bicycle animation!
November 27, 2011 12:33 AM Subscribe
"This is a piece created to question whether it was possible to film animation in realtime."
Actually pretty cute. Basically the same thing as a zotrope, except using the camera shutter to separate frames, as opposed to a narrow slit or a strobe light.
(and even though it's had its own FPP somewhat recently, I can't resist linking to the Toy Story animated sculpture thingy)
posted by ShutterBun at 1:32 AM on November 27, 2011
(and even though it's had its own FPP somewhat recently, I can't resist linking to the Toy Story animated sculpture thingy)
posted by ShutterBun at 1:32 AM on November 27, 2011
Expected this to be a long drawn out argument. Turned out to be a moving experience.
posted by hal9k at 1:35 AM on November 27, 2011 [2 favorites]
posted by hal9k at 1:35 AM on November 27, 2011 [2 favorites]
Ugh...please disregard my link, it's terrible. There are much better versions available. Apologies.
(also please forgive my misspleing of zoetrope)
posted by ShutterBun at 1:37 AM on November 27, 2011
(also please forgive my misspleing of zoetrope)
posted by ShutterBun at 1:37 AM on November 27, 2011
I remember seeing hubcaps do this on the Dukes of Hazzard back in 1981.
posted by scrowdid at 1:42 AM on November 27, 2011
posted by scrowdid at 1:42 AM on November 27, 2011
Yeah, I've seen stuff like this before. Nice shots but this isn't groundbreaking.
posted by delmoi at 4:05 AM on November 27, 2011
posted by delmoi at 4:05 AM on November 27, 2011
I hope the instructor returned a note saying "OF COURSE YOU CAN FILM ANIMATION IN REAL-TIME WE COVERED THAT IN WEEK TWO" and gave them an A+ anyway.
posted by bicyclefish at 6:14 AM on November 27, 2011
posted by bicyclefish at 6:14 AM on November 27, 2011
I think this will be the definitive Pixar Zoetrope video: it includes a few examples of the thing's historical inspiration and none of that blinking effect when the Toy Story crew is spinning.
posted by pokermonk at 6:40 AM on November 27, 2011 [2 favorites]
posted by pokermonk at 6:40 AM on November 27, 2011 [2 favorites]
Thanks for that better-quality video, Pokermonk. The first one gave me a headache. That Studio Ghibli zoetrope is so charming, too. I missed the FPP on the Pixar one first time around; off to read it now.
posted by Gordafarin at 8:27 AM on November 27, 2011
posted by Gordafarin at 8:27 AM on November 27, 2011
Funnily, I think that this would not work in real life, you need a second rythym to get in phase with the first, like the slits on one side of a zoetrope through which you see the animation. In this case, the video frames make it happen. Works also at night, under fluorescent light (which pulses), or if you saw the bike zip past behind an iron railing.
posted by Tom-B at 10:13 AM on November 27, 2011
posted by Tom-B at 10:13 AM on November 27, 2011
Peter Hudson does these on enormous scale at Burning Man.
posted by inpHilltr8r at 11:06 AM on November 27, 2011
posted by inpHilltr8r at 11:06 AM on November 27, 2011
Yeah, I'm not sure that Tom-B's comment is necessary, but in case it is, it's not the case that you'd see these animations directly. It only works because of the relative timing of the captured frames of the video and the angular velocity of the wheel.
However, as Tom-B says, you could rig up a second spinning wheel with slits to create for your eye the equivalence of a sequence of "still frames".
This also may not be necessary for anyone, but in case it is, I'm pretty sure that the human visual system (or, for that matter, any animal's visual system) doesn't "see" like a motion camera does, as a sequence of complete frames. It is true that different portions of the system, from the receptors in the eye to the sequences of neurons in through the brain, have firing rates—it's not a continuous signal. But those rates are different from each other, I'm pretty sure; and, furthermore, it's not the case that the whole visual field is processed at once, nor even that the various components of it (such as apparent lines, colors, apparent motion) are processed at the same time. There's different timings and rates for all of this.
Because of this, there are a number of interesting perceptual illusions that can be generated that exploit particular timing sensitivities in human vision. You can generate false colors, you can generate false apparent motion, many things like that.
