Slashdot Mirror
Ars Technica Builds Make Magazine's Steadicam
An anonymous reader writes "Make magazine has been out for a little over a month now and was given high marks in a Slashdot review. Ars Technica has taken their review one step further by building the $14 steadicam project and testing it out. (be sure to check out the QuickTime video at the end to see their results...)"
make: *** No rule to make target `magazine'. Stop.
no, ms open sources windows under gpl
I was a bit disappointed to see the article, actually -- when a "cool" new print mag recycles the Internet, you know the end of paper is nigh.
So why would you go to the trouble?
Because there are many thousands of us who shoot video on a level between "dad in the backyard with the kids" and "Oscar nominee" who might want some improvement in our quality?
Not to review a review of an instruction, but I think Ars Technica is being a little hard on the Chung. Operating a steadycam is a bit of an artform unto itself.
A steadycam will not turn Shakes the Clown into the next Scorsese, but once you learn the limitations of the axes you'll get results like Mr. Lee posts as samples on his site (see the bottom of the page, under "Using Your Steadycam").
Snickersnee3: Build your own 3-watt Luxeon Star headlamp from scratch
One could easily add some small gyros to a handheld steadicam design and still come out fairly inexpensive. Take three 6v electric motors and a 6v battery. Add a wheel on the end of each motor, and mount them so that the wheels spin in three planes. Add weights to the wheels to balance them. Continue adding weights until you have enough stabilization. Place this apparatus in a padded box to keep it quiet, and mount the box.
Check out my sci-fi/humor trilogy at PatriotsBooks.
I appreciate the effort, but a test video with no subject to focus attention on doesn't really show off how well any of the four tests are working. All I see are tarred cracks in the pavement jiggling in all four.
Whereas, had they walked with someone down the street, it would have shown off the differences quite well.
Jory
Ars' video samples are really subpar for this type of device. Slashdot covered a story on a nearly identical setup here about a year ago and the results are *much* more impressive. As with anything, experience and practice make all the difference in the world. Ars is a great site, but film isn't exactly their focus. :-)
Unfortunately, it's still far too shaky to consider it useful for any indie film that doesn't want to be branded with the Blair Witch style. So why would you go to the trouble?
I'll offer up an answer, since I wrote the review.
I realized today that there was a small problem with our video test: we weren't following any subject. Most steadicam shots are either following a subject or moving around a subject within a few feet of the camera. As you saw in the test, there was no subject. Consequently, even the tiniest movements seemed to make the whole world shake.
The other half (as I mentioned in the review) was that we didn't practice much with the steadicam. I imagine with a little work, we could have gotten a really nice shot. One of the areas that could have used some practice was paying attention to how the side bar is held (since it controls the side-to-side motion). You'll see in the last shot that overall it is very smooth, except that there is a little side to side movement.
All in all, the $14 steadicam was a fun build and worthwhile if you don't have a few thousands dollars for a real steadicam.
Just because they built the tool doesn't mean they're pros at using it. I've read that even after you build the steadycam, it'll take awhile to get used to and become good at. For a better idea of the usefulness of building the steadycam, check out the movies near the bottom of the page by the original creator (especially where he follows the walking subject):
http://www-2.cs.cmu.edu/~johnny/steadycam/
Ars' video samples are really subpar for this type of device. Slashdot covered a story on a nearly identical setup here about a year ago and the results are *much* more impressive. As with anything, experience and practice make all the difference in the world.
This is probably key - professional steadicam operators are trained specifically in how to operate a steadicam (they're not just camera operators who decide to strap on a rig one day for kicks). If you've ever seen any behind-the-scenes footage of steadicam shots being filmed, it's pretty amazing how smoothly these guys move.
A lot of amateur camera operators - be it still or motion picture cameras - think good camera work is almost entirely dependent on the equipment. In fact, I'd say way more than half of what it takes to get good results lies with the operator. You can't put together a steadicam rig and then walk down the street like you'd walk normally and expect a steady shot - that won't work even with a real steadicam. You need to walk as smoothly as possible and make smooth, even camera movements. It doesn't look like that was done here, although to Ars' credit, they do note that they probably could have gotten better results with a bit of practice.
I do think that a rig like this could be a pretty decent option for indie videographers willing to actually learn and practice the proper techniques.
It's just a little Physics, not magic: from Johnny Chung Lee's website (and also in the magazine):
How you use it is 80% of the smoothness. This even is true for the professional stuff with all the fancy shocks and hydraulics. Don't expect this thing to perform miracles, you have to practice using your arms and body to create a smooth motion. Watch your hands while you walk, and see how level you can keep them relative to the ground. Watching the shadow of your hands on a sunny say is an easy way to isolate thier movement. Keep your legs bent and learn how to "glide". I talked with someone who has used professional steady-cams and they said this was, "really, just as good." Getting good results is not so much about the equipment, but how you use it. That's really true about everything.
Sola Deo Gloria!
Comment removed based on user account deletion
Since all you really need to dampen is movement along the pitch and roll axes, a single gyroscope with the axis mounted vertically would be adequate. And rather than trying to build and balance your own, you'd do well to save yourself the headache and pick up a nice surplus military missile guidance gyroscope, like this. Knock that spinner out of its gimbals and I bet it'd be just right...
If a job's not worth doing, it's not worth doing right.
If you want to take home build stabilizers to the next level, check this out. The test videos are VERY impressive.
Professional steadicam operators use equipment that is completely different from consumer/prosumer 'steadicam'-ish gear. They come in teams of two: The one carrying the rig isn't able to look through a viewfinder, so a second person has to control focus, aperture etc. The apparatus has so much inertia that the carrier can actually hop up and down and the camera will still hardly move, let alone rotate. Lots of practice is only half of the story: the huge weight strapped to the body kind of forces the operator to move smoothly, anything else would be totally exhausting.
Equipment in the two figure dollar range, on the other hand, isn't more than a handheld tripod that keeps the centre of gravity around your hands by introducing a counter-weight that roughly equals the weight of your camera. With consumer cameras becoming ever smaller and lighter, that isn't much. Jerks introduced by shaky hands are pretty much all they can compensate for somewhat adequately.
However, this isn't what ars technica actually tested: They walked down a street. The picture resulting from this has to be shaky hand cam style. The "little bit of shake" that the reviewer attested is actually insanely wild. If you haven't got a more expensive steadicam setup, you really only have these two options: Either purposely go for the blair witch style, or let only the actor walk (when shown in the picture, as opposed to the 1st person view) and put the camra person plus camera on some sort of wheeled vehicle, like a dolly with a tripod.
It seems like the ars techinca reviewers had too much fun doing their oh-so-cool project; not only were they too enthusiastic about the marginal improvement in image quality, they also didn't really factor in the problems they had (they needed much longer than the make mag instructions projected and ran into issues the instructions didn't even touch on) when writing the concluding summary.
Not that there's anything bad about this: With a USD 14 handheld tripod, you're supposed to have fun, not emulate Oscar-league steadicam gear. And yes you can train to move super-smooth, but still steadicam and steady camcorder are two entirely different issues.
but what do i know, i'm just a model.