Interesting article on why 3D may never work (I happen to completely agree)

I follow Ebert's blog and came across that article last night.  It's pretty fascinating.  I'm not nearly as anti-3D as Ebert is, but I do think it's pretty much just a gimmick, and haven't seen many movies that were really improved by being in 3D.

But the information about how watching a 3D film differs from looking at an actual three-dimensional object--the difference between focus and convergence--is really interesting.  It's something I didn't realize before.

Thanks for sharing.

I Knew It!

I knew Roger Ebert is reading this forum! 

(Or perhaps it is Walter Murch  )

I posted about this more than a year ago. I remember there was someone who didn't believe me (found it: it's on the next page of that thread).

The problem with "3-D" (even the name is actually wrong, it's stereo-viewing) is that everything - literally - takes place in one plane only. There's nothing being projected or even remotely present (except light-waves passing through) in another plane.

This doesn't "prove" that 3-D movies are bound to fail in the long term, but to me it suggests that the current technique is not on par.

Cees

I saw a 3-D film once or twice  [Space Hunter and the Imax Galapagos Islands thing, from the projection booth.]  And numerous ViewMaster and Steriopticon thingies, too.  While they're a neat gimmick, they never struck me as being particularly realistic.  The stereo-slides and stereo-"cards" to me, always presented more like a shoe-box diorama, where different elements of the picture are cut out of paper, and glued into place "in space."  Not, for example, a "doll" standing in front of a mural, but a paper-doll standing in front of a mural. 

Something that came to me just as I was reading the Ebert-blog, and remembering Cees' post, was one of the most pro-3D people I know is also the person with the worst eye-sight I know.  His glasses make old glass Coke-bottles seem thin and delicate.  And his vision has been crap since he was born.

And he's wildly in favor of 3D. 

I, on the other hand, have always had excellent vision, and never thought much of it.

Granted, this is a sample-size of two.  Is there anyone else willing to compare their three-d experiences with their vision? 

Leo

Saw Tron:Legacy on Friday night in 3D. While the 3D of today is more advanced than yesteryear (and FTR, the image for Tron:Legacy was surprisingly bright compared not only to previous 3D presentations, but "flat" presentations as well; of course, I got there early enough to get my desired 3D "sweet" spot), it still possesses the inherent limitations of stereoscopic viewing, which at its worst is described as "cardboarding." I love 3D, but I think we will get to a point where audiences will begin refusing to pay the premium for a 3D movie, arguing that after 4+ years of premium pricing on 3D films, the equipment should be paid off by now. Conversions will not help this mindset.

Avatar *almost* convinced me 3D could become a permanent art form, and Resident Evil Afterlife *did* convince me.

If done well, it is slightly better than flat film.

And it wasn't even an action shot that convinced me, it was a scene in which Alice walks through a deserted airfield (or rather an overgrown field serving as a parking lot for small abandoned airplanes). All the reflections coming off the various planes' windows and chrome bits, reflecting Alice as she walked cautiously through this ominously quiet place, made for an absolutely astounding 3D experience. Probably because if those reflections had all been computer generated with fake 3D, it would have taken weeks of processing to get it right -- or rather, no one would have bothered to get it right. But shooting it with two cameras from the get-go resulted in an image that was beyond our present technology to fake. It was astounding, and simple, all at the same time.

It just seems like the work-to-benefit ratio is kind of difficult right now. The cinematographer has to be so much more knowledgeable and careful...

Murch's critique of 3D is very well done, and I agree with him.  In addition to the focus-convergence issue, I agree with his comments about darkness and reduction of scope.

Quote:

Originally Posted by Leo Kerr 

I saw a 3-D film once or twice  [Space Hunter and the Imax Galapagos Islands thing, from the projection booth.]  And numerous ViewMaster and Steriopticon thingies, too.  While they're a neat gimmick, they never struck me as being particularly realistic.  The stereo-slides and stereo-"cards" to me, always presented more like a shoe-box diorama, where different elements of the picture are cut out of paper, and glued into place "in space."  Not, for example, a "doll" standing in front of a mural, but a paper-doll standing in front of a mural. 