But what you're not going to be able to do is to exploit some sort of general timing of a whole-frame visual image produced in human vision in order to generate apparent motion as we see in this video. To do that, you actually need to generate what to the eye (but not further along the pathway to the brain) appears to be a sequence of still images. Film and video do this, a zoetrope does this.
And that works because further along the processing pathway, our visual system above a certain rate cannot distinguish a sequence of still images as still images and, instead, if there's congruency from one image to the next in a way that corresponds to the continuous imagery in the real world that we've evolved to interpret, we perceive it as continuous imagery and (if there is any apparent) motion.
What's really interesting about human vision and imaging technology to me is that all our imaging technology, both still and motion, exploit what you might consider "idiosyncratic" features of human vision in order to reproduce the experience of viewing a scene. What our technology absolutely doesn't do is reproduce with any remote accuracy the actual physical conditions that would have generated that visual experience in person. Put another way, to many other animals on Earth our still and motion images are uninterpretable, meaningless, because their visual systems work just differently enough from ours that either they can see the image more for what it really is rather than what it pretends to be (for example, as a series of still images rather than apparent motion), or the differences produce the wrong perceptions and thus something entirely different from the real-life scene.
And that's just animals on Earth. There's no reason to assume that aliens would have visual systems that work enough like ours that our images, still or motion, would make any sense to them. If they're intelligent and technological and imaginative, then given our imagery they'd eventually be able to figure out much of what it's trying to represent. But to do that, they'd first have to figure out what these images are telling them about how humans eyes and visual perception works. Until they did that, television images would be gibberish to them. I don't mean the electronic signal, I mean even if they had a television in front of them to watch the signals with.
They'd quickly figure out that at about thirty frames a second, we perceive apparent motion. But they'd also have to puzzle out things like what is the smallest resolution (expressed as fractions of degrees of our visual field) beyond which we can't see further detail. You know, why it is that you perceive lines forming letters on the screen in front of you rather than dots in patterns.
More difficult would be our color vision. All our imagery reproduces only the (partial) experience of seeing colors, not the light that would have created that experience of color in the original scene. Our three-color system of producing the experience of the full spectrum of color in human vision is a sort of code that aliens would have to decode. And, even then, without knowing the spectrogram of Sol in general, and its spectrogram after it passes through the Earth's atmosphere in particular, they'd not be able to figure out where, exactly, on the spectrum these apparent colors were intended to represent.
So the point here is that what seems like an accurate re-creation of reality is far, far from it. It's not an accurate re-creation of reality at all. It's at best a semi-accurate re-creation of a human being's visual experience of that piece of reality. And if you've digested that, then consider something more weird: our actual visual experience of reality is itself a sort of a hack of the idiosyncratic peculiarities of how things look on Earth, on land. We mostly don't even see the things we think we see—our visual system is filling in lots of missing data with what's most likely, based upon the history of evolution of life on this planet.
And so there is no human visual experience (or any experience) that isn't very heavily mediated. It's all an artifact—a long, long way from reality itself.
posted by Ivan Fyodorovich at 11:19 AM on November 27, 2011 [1 favorite]
However, as Tom-B says, you could rig up a second spinning wheel with slits to create for your eye the equivalence of a sequence of "still frames".
This also may not be necessary for anyone, but in case it is, I'm pretty sure that the human visual system (or, for that matter, any animal's visual system) doesn't "see" like a motion camera does, as a sequence of complete frames. It is true that different portions of the system, from the receptors in the eye to the sequences of neurons in through the brain, have firing rates—it's not a continuous signal. But those rates are different from each other, I'm pretty sure; and, furthermore, it's not the case that the whole visual field is processed at once, nor even that the various components of it (such as apparent lines, colors, apparent motion) are processed at the same time. There's different timings and rates for all of this.
Because of this, there are a number of interesting perceptual illusions that can be generated that exploit particular timing sensitivities in human vision. You can generate false colors, you can generate false apparent motion, many things like that.
But what you're not going to be able to do is to exploit some sort of general timing of a whole-frame visual image produced in human vision in order to generate apparent motion as we see in this video. To do that, you actually need to generate what to the eye (but not further along the pathway to the brain) appears to be a sequence of still images. Film and video do this, a zoetrope does this.
And that works because further along the processing pathway, our visual system above a certain rate cannot distinguish a sequence of still images as still images and, instead, if there's congruency from one image to the next in a way that corresponds to the continuous imagery in the real world that we've evolved to interpret, we perceive it as continuous imagery and (if there is any apparent) motion.
