Back in early 2005, at the height of my HDV camcorder infatuation, I came across an article titled “$14 Video Camera Stabilizer” in Volume 1 of Make: Magazine (I’ve been a faithful subscriber for a long time). The author, “maker” Johnny Lee (who subsequently found well-deserved fame and presumed fortune at Google), prefaced his piece with this tempting prose:

You don’t have $10,000 to spend on a Steadicam? Make this ultra-low-cost video camera stabilizer and see how much better your video shots turn out.

I bit. Hard, actually…instead of trying to make my own, I went ahead and bought a prefabricated “Poor Man’s Stabilizer” kit from his (now apparently defunct) Little Great Ideas LLC side company. It’s still sitting on the bookshelf behind me as I type these words, actually. It worked passably, and certainly was much less expensive than the real thing. But it was bulky and heavy, and anyway, my aspiration to become the next Robert Rodriguez hasn’t amounted to much, at least to date.

Hope springs eternal, however, which is why gimbals have long caught my eye. Unlike a traditional stabilizer, which harnesses weights and handles to compel the camera operator to hold steady, a gimbal actively counteracts the operator’s movement, whether inherent or caused by his/her environment, such as by an earthquake or (less dramatically) the pitching and rolling of a watercraft. Speaking of which, as Wikipedia notes, although the term is nowadays most commonly applied to the camera accessory (at least in tech circles), its conception and subsequent use long predates that particular implementation of the concept:

A gimbal is a pivoted support that permits rotation of an object about an axis. A set of three gimbals, one mounted on the other with orthogonal pivot axes, may be used to allow an object mounted on the innermost gimbal to remain independent of the rotation of its support (e.g. vertical in the first animation). For example, on a ship, the gyroscopes, shipboard compasses, stoves, and even drink holders typically use gimbals to keep them upright with respect to the horizon despite the ship’s pitching and rolling.

The gimbal suspension used for mounting compasses and the like is sometimes called a Cardan suspension after Italian mathematician and physicist Gerolamo Cardano (1501–1576) who described it in detail. However, Cardano did not invent the gimbal, nor did he claim to. The device has been known since antiquity, first described in the 3rd c. BC by Philo of Byzantium, although some modern authors support the view that it may not have a single identifiable inventor.

Gimbals’ use for various video capture devices has dramatically increased in recent years, driven by a combination of increased demand and increasingly robust capabilities. Demand first: as anyone who’s perused Facebook (and/or its Instagram subsidiary), YouTube or (particularly of late) TikTok can attest, the amount of user (vs professional)-created video content is exploding, as are viewers’ quality expectations of it. Hardware capable of capturing high-quality images is no longer restricted to pro-grade camcorders, either; prosumer HD and 4K camcorders selling for under $500 are now widely available, for example. That said, they’re being squeezed out of the market by DLSRs, mirrorless interchangeable-lens and even fixed-lens point-and-shoot still cameras that also capture high-res, high-quality video footage. And they’re being squeezed out, as I recently wrote, by modern video-capable smartphones.

What about the gimbals themselves? Here I’m focusing (pun intended) on the ones used for videography, although they also have plenty of applications for still imaging, especially in conjunction with telephoto lenses (in wildlife photography, for example, or astrophotography). Wikipedia again:

Handheld 3-axis gimbals are used in stabilization systems designed to give the camera operator the independence of handheld shooting without camera vibration or shake. There are two versions of such stabilization systems: mechanical and motorized. Mechanical gimbals have the sled, which includes the top stage where the camera is attached, the post which in most models can be extended, with the monitor and batteries at the bottom to counterbalance the camera weight. This is how the Steadicam stays upright, by simply making the bottom slightly heavier than the top, pivoting at the gimbal. This leaves the center of gravity of the whole rig, however heavy it may be, exactly at the operator’s fingertip, allowing deft and finite control of the whole system with the lightest of touches on the gimbal.

Powered by three brushless motors, motorized gimbals have the ability to keep the camera level on all axes as the camera operator moves the camera. An inertial measurement unit (IMU) responds to movement and utilizes its three separate motors to stabilize the camera. With the guidance of algorithms, the stabilizer is able to notice the difference between deliberate movement such as pans and tracking shots from unwanted shake. This allows the camera to seem as if it is floating through the air, an effect achieved by a Steadicam in the past. Gimbals can be mounted to cars and other vehicles such as drones, where vibrations or other unexpected movements would make tripods or other camera mounts unacceptable. An example which is popular in the live TV broadcast industry is the Newton 3-axis camera gimbal.