 

Something that came to me just as I was reading the Ebert-blog, and remembering Cees' post, was one of the most pro-3D people I know is also the person with the worst eye-sight I know.  His glasses make old glass Coke-bottles seem thin and delicate.  And his vision has been crap since he was born.

 

And he's wildly in favor of 3D. 

 

I, on the other hand, have always had excellent vision, and never thought much of it.

 

Granted, this is a sample-size of two.  Is there anyone else willing to compare their three-d experiences with their vision? 

 

Leo

Viewmaster unfortunately did a lot of "fake" 3D reels just to have movie tie ins. (E.T. was a good example of the paper doll cutout effect.) The reels shot with their 3D cameras are the real thing. Pixar seems to have always provided them with true 3D renderings from their films.

I have not been impressed with the few 3D films I have seen, nor with the 3D TV demonstrations at the stores. However, I did find this one comment by Murch in the linked article interesting:

"Consequently, the editing of 3D films cannot be as rapid as for 2D films, because of this shifting of convergence: it takes a number of milliseconds for the brain/eye to "get" what the space of each shot is and adjust."

If one of the outcomes of editing films for 3D is that Hollywood gets way from the current style of overly rapid quick cut editing, I am all for it. The last Bond film Quantum of Solace, for an extreme example, gave me a headache trying to watch in that editing style.

I've heard a lot about serious medical concerns over the long-term effects of more frequent 3D viewing.  Murch's comments about the plane of convergence make me think that's the point that a lot of doctors are expressing concern over.  What real scientific evidence can anyone have at this point though about what it can do to your vision in the long-term?

Seems like the convergence issue is only a concern for 3D movies that have the gimmick of things "flying out of the screen". Though these objects seem to be near to us, and therefore would seem to require us to shift our eyes to a slightly cross-eyed position, we do not need to -- that would be a waste of effort. So it causes some fatigue.

But if we're looking into a "deep" 3D picture, the technique used in Avatar and Resident Evil (in which things go not "fly out" but rather extend back into the distance), our eyes are already pretty much staring straight ahead into the infinite distance already.

No doubt someone needs to hook up some viewers to machines to measure how their eyes are moving during 3D films, to get a better understanding of whether people's eye muscles are getting an unneeded workout. But even if they are, I doubt this is a problem that "many doctors" would care about -- after all, what about the long term damage caused by watching movies in black and white? It is unnatural, but we do just fine. What about 2D films? Also unnatural. Our mind expects us to need to adjust our focus, but some camera operator already did it for us!

Quote:

Originally Posted by Will_B 

Seems like the convergence issue is only a concern for 3D movies that have the gimmick of things "flying out of the screen". Though these objects seem to be near to us, and therefore would seem to require us to shift our eyes to a slightly cross-eyed position, we do not need to -- that would be a waste of effort. So it causes some fatigue.

 

But if we're looking into a "deep" 3D picture, the technique used in Avatar and Resident Evil (in which things go not "fly out" but rather extend back into the distance), our eyes are already pretty much staring straight ahead into the infinite distance already.

 

No doubt someone needs to hook up some viewers to machines to measure how their eyes are moving during 3D films, to get a better understanding of whether people's eye muscles are getting an unneeded workout. But even if they are, I doubt this is a problem that "many doctors" would care about -- after all, what about the long term damage caused by watching movies in black and white? It is unnatural, but we do just fine. What about 2D films? Also unnatural. Our mind expects us to need to adjust our focus, but some camera operator already did it for us!

 

You are still converging on a point beyond where you are focusing. The focus plain being the surface of the screen, and the convergence point being some distance past it.