What's really interesting about human vision and imaging technology to me is that all our imaging technology, both still and motion, exploit what you might consider "idiosyncratic" features of human vision in order to reproduce the experience of viewing a scene. What our technology absolutely doesn't do is reproduce with any remote accuracy the actual physical conditions that would have generated that visual experience in person. Put another way, to many other animals on Earth our still and motion images are uninterpretable, meaningless, because their visual systems work just differently enough from ours that either they can see the image more for what it really is rather than what it pretends to be (for example, as a series of still images rather than apparent motion), or the differences produce the wrong perceptions and thus something entirely different from the real-life scene.
And that's just animals on Earth. There's no reason to assume that aliens would have visual systems that work enough like ours that our images, still or motion, would make any sense to them. If they're intelligent and technological and imaginative, then given our imagery they'd eventually be able to figure out much of what it's trying to represent. But to do that, they'd first have to figure out what these images are telling them about how humans eyes and visual perception works. Until they did that, television images would be gibberish to them. I don't mean the electronic signal, I mean even if they had a television in front of them to watch the signals with.
They'd quickly figure out that at about thirty frames a second, we perceive apparent motion. But they'd also have to puzzle out things like what is the smallest resolution (expressed as fractions of degrees of our visual field) beyond which we can't see further detail. You know, why it is that you perceive lines forming letters on the screen in front of you rather than dots in patterns.
More difficult would be our color vision. All our imagery reproduces only the (partial) experience of seeing colors, not the light that would have created that experience of color in the original scene. Our three-color system of producing the experience of the full spectrum of color in human vision is a sort of code that aliens would have to decode. And, even then, without knowing the spectrogram of Sol in general, and its spectrogram after it passes through the Earth's atmosphere in particular, they'd not be able to figure out where, exactly, on the spectrum these apparent colors were intended to represent.
So the point here is that what seems like an accurate re-creation of reality is far, far from it. It's not an accurate re-creation of reality at all. It's at best a semi-accurate re-creation of a human being's visual experience of that piece of reality. And if you've digested that, then consider something more weird: our actual visual experience of reality is itself a sort of a hack of the idiosyncratic peculiarities of how things look on Earth, on land. We mostly don't even see the things we think we see—our visual system is filling in lots of missing data with what's most likely, based upon the history of evolution of life on this planet.
And so there is no human visual experience (or any experience) that isn't very heavily mediated. It's all an artifact—a long, long way from reality itself.
posted by Ivan Fyodorovich at 11:19 AM on November 27, 2011 [1 favorite]
Also previously: Tim Wheatley's bicycle-wheel "cyclotrope."
posted by arm's-length at 12:40 PM on November 27, 2011
posted by arm's-length at 12:40 PM on November 27, 2011
"This is a piece created to question whether it was possible to film animation in realtime." -- And was the answer is "no?"
posted by crunchland at 12:55 PM on November 27, 2011
posted by crunchland at 12:55 PM on November 27, 2011
Come on ReeMonster, I'm sure they know that. Keep your scholastic regrets and bitterness to yourself, you don't need to impress us.
I think it's interesting that like gifs, all these animations can do is repeat over and over.
posted by victory_laser at 1:20 PM on November 27, 2011
I think it's interesting that like gifs, all these animations can do is repeat over and over.
posted by victory_laser at 1:20 PM on November 27, 2011
Well, just think of a film strip as a very large bicycle wheel that's been wrapped around some smaller bicycle wheels.
posted by RobotHero at 2:10 PM on November 27, 2011
posted by RobotHero at 2:10 PM on November 27, 2011
Pretty neat effect! Especially liked the ones that started about 1:20.
posted by darkstar at 4:25 PM on November 27, 2011
posted by darkstar at 4:25 PM on November 27, 2011
I watched the We Got Time video just the other day, another awesome use of zoetrope.
posted by mrzarquon at 5:49 PM on November 27, 2011
posted by mrzarquon at 5:49 PM on November 27, 2011
« Older "Suppose there was a place where there wasn't even... | Free speech is only as strong as the weakest link Newer »
This thread has been archived and is closed to new comments
posted by ReeMonster at 12:54 AM on November 27, 2011