Is 3D dying

[Alec Morris]

Ivan

Unfortunately, you are certainly not the only person who has that reaction. The worst part is this is actually completely avoidable! (stay with me here, I'll try to explain this)

It starts with what we call stereo budget, which (in technical terms) is the maximum amount of parallax we allow and is calculated in relation to screen size or viewing distance and the amount of divergence and convergence we allow.

In regular words, that basically means this: 'Convergence' is when you go cross eyed, and 'divergence' is when you go bug eyed. When you focus (in real life) on an object thats far away, your eyes move outward, and its the opposite when things are close. You can experience this by putting your finger a few feet away and focusing on it as you slowly move it in to touch your nose.

Now, to illustrate "parallax" put your finger a foot in front of your face. Stay focused on it, but close one eye. Now, switch eyes. You will notice your finger "jumps" in relation to the background. There is a difference in where you see your finger in your left eye and your right eye in relation to whatever is behind it. The difference between these two pictures, when overlaid on each other, is calculated by your brain as depth, and is what we refer to in 3D as parallax.

The stereo process is making a computer act as our brain and two cameras act as our eyes. We can adjust it by moving the distance between the cameras and by adjusting the images in the computer (as long as we shoot with the cameras completely parallel). Now, once we have the DI finished, it's locked. So the distance between the two images isn't going to change.

But here's the problem. The reason 3D works is that our eyes are processing the distances between two different images on the screen (one through each polarized lens in the glasses). The distance between our eyes never changes. But guess what? The distance we VIEW them from does. So take your finger again and place it a foot in front of your face. Now move "closer to the screen" (move your face about two inches away from your finger while continuing to focus on it).

Hurts doesn't it? Guess what most people do when they walk into a theater? They get seats as close to the screen as they can. When you do that, you are literally going crosseyed and bug-eyed, which HURTS! Theaters should do away with seats that are within a certain distance of the screen, but that obviously wont ever happen.

You know how you can spot a stereographer at a screening? They run for the back of the theater.

So, in a lot of cramped theaters, this is a problem.

The second and much bigger problem lies in editing. It takes your brain three seconds to process depth. So, if I am editing something in 3D, each cut MUST be at least three seconds long, or I must cut to a shot of equal depth with the point of focus in a similar place. Otherwise, my brain is forced to adjust to a new image before it has processed the old one. Guess which editors and directors follow this rule. Almost none. I see it all the time. I read an article today on a major release that is about to come out and was post converted, in which the editors said they didn't edit in 3D at all.

And then of course there are the crappy dim projectors.

If all director, editors, and cinematographers that are doing a 3D film would just take a weekend retreat to learn this stuff before they shoot, we wouldn't have these problems. You would think, with a large percentage of the world viewing their work in 3D, they would take the time. But no. They just treat it like 2D and let the post guys worry about it.

Can't wait for Avatar 2 though, because Cameron knows more about 3D than anyone else on the planet. He had to blaze a lot of trails and figure stuff out as he went on the first film. Part 2 is gonna show you what 3D can really do.

Oh and that's the whole point of High Frame Rate 3D btw. The more frames there are, the easier it is on your brain. I have shot HFR3D, and I think it looks incredible. (It certainly takes some getting used to though, I though it looked bizarre for the first five hours or so of shooting. Then I loved it). I think the problem people had with the hobbit was that it is a world they are already familiar with from the first three films, and not only were they adjusting to HFR, but 3D and digital as well. That's a lot to change when we've all spent 8+ hours in middle earth with the first films.

 

[Alec Morris]

Sorry for the length of the post, just hoping if I can explain a bit of the tech and terminology we can all have a real discussion.

Ivan, to your other point about only some things being 'appropriate' for 3D, I think after a few years you might change your mind. Actually, I had a producer that worked on something I shot in 3D that raved about how much he loved it. Then later, we were talking about the great gatsby and he said the 3D didn't work for him with that film.

It made me really reexamine why I like it so much, as I though he was 100% gung-ho on 3D. And I remembered a time I had been locked in a room editing in 3D for like four days when I decided to go see a movie to clear my head. It was in 2D, and after viewing 3D for so long, the 2D looked bizarre. I was conscious of how much work our brains do to turn a 2D image into what we recognize as reality (and obviously we experience reality in 3D). It kind of clicked with me then and there that I had seen enough 3D that my brain stopped trying to process the depth. As in, I didn't notice when I was watching 3D anymore, it just seemed 'natural.' Then everything seemed appropriate for 3D because 3D felt natural.

So, I guess my advice is lock yourself in a room for four days and watch nothing but 3D

 

[Ivan Kovax]

Wow. Great answer.