Black and White doesn't force our eyes to do something they don't do naturally. Nor do films themselves. Seeing motion through persistence of vision is a normal human process. Now if the frame rate were too slow, that would be fatiguing and again people would get head aches. The camera changing focus has no effect on what our eyes are doing, which is focusing on the screen. Again nothing unnatural there.

Doug

This all sounds like the concerns back in the 50's about staring at little flickering cathode ray tubes too long. How long before some unscrupulous individual sues a studio claiming little Suzie has gone cross eyed from watching a 3D film? I imagine there will be a lot of psuedo-scientific baloney spouted pro and con in the meantime. I've already heard the suggestion somewhere that a moviegoer driving home right after seeing a 3D film might be "visually impaired".

I'm more worried about those supermarket checkout scanners flashing in my eyes than any prolonged 3D viewing.

I don't argue with Walter Murch's reasoning, but I respectfully disagree with his conclusion.

His experience on Captain Eo may be discounted because that film was limited in its technology and does not represent the technology available today.

Roger Ebert is confused and irrational, so I don't bother with him at all.

Filmmakers need to learn how to shoot and edit in 3-D. It is a discipline that needs to be practiced and learned. Then you can avoid the mistakes detailed by Walter Murch. The problem is few filmmakers take the trouble to learn the discipline. They shoot 3-D before they fully understand what it does and doesn't do. I include Robert Zemeckis and James Cameron among them (although they're getting better at it). When properly practiced, 3-D is not too small, too dark, nor too at odds with focus and convergence. Properly shot and edited 3-D leads the eye gently and naturally into depth and is smooth as silk.

Standard flat films are made in a different syntax that is at odds with 3-D should never be converted because they can't possibly work in 3-D. It brings 3-D into discredit. This trend has to stop.

Premium pricing is huckstering and should also be stopped.

I heard someone say.. and I'm being vague, here, 'cause it was second hand, and I can't remember who was being attributed...

But in short, when you're piecing together a 3D film, you need to make sure that from shot-to-shot, you aren't "jumping" too far in apparent depth.  Or that the shot itself changes, allowing the eye to "adjust" to the "location" in depth.  Yo-yo-ing too much is a guaranteed way of making the audience sick.

Which is, in itself, a perfectly fine and understandable statement, and probably mostly true.

The challenge then becomes, how do you produce this film?  Normally, the film is written, storyboarded, shot, edited, and so on.  And the finished film may look nothing like the storyboards.  But if you're storyboarding with the "depth" element, and shoot for that, and then the editor says, "we need to do this," but the depth-cues are all wrong, do you re-shoot?  Screw the audience?

Or shoot 2-D and do a conversion?

The argument was it was generally easier, and gave better results, to do a 2-D conversion than it was to do a proper 3-D shot, 'cause then when you're doing the conversion, you could control the depth-cues, without having to re-shoot.

Leo

Roger Ebert has hated 3-D for decades. There's been just a few 3-D titles where he's praised the 3-D efforts. But when the next 3-D film rolls out that he dislikes, it's then back to bashing the process entirely.  I'm pretty sure the World 3-D Film Expo (I don't remember if it was the first or second Expo) tried to get him to attend, just to be able to have him sample the Golden Age 3-D titles of the 1950s, but he never showed. Leonard Maltin attended, had a wonderful time, and has written a lot of positive comments on the process.  I saw over 40 hours of 3-D features in ten days and never suffered eyestrain.

There's a few areas where current stereoscopic shot features can be improved on in the theatrical chain. The main one in my book is light output. This is one of a few advantages of dual projection IMAX 3-D, when compared to single projection Real-D or Dolby.

Regarding light output, digital capture and digital projection is too dim for 3-D. I find it too dim for flat films, and my vision is fine (excellent in fact, I don't need to wear glasses unless I'm reading). Of course, the luminosity of a monitor / home display compensates. Digital capture looks much better at home than on a theater's flat matte screen.A luminous monitor is the only way to watch 3-D captured digitally, whether the industry practitioners admit to it or not. A lot of them are lying to themselves when they say these million-dollar digital projectors are up to industry standards. They're not, and that's that.

