ritter@versatc.versatec.COM (Jack Ritter) (08/04/90)
About 5 years ago, I saw a Siggraph presentation on computer animation. I forgot who the deliverer was, but he showed a computer film of a complex, stick & ball type molecule, rotating. First he showed a version WITHOUT motion blur; all frames were sharp. Then he showed a version WITH motion blur. The upshot was: "Look how much 'smoother' the motion blurred version is, as opposed to the 'stroboscopic' effect of the non blurred version." Well, I LIKED the stroboscopic one BETTER than the motion blurred one. It was simply crisper and more enjoyable than the mushiness of the blurred version. Some of the literature implies that it's good for computers to immitate artifacts left over from cameras, like motion blue (and depth-of-field.) Of course, m.b. helps alleviate "popping fragments", but this was not a factor in the molecule rendering. Are there sound physiological principles that support the need for motion blur? My eyes didnt seem to need it, at least not at the frame/sec rate he showed. Maybe there's another factor: We're so used to seeing motion blur all our lives, from film & tv cameras, that when you take it away, we MISS it. ---------------------------------------------------------------- (Personally, I think the eye-brain system behaves digitally AND analogically, AND in accordance with all responsible, future theories that ever come along.) -- Versatec, Inc. Jack Ritter, M.S. 1-7 2710 Walsh Ave. P.O. Box 58091 Santa Clara, CA 95052-8091 (408)982-4332, or (408)988-2800 X 5743 UUNET: {ames,apple,sun,pyramid}!versatc!ritter --( / __ - .. (( / / / -- ) . \ \ // . ( / ** ) // _*_ // * .. ) (( . \ / . * ) //
J.COOK@ENS.Prime.COM (08/06/90)
Re: Motion Blur and what we are imitating. I once read an article that mentioned "life imitating art imitating life." The big example was that early focal plane camera shutters (a moving slit) when used to take pictures of race cars, produced pictures that showed the car slanting forward. The effect had to do with the movement of the car while the exposure was being made. This was the art imitating life. The outgrowth (claimed by the author) was the BMW hood/grill. It rakes forward to resemble these photos. So life (the BMW hood/grill) imitates art (the early race car photos) imitates life (race cars). Amusing. J.Cook@ENS.Prime.COM "Just my recollections. Prime doesn't have a corporate recollection."
mgreen@cs.toronto.edu (Marc Green) (08/06/90)
>From: ritter@versatc.versatec.COM (Jack Ritter) >Are there sound physiological principles that support >the need for motion blur? My eyes didnt seem to >need it, at least not at the frame/sec rate he showed. >Maybe there's another factor: We're so used to seeing >motion blur all our lives, from film & tv cameras, that when >you take it away, we MISS it. I think I can answer your questions. First, the visual system has a mechanism which suppresses blur. The persistence characteristics of the retinal image should produce long streaks trailing any moving object - but it doesn't. There is some sort of inhibitory mechanism minimizing the streaking. The reason that bluring images reduces motion aliasing is that it removes high spatial frequency components from the image. The high frequency components have longer persistence, so they create longer trails and disrupt perception of smooth motion. There are numerous papers on this, including some of mine, in the vision literature. One last thing. This business about emulating a camera is a waste of time. The real issue is how the make the images match the capabilities of the visual system. Fixation with camera emulation gets in the way and obscures the real issues. Your question is a perfect example. >Well, I LIKED the stroboscopic one >BETTER than the motion blurred one. >It was simply crisper and more enjoyable than the >mushiness of the blurred version. Your terminology is wrong. Stroboscopic motion refers to any motion creeated by a sequenve of still images. There are many types of stroboscopic motion, which people seem to confuse. In particular people (even some experts) confuse beta motion and phi motion. Beta motion is stroboscopic motion which is indistinguishable from real motion; phi motion is "disembodied motion, where you feel as if the something moved, yet the images appear to flicker. Marc Green Trent University
dpawson@oracle.com (08/07/90)
Motion blur seems to be the kind of thing that is much more appropriate in some situations than others. Maybe you just looked at the wrong situation. In all seriousness, a class I took last semester demonstrated the value of motion blur using two films of rotary fans, one with and one without motion blur. This was certainly a case where it was appropriate, since the one without blur generated vicious aliasing, and just plain looked wrong. For another good example of where it can be used effectively, check out Pixar's Knickknack (sp?) When the snowman bashes at the wall with the hammer, there is some really sweet blurring on the hammer. Also when he gets shaken up by the jackhammer. Maybe it is just my preference, but I can't help but think the scenes would not be the same without the blur. Dave Pawson dpawson@oracle.com These views and opinions are entirely mine. I don't think Oracle has even established a corporate opinion on motion blur :-)
musgrave-forest@cs.yale.edu (F. Ken Musgrave) (08/13/90)
John Amanatides had some very intersting info on the appropriateness of motion blur in his presentation of his paper with Don Mitchell in Dallas last week. Presumably this is covered in the paper appearing in the proceedings. I found it quite surprising - I recommend having a look at the paper. Ken -- *===============================================================* F. Kenton ("Ken") Musgrave arpanet: musgrave-forest@yale.edu Yale U Depts of Math and CS (203) 432-4016