distrust you have for humanity.

  • Rob Dunlop

    Ok looking at his other videos I’m prepared to admit that I was wrong. This thing must be incredibly light, but I would have expected a dip when he shifted to one handed from the side, even with carbon fiber materials and a very small camera. FYI the Movi etc are NOT designed to be handed off with one hand only supported the device from the side at any time. All these devices have long handles on the sides and top, so two people can share a grip – you hand it off top to top or sides to sides or sides to top or top to sides, as you can see in the Movi clip above. Stick a 5D + 16-35 with heavier motors on there, like a Movi, and holding it one-handed from the side for even a moment would cause an immediate dip I’m sure. Stabilisation is one thing, but these things don’t have rocket engines underneath :) On his other clip, this guy is holding the thing from the side for much longer, I still can’t get my head around it.

    As for elaborate fakes on Youtube, dude, are you kidding? It’s par for the course.

  • JudeJackson

    I think you’re vastly underestimating the speed of light. The speed of light is 299,792,458 meters per second, which means if the operator is standing one meter from the mirror and the camera is filming at 60 frames per second, the light could still travel the distance between the camera and its own reflection 2,498,270 times between each frame.

    But it’s true that the stability is somewhat deceptive, because a head-one reflection will always appear stable head-on. You can see in the background the lag in rotation, which is still pretty impressive all things considered when you see how smooth it is.

  • LeChatNoir

    Since I’m bored I actually did some(quite questionable calcs).
    In a mirror with the light time lag there is relative motion, it’s just undetectable with most equipment.

    I used a full frame to amplify the motion difference:
    Full frame at 5760×3840, 6.25µm square pixels(5D Mark III).
    Camera rotating at 2/3 rotations a sec.
    c in air (STP).
    distance 2 meter total.

    Corner pixel outer edge movement: 0.3023507 nm
    Corner pixel inner edge movement: 0.3022902 nm

    ^ That’s roughly how far the corner pixel would move in the time it takes for light to bounce off the camera to bounce off the mirror to hit the sensor.

    Which considering the pixel size is insignificant, about 0.06% light shift off the pixel, then again I already knew it would be <0.1%.

    This is also assuming perfect rotation on center of image sensor, which with only a 3-axis you can't do with such a rig at these precisions, not to mention varying degrees of gas compositions affecting the refractive index of the air as well as the different refractive indexes of the lenses affecting the flight time leading to more of an error.

    Then again none of this would show up on a video of less than a 9,455.08 terra-pixel sensor crammed into a 35mm format, at which point you would have a 1 pixel shift, but it would be blurred out over the exposure time.
    —–
    It's that somewhat large amount of lag(or flexing of the rig in comparison to the gyro sensor or microcontroller lag, or a bunch of other stuff) that I called it not particularly impressive.

    As for whether it does what it tries to do, I agree that is it usable, if far from rock steady.

  • Raul

    What is truely amazing is that people spend thousands of dollars to buy or hundreds of hours to build these rigs and then do generally nothing but wiggle them in front of a mirror…

  • Salvador Peña

    You can even skip the inverse kinematics and just use a simple PID control, or even a really well tuned PD for each joint, load them in a fast microcontroler (16 MHz at least) and let them run, inverse kinematics would require a little bit more processing power, thus ending up being more expensive.

    Salvador Peña
    Robotics Engineering Degree
    Mexico

  • JudeJackson

    Whoops now that I think about it that’s completely obvious. All those robot arms made me want to overcomplicate things.

  • http://www.vincentmorretino.com/ fast eddie

    Bathroom mirrors, at that. I can’t be the only one with a large mirror outside of my bathroom.

  • Brian

    Noisy. Worthless for dialogue scenes.

  • Jacob Dole

    $800 I’ll make you one from carbon fiber… using Alexmos firmware. Basically it’s a 3 axis frame with 3 servos or motors + ESCs and a 3 axis gyro controller board.

  • csmith

    is there a reason why the image quality in that film form movi is so atrocious? almost no contrast and washed out colors…the gimble seems to be working great but what’s wrong with the camera?