Photochemical 3-D capture looks fine when the celluloid is projected on an aluminized screen, and it looks fine when transferred to digital medium and played through a luminous monitor.

Two rules of thumb: project theatrical features on aluminized (silver) screens to get proper 3-D, and make sure the light is cranked up to industry standard.

Quote:

Originally Posted by Leo Kerr 

I heard someone say.. and I'm being vague, here, 'cause it was second hand, and I can't remember who was being attributed...

 

But in short, when you're piecing together a 3D film, you need to make sure that from shot-to-shot, you aren't "jumping" too far in apparent depth.  Or that the shot itself changes, allowing the eye to "adjust" to the "location" in depth.  Yo-yo-ing too much is a guaranteed way of making the audience sick.

 

Which is, in itself, a perfectly fine and understandable statement, and probably mostly true.

 

The challenge then becomes, how do you produce this film?  Normally, the film is written, storyboarded, shot, edited, and so on.  And the finished film may look nothing like the storyboards.  But if you're storyboarding with the "depth" element, and shoot for that, and then the editor says, "we need to do this," but the depth-cues are all wrong, do you re-shoot?  Screw the audience?

 

Or shoot 2-D and do a conversion?

 

The argument was it was generally easier, and gave better results, to do a 2-D conversion than it was to do a proper 3-D shot, 'cause then when you're doing the conversion, you could control the depth-cues, without having to re-shoot.

 

Leo

A storyboard drawn in 3-D with a diagrammed camera set-up to match it is a necessary step when shooting in 3-D. You plot your convergence together with your focal length the way you plot your angles and motion. There is a way to lead the eye from one convergence to another, in fact, you can lead the eye out of one convergence and into another in the same shot, without cutting to another shot, but you have to plan it, and time it, in relation to the shots that come before and after so that the eye is led easily in. This is too much homework for some filmmakers. You have to plan and plot your shots like Hitchcock did, only thinking in terms of 3-D all the time.

Never, never shoot 2-d and then convert to 3-D. Never. If you have to shoot in 2-d for the purpose of converting to 3-D you are doing it wrong.

Split-second cutting doesn't work in 3-D -- many people will argue it doesn't work in 2-d films -- because the human brain needs at least three-seconds to absorb a shot, and some people need longer than that. Personally I would not cut faster than 3-5 seconds in 3-D and only then under unusual circumstances. Most viewers find a seven-second shot is a very satisfying experience in 3-D.

Kind of a silly article and conclusion, regardless of Murch's pedigree.

3D doesn't work and will never work...sure, except for all the people for which it already does.

Ebert's anti-3D agenda is petty at best

Quote:

Originally Posted by Douglas Monce 

You are still converging on a point beyond where you are focusing. The focus plain being the surface of the screen, and the convergence point being some distance past it.

I'm suggesting that at a certain distance, there is no more "convergence" of the eyes taking place anymore. I don't know where the cut off is, but clearly there's no difference in convergence between, say, a hundred feet and a mile. It's all pretty much looking straight ahead, eyes parallel.

So the question is, where is this spot when convergence is no longer relevant? In a big movie theater it might not be a problem. In a living room with a 3D tv close by, it may be.

Quote:

Originally Posted by Douglas Monce 

 

Black and White doesn't force our eyes to do something they don't do naturally. Nor do films themselves. Seeing motion through persistence of vision is a normal human process. Now if the frame rate were too slow, that would be fatiguing and again people would get head aches. The camera changing focus has no effect on what our eyes are doing, which is focusing on the screen. Again nothing unnatural there.

 

 

Doug

But it is unnatural. You'd expect our brains would be trying to fill in colors when it sees a black and white basket of fruit. And when a camera is focused on something close up, and the background is thrown into a nice bokeh of blurriness, there should be a conflict in our eyes and brain if we try to "focus" that background -- which we cannot do. Yet somehow we accept that we cannot pull it into focus, and we are not fatigued. It would be interesting to know if people adapted, or if there was never any fatigue to begin with.