With HFR 3D - it took me about 10 minutes for my brain to get up to speed in the Hobbit. The dialogue looked out of sync. And the characters seemed to be walking around at a different framerate... 20FPS... like a charlie chaplin movie. There was some serious neurological clusterfucking going on in those 10 minutes. My question is - will this happen everytime I go to see HFR in 3D? I disliked the hobbit for many other reasons too, which are have probably been discussed ad nauseum on other threads (I am sure you know what I would have said... heh).

I generally go for the middle of the theater which is what I have heard most stereographers suggest as being the sweet spot. It was once explained to me by a stereographer on a 3D Coke commercial I worked on.

I would actually like to see somebody do a neurological study on HFR 3D with an MRI to see the effects on the human mind. If there is a way to escape the hyper-realism that the hobbit created then I am all for that. I think the Hobbit failed psychologically on many levels with their HFR experiment, but I am eager to see how Cameron takes a different approach at 60FPS.

One interesting study placed a plasma TV with an HD image of a male octopus next to a physical fishtank containing a real male octopus.

They played the video of the male octopus on the plasma at 24hz (or equivalent) yeh the real octopus didn't react at all. Didn;t even see it.

WHen played at 60hz, the real octopus freaked the shit out and started attacking the plasma screen.

I don't know if this is a good thing or a bad thing

 

[Sergio Perez]

Avatar, Hugo, The Great Gatsby where the films in the middle of hundreds I've seen in this few years of post Avatar 3D madness that made the experience worthwhile. As a filmmaker and communicator I fail to see the necessity of 3D in order to make a great, commercialy successful film. Its still a good marketing hook, but not "the future". That will be 8K Projection in my view. And HFR sucks.

 

[Alec Morris]

From my experience, yes you will get used to HFR. However, it is only after a significant amount of viewing, which is problematic.

If the theaters do a 'marathon' screening of the first two hobbit films before a midnight screening of the third film, by the time the third one starts I am confident you will be used to it.

I would tell you to watch the bluray a few times, but unfortunately bluray3D can only be mastered at 24fps and 60fps (the hobbit was 48fps, so the bluray doesn't have an HFR version)

I have spoken with some people that have worked with Douglas Trumble on his "Digital Showscan". They said he has some studies that showed the brain responding significantly to HFR 3D (he's doing 4K 120fps)

 

[Alec Morris]

Honestly it's hard enough to tell the difference between 2K and 4K. As cool as 8K may be on massive screens, it probably won't really make a noticeable difference on smaller screens.

High Dynamic Range, High bit depth color, Glasses free 60fps 3D on an ultra-bright laser 4K projector (or 8K if you REALLY have the screen size) with Dolby Atmos surround seems like a hell of a future to me.

 

[Aaron McLane]

We did a study a few years ago and discovered that dogs, with glasses, could actually see S3D perfectly....... in a related study we also discovered that only dogs and people with a financial interest in S3D can actually enjoy S3D but even then only for a maximum of 2 min.

 

[Aaron McLane]

we did discover a good use for 3D rigs though..... about 21 stops of DR.....true motion HDR

 

[Alec Morris]

Thanks aaron, that was incredibly helpful.

 

[Ivan Kovax]

From my experience, yes you will get used to HFR. However, it is only after a significant amount of viewing, which is problematic.

...

I have spoken with some people that have worked with Douglas Trumble on his "Digital Showscan". They said he has some studies that showed the brain responding significantly to HFR 3D (he's doing 4K 120fps)

Yeh, it is problematic. If one risks a 10 minute freak out period at the start of every 3D HFR movie then this needs to be reexamined.

I have heard of Douglas Trumble's Digital showscan system (I am not sure what the difference between 120 and 60 would be though. Apparently the brain caps out ar around 45 or so (i forget the number they settled on)... wouldn't everything beyond that be the same perceptually?

I would like to read up on these studies he is doing at 120fps. When you say significant response, what do you mean in particular?

 

[Alec Morris]

I can't remember exactly, but it was something along the lines of what you are talking about.

They were hooked up to machines that read brain activity and responded most to 120fps. I think they measured heart rate too. I'll see if I can find an article on it.

 

[Ivan Kovax]

That sounds insane. From what I understand Douglas is also experimenting with variable frame rates within the same movie... if that can be perfected, that will be amazingly useful...

 

[Hrvoje Simic]

I am not sure I completely understand your conclusion. Are you saying that because there is a natural tendency for the eyes to wander more in 3D, the director/creator has to place even stronger emphasis (technically) to focus the viewers attention on the subject than one would in 2D?