Perhaps obviously, from my prior comments, motorized gimbals are of most interest to me. And the ongoing innovations in those brushless motors’ form factors, power output and other parameters, along with that of the batteries that power them and the IMUs (accelerometers, gyroscopes and sometimes also magnetometers, all now also prevalent in smartphones and the like) that modulate their operation, are fundamental to gimbals’ price, size and weight, between-recharges operating life and other characteristics critical to adoption by the masses.

By the way, those same motors can find use not only in keeping the camera still but also in moving it in 3D free space while footage capture is in process…under videographer power via input buttons integrated into the gimbal’s grip, and/or fully handled by the gimbal itself after the videographer selects from among the pre-programmed patterns. And speaking of full gimbal control, it can also find use in keeping the camera pointed at the human or other subject being videorecorded, either by tapping into the video stream already being captured by the camera (when using a smartphone, for example, which is amenable to user installation of applications developed by the gimbal manufacturer) or via a camera directly integrated into the gimbal, in either case in conjunction with face/object detection and tracking algorithms.

Speaking of smartphones, I already last fall mentioned (within my annual holiday shopping guide for engineers) one case study example, DJI’s Osmo Mobile 3 Combo gimbal:

The latest and greatest in the company’s product line is already up to v5, with iterative updates (magnetic coupling between the phone and gimbal, versus a traditional bracket mount, and a built-in extension rod), none of which I personally found sufficiently essential to justify the incremental price tag versus the refurbished unit (plus a $10 gift card) that I bought last year for $59. I’ve only used it a bit, but it definitely produces superior stability results to those obtainable with my more elementary (but, at $10, even lower priced) JOBY handgrip:

Which, to be clear, are still far superior to my Neanderthal attempts to handhold the smartphone without any accessory assistance whatsoever!

And for my DSLR? Here I also could have gone with DJI…the company seemingly pretty much single handedly created the gimbal category with its Ronin series, which vary in (along with their price tags) the maximum weight of the camera-plus-lens-plus-accessories payload they can reliably handle, along with other parameters. More generally, DJI deserves kudos for the degree to which it’s leveraged the core audio, still image, video, battery and motor competencies initially developed for its drones into other product lines, today also including:

But its historical gimbal dominance is increasingly being encroached on by equally innovative competitors such as another Chinese manufacturer, Zhiyun-Tech. In mid-December of last year, B&H Photo Video briefly put the company’s CRANE 2S gimbal on sale for $299, $300 off the normal price.

I’ve subsequently obtained the “Combo” variant of the various available CRANE 2S “bundles”; it includes a supplemental grip that enables two-hand operation, along with a second set of rechargeable batteries. The “Pro” package includes even more useful accessories—not including, I should be clear, the cameras, lenses and microphone shown in this “stock” photo:

Among them, believe it or not, is something called the TransMount Image Transmission System:

It enables a second “shoot” participant (such as the director) to live-view the video footage being captured by the videographer:

Including, if necessary, even taking over gimbal control for him/herself:

On that point, I should point out one shortcoming of the system in my specific situation, although it has nothing particularly negative to do with the CRANE 2S or, bigger picture, even with Zhiyun-Tech. Ideally, a gimbal will be able to mate up with the standalone or smartphone-inclusive camera either via a wired (USB-C, Lightning, etc.) or wireless (Bluetooth, etc.) connection, so that the user can keep both hands on the gimbal while still controlling the camera through the gimbal’s UI. Zhiyun-Tech understandably focuses its development attention on the larger hardware manufacturers and models, to get maximum return on its investment. Peruse the support list for the CRANE 2S and you’ll see predictable company names—Canon, Nikon, Panasonic, Sony, etc.—just not Pentax. Although the base functionality of the CRANE 2S remains usable in my setup, I need to control the camera standalone: lens zoom, for example.

Looking forward, I can imagine that gimbals will incrementally improve in their size and weight for a given target camera payload, along with operating life and other parameters, via further evolution in batteries, motors, materials and such. That said, harkening to my earlier comments on the DJI Osmo Mobile 3 versus its successors, at this point in gimbals’ evolution, further improvements in fundamental functionality may be modest at best. Having said that, however, I admit that whenever I’ve made similar product observations in the past, I’ve ended up pleasantly surprised with the more significant-than-expected impacts that engineers in-the-know (i.e., far more engaged with markets, products and customers’ needs than me) end up coming up with. So, I suspect I’ll be pleasantly surprised in the future here, as well. What do you see coming from DJI, Zhiyun-Tech and others in this space? Let me know in the comments.

—Brian Dipert is Editor-in-Chief of the Edge AI and Vision Alliance, and a Senior Analyst at BDTI and Editor-in-Chief of InsideDSP, the company’s online newsletter.

You must Sign in or Register to post a comment.