Quote:

Originally Posted by Richard--W 

His experience on Captain Eo may be discounted because that film was limited in its technology and does not represent the technology available today.

In reality he was working under the best possible conditions, even better than 3D today. First of all Captain EO was shot in 65mm, meaning higher resolution and much more light output from the projectors. Secondly the film was shown in a theater that was designed and built only to show that film. Again the best possible conditions. I must say that Captain EO was by far the best 3D I've ever seen, and even by the end of that film, I had a head ache.

Doug

Quote:

Originally Posted by Will_B 

But it is unnatural. You'd expect our brains would be trying to fill in colors when it sees a black and white basket of fruit. And when a camera is focused on something close up, and the background is thrown into a nice bokeh of blurriness, there should be a conflict in our eyes and brain if we try to "focus" that background -- which we cannot do. Yet somehow we accept that we cannot pull it into focus, and we are not fatigued. It would be interesting to know if people adapted, or if there was never any fatigue to begin with.

The brain doesn't work that way.

Doug

Quote:

Originally Posted by Douglas Monce 

Quote:

Originally Posted by Richard--W 

His experience on Captain Eo may be discounted because that film was limited in its technology and does not represent the technology available today.

In reality he was working under the best possible conditions, even better than 3D today. First of all Captain EO was shot in 65mm, meaning higher resolution and much more light output from the projectors. Secondly the film was shown in a theater that was designed and built only to show that film. Again the best possible conditions. I must say that Captain EO was by far the best 3D I've ever seen, and even by the end of that film, I had a head ache.

 

Doug

You make a valid point about the best of conditions. I saw it a few times as well. It's a good 3-D film but by no means the best I've ever seen. The sections shot with the old SpaceVision lens were the most impressive for depth. Different methods of capture were implemented, and they were not consistent, according to one of the 3-D engineers on the film who works for me. As I recall it was only 18 minutes long? There were some jarring moments, but it did not induce a headache or eyestrain in this viewer. After 20 + years I would like to see it again.

Several of the best -- the best as in the most submersive experience -- stereoscopic films are documentary shorts (45-52 minutes) by Imax, which, like Captain Eo, are in 65mm.

3-D is a deep-focus discipline. Do not blur the backgrounds. Do not cut from a blurred background to a focused background. The background should be in focus during every shot, whether it be a close-up or a long angle. There are other ways of leading the viewers' eye into the frame, or toward the subject of the frame, without blurring backgrounds. To blur the background in a stereoscopic film is to induce eyestrain, eye-jarring, headaches, and to screw up.

Depends on the shot. It will be the same shot whether you sitting in a theater or watching at home. Eventually, long distance shots flatten out in 3-D. You can set the lenses to infinity just like in 2-d, but the depth flattens out in the distance just like 2-d. The point is to maintain depth where it matters. Using matched lenses on two interlocked cameras facilitates the greatest depth and provides the most latitude in convergence. It is possible to adjust the convergence during the shot, leading the eye toward or away from the distance. You can accomplish almost any shot that requires a long-distance focus. For example, that famous shot in Lawrence of Arabia of Omar Sharif riding the toward the camera over a great distance could be accomplished nicely in 3-D, but you would perceive the most depth the closer he gets, and the least depth the further back he is, even if you are adjusting convergence as he travels. I'm sure the floating-on-the-haze effect would be palpably enhanced. I'd love to try a shot like that in 3-D.

Shall we tell the truth about 3-D.  We appear to be beating around the bush, skirting around the edges, and not getting to the core of the matter.

3-D doesn't work.

Let me clarify that, because you might bethinking "Hold on, I don't like 3-D, but I see things flying off the screen. It'd definitely 3-D and it definitely works".