Yes.

 

[Hrvoje Simic]

Alec, just a few additons to your findings, to avoid generalizations and misinterpretations:

In regular words, that basically means this: 'Convergence' is when you go cross eyed, and 'divergence' is when you go bug eyed.

It would be more precise to perceive convergence as rotational movement of the eyes with vergence point going inward and divergence as vergence point moving outward. Convergence- movement towards joining, Divergence - movement towards separation (of viewing direction axis). Cross-eyed and bug-eyed would be the extreme points passing the limits or natural eye movement.

Also, it would help informing both the viewers and the content creators that enjoying the object floating in front of the screen necessitates doing precisely what their parents told them not to do, i.e. forcing cross eyed viewing.

stereo process is making a computer act as our brain and two cameras act as our eyes. We can adjust it by moving the distance between the cameras and by adjusting the images in the computer (as long as we shoot with the cameras completely parallel).

It should be stressed that by adjusting IO in post one is simply moving two flat images without ability to affect a natural parallax. Which leads to other issues.

second and much bigger problem lies in editing. It takes your brain three seconds to process depth. So, if I am editing something in 3D, each cut MUST be at least three seconds long, or I must cut to a shot of equal depth with the point of focus in a similar place.
.

It is beneficial to perceive these 3 secs as a very general orientation point, as mind is adaptable and perception varies depending on the circumstances, in this case affected by picture or content properties and viewing setup and environment. Few examples would be obscure scenes and exaggerated DOF.

 

[Mike Garrick]

Alec as a dummy in 3D. I found your description of 3D editing and the viewers reaction to it insightful.

So for the dummy's the shortest shot duration should be 3 Secs minimum in 3D land to give a chance for the viewers brain to re-calibrate?

 

[Aaron McLane]

Hrvoje, are you saying convergence is moving the distance between your eyes in and out? Most humans cannot do that.....but humans can converge their eyes on a closer target by going a little cross eyed.

Also it is impossible to change the IOD or IA in post regardless if you shoot converged or parallel. You can change the convergence in post if you shoot parallel or converged, although if you shoot parallel you must converge in post which does cost you up to 10% of your sides where if you shoot converged you only loose 4% of your sides which is why on 1080p resolution limited cameras they always shoot converged and on Red cams people almost always shoot parallel.

Many people involved with S3D still confuse IOD (Interocular Distance), which is the distance your eyes or two cameras are placed apart and can't be changed in post, to HIT (Horizontal Image Translation), which can be done in post to change or create convergence. I know some programs incorrectly state that you can change IOD when they mean convergence but it's a matter of physics and not possible.

But back to the point of this thread, here are a few ways to tell that S3D is dead or dying.

Directors that have shot successful well done 3D movies, Marty HUGO, Ang PI, Ridley PROMETHEUS have stated they will not do another 3D movie. CHECK

Directors (JJ STAR TREK, Shane IRONMAN, Zak SUPERMAN) that have had bad 3D conversions forced on their movies that they intentionally shot as 2D as possible (anamorphic, on film, quick cutting, no 3d framing) and who advise audience members to see in 2D. CHECK

Exit polls showing that aprox 50% who see 3D movies did so because that was the only/most convenient screening they could get into and would prefer to have gone to a 2d screening. CHECK

This past NAB which had almost no 3D presence compared to the last two years which had a massive presence. CHECK

 

[Hrvoje Simic]

Hrvoje, are you saying convergence is moving the distance between your eyes in and out?

Nope. That wouldn't make sense. Rotational movement of the eyes around their center towards one another.
The distance which changes would be the one between the corneas. Post edited for clarity.

You can change the convergence in post if you shoot parallel or converged, although if you shoot parallel you must converge in post which does cost you up to 10% of your sides where if you shoot converged you only loose 4% of your sides which is why on 1080p resolution limited cameras they always shoot converged and on Red cams people almost always shoot parallel.

I would suggest optimizing any 3D settings on set and leaving it out of post. The subtle differences which some may not perceive consciously, relying on post work to compensate for shot inadequacies, are still perceived as unnatural by subconscious, which is vastly more sensitive to imperfections.

But back to the point of this thread, here are a few ways to tell that S3D is dead or dying.