Okay, what is 3-D for?  Is 3-D designed for 3-D's own sake?  No, not at all.  The whole idea is that real life is 3-D.  We have stereoscopic vision, we see slightly different things with each eye, and that gives us a sense of depth.  So the whole idea behind 3-D is that it is supposed to be more like what we see in real life.

And it isn't.

If I watch a 3-D film, yes it's 3-D.  But no, it doesn't look more like the real world.  If anything it looks less natural than 2-D.  I don't ever remember watching a 2-D film and feeling it didn't look natural, or that the image was lacking real world depth.  Not once.  Not for a second.  Not even a hint of it.  But watching a 3-D film it just looks like a series of layers placed on top of each other.  Real life doesn't have that 'layered' appearence.

And that's the real problem.  Even though the real world is in 3-D, films made in 3-D is less like real vision than 2-D.

Steve W

Quote:

Originally Posted by urbo73 

Quote:

Originally Posted by GregK 

Roger Ebert has hated 3-D for decades. There's been just a few 3-D titles where he's praised the 3-D efforts. But when the next 3-D film rolls out that he dislikes, it's then back to bashing the process entirely.  I'm pretty sure the World 3-D Film Expo (I don't remember if it was the first or second Expo) tried to get him to attend, just to be able to have him sample the Golden Age 3-D titles of the 1950s, but he never showed. Leonard Maltin attended, had a wonderful time, and has written a lot of positive comments on the process.  I saw over 40 hours of 3-D features in ten days and never suffered eyestrain.

 

There's a few areas where current stereoscopic shot features can be improved on in the theatrical chain. The main one in my book is light output. This is one of a few advantages of dual projection IMAX 3-D, when compared to single projection Real-D or Dolby.

 

I wonder why he's not liked it....Hmm.. On the other hand, is there something Leonard Maltin doesn't like? I can't even begin to compare the two...Ebert is a critic. Maltin is commercialism. Just because YOU didn't suffer eyestrain doesn't mean anything. There are plenty that do. Explain that. Until you can, then 3D is indeed a failure. I knew one kid years ago who played these RPG games and claimed they never bothered him. All that fake up/down/sideways movement that is not real. Then he started getting tired, restless, etc. Ultimately he couldn't play them anymore without experiencing fatigue and headaches. So be careful. While some are sensitive right off the bat and others may be fine, in the end its unhealthy. It just makes common sense.

The average person is going to sit in front of an RPG game a lot longer than they would for a 3D movie. And there are flat films that have given people headaches and nausea, Blair Witch Project for example.

I can't read in a moving vehicle without dizziness and headaches, therefore automobile travel is a failure.

Quote:

Originally Posted by Yorkshire 

Shall we tell the truth about 3-D.  We appear to be beating around the bush, skirting around the edges, and not getting to the core of the matter.

 

3-D doesn't work.

 

Let me clarify that, because you might bethinking "Hold on, I don't like 3-D, but I see things flying off the screen. It'd definitely 3-D and it definitely works".

 

Okay, what is 3-D for?  Is 3-D designed for 3-D's own sake?  No, not at all.  The whole idea is that real life is 3-D.  We have stereoscopic vision, we see slightly different things with each eye, and that gives us a sense of depth.  So the whole idea behind 3-D is that it is supposed to be more like what we see in real life.

 

And it isn't.

 

If I watch a 3-D film, yes it's 3-D.  But no, it doesn't look more like the real world.  If anything it looks less natural than 2-D.  I don't ever remember watching a 2-D film and feeling it didn't look natural, or that the image was lacking real world depth.  Not once.  Not for a second.  Not even a hint of it.  But watching a 3-D film it just looks like a series of layers placed on top of each other.  Real life doesn't have that 'layered' appearence.

 

And that's the real problem.  Even though the real world is in 3-D, films made in 3-D is less like real vision than 2-D.

 

Steve W

That's exactly my feeling. It's like watching a moving pop-up book.