Directors that have shot successful well done 3D movies, Marty HUGO, Ang PI, Ridley PROMETHEUS have stated they will not do another 3D movie. CHECK

Directors (JJ STAR TREK, Shane IRONMAN, Zak SUPERMAN) that have had bad 3D conversions forced on their movies that they intentionally shot as 2D as possible (anamorphic, on film, quick cutting, no 3d framing) and who advise audience members to see in 2D. CHECK

Exit polls showing that aprox 50% who see 3D movies did so because that was the only/most convenient screening they could get into and would prefer to have gone to a 2d screening. CHECK

This past NAB which had almost no 3D presence compared to the last two years which had a massive presence. CHECK

I see these as valid points, but related to this type of approach to 3D storytelling.
Which is flawed in many ways.

Directors which made successful 3D movies have been learning and perfecting their skills to express on canvas all their life. And in this medium they have to throw half of the highly attuned skills away and learn to express within new boundaries. Forced conversions of 2D movies are pretty much self explanatory and a consequence of greed overshadowing reason. Viewers' reaction is natural since they have been watching canvas stories pushed in space. NAB presence is a bounce back of overeagerly forced technology while the perception and utilization of the medium it depends on is still very much immature in most cases. I would say.

This medium is very specific yet very generalized so far and pushed where it doesn't belong. Storytelling potential it offers differs significantly, it cannot substitute canvas as it was pushed to do. It is another creative realm. Offering new things with a cost of trading them for a lot of others. So far the storytellers' intent has been to gain without the willingness to trade.

If this medium is to be mastered, now that technology allows much more practical access to it, it will require many years of individual and joint effort of storytellers to develop sensibility and expertise to perfect the craft. Anyone who expects great results within few years of experience in this medium has unrealistic expectations. Pretty much as in any medium.

 

[Simon Dunne]

BBC to suspend 3D Broadcasts

BBC to suspend 3D Broadcasts

http://www.bbc.co.uk/news/entertainment-arts-23195479

 

[Dave Blackham]

We enjoy our 3D work but then we also enjoy our work in 5k, Anamorphic, for internet, television and Cinema. 3D is to us is just one of the formats which clients wish to acquire and deliver. Its sad if its taking a while to get momentum and Im not sure the big boys dropping the ball will move us forward. The fact is we can do it if we are required to and we can do it well, especially with Epic.

 

[Alec Morris]

Hrvoje, thanks for posting and going more in depth. I was just trying to keep that post simple.

Mike, yes the shot should, as a rule of thumb, be 3 seconds. UNLESS you are careful enough in your previz to exactly match the depth and point of focus between two shots you plan to cut more rapidly. I have rarely seen this done, but every time I do I smile because it works very well.

Aaron, the reason there wasn't a 3D presence at NAB was because the market is already saturated. 3D is expensive, and there are already enough rigs in the marketplace to satiate the needs of productions that can afford to shoot in 3D. This has been covered before in much more depth.

To say you can't adjust IA in post is an argument in semantics. Thats like saying you can't change the zoom of a lens in post. While technically true, every day people are cropping and applying zooms in post production. Up to 10%? Again, while technically true I guess, that is an insanely inflated percentage in practice.

I could find absolutely nothing from ang lee other than his absolute adoration of 3D. As for scott, I read he felt the counselor wasn't appropriate for 3D, but that prometheus 2 would be shot in 3D. I can't find anything from scorcese saying one way or the other in the last few months (actually, everything I read from him before suggested he thought it would be a huge part of his future, but the wolf of wall street is 2D. again, that feels like a '2D' movie to me) I also notice you conveniently left our Peter Jackson and James Cameron.

Of course guys that shot for 2D are gonna say 'see it in 2D'. They also can only hurt their reputation by saying otherwise, because the 3D experience is subpar. They are already established directors, why would they care to take the time to learn 3D if they didn't have to? (cameron, lee, scorcese, scott, and jackson all did because they see the beauty, but I can understand why an established director might not want to put in the effort)

please link to your "exit poll". Something tells me it's a bullshit poll.

Simon, the reason 3DTV has been such a shitshow is that there was so little available content. They asked people to pay thousands of dollars for TV's with only a few pairs of glasses and then made it incredibly difficult to get any content for it. There are a few blurays, but the TV programming was a joke. Of course people weren't going to jump on the bandwagon. Do you know why HD was so successful? Guys could buy a new TV, have their buddies over for a beer, and put it on HDNet, which showed how cool HDTV was. They could actually watch HD channels, and all their friends then wanted to be cool HDTV owners. Sports were huge for HD. ESPN3D cost extra and wasn't even available from a lot of subscribers/ Plus, you had to get a box to make it work! I have a 3DTV and didn't want to do that. (can't say anything to BBC3D as I am in America, although I just watched the two seasons of Black Mirror and they were incredible)

dave, i definitely agree. I have the hobbit 3D bluray, but I watched it the other night in bed on my iPad. A good movie is a good movie.