What are your guidelines for positive parallax %?

[Phil Bates]

For cinema is it 1%, 1.5% or more? Do you go up to 5% or more for TV?

If it is for a few isolated shots, do you push these numbers?

I ask because I am hearing different numbers from different sources. It's a little confusing out there.

Thanks,
Phil

 

[Pedro Guimaraes]

Discovery 3DTV specs./requirements lists 3% total. 2% positive and only 1% negative. They are also allowing and additional 1% negative for titles.

Specs are specs. If your given them follow them of course.....if your not.....experiment! Each setup is different, lighting, colors, contrast these and other factors could all affect how much you can push those numbers.

 

[Phil Bates]

Discovery 3DTV specs./requirements lists 3% total. 2% positive and only 1% negative. They are also allowing and additional 1% negative for titles.

Specs are specs. If your given them follow them of course.....if your not.....experiment! Each setup is different, lighting, colors, contrast these and other factors could all affect how much you can push those numbers.

Interesting, thanks Pedro.

P

 

[Phil Bates]

Just wondering, is there a penalty for being too conservative, like 1% positive where there is a lot of distance in the scene? Other than the scene not looking as deep, do clients complain?

Thanks,
Phil

 

[Stefan Christou]

1% neg 2% possitive is the same for Sky over here in the Uuuukaaaaaay.

BTW for some reason when they say negative they mean out of the screen, and positive they mean into the screen. You can also go beyond that for certain shots etc etc, yada yada yada

 

[Charles Angus]

I don't personally have any problem with shots not being deep enough. I find I normalize my S3D depth perception to the depth presented. What I mean is that I've never felt the horizon to be unrealistically close because of narrow IA.

I also HATE it when stuff is in front of the screen.

My personal feeling is 2% positive, and never any negative. More than 2%, and you'll be causing problems on bigger screens.

Some simple math for you:
Figure the biggest screen size you're targeting. Say, 40 feet.
Determine what % of that 2.5 inches is (an average adult viewer's IO - use less for a child). In this case, it would be .5%.

That's your zone of absolute safety - this will never require anyone to diverge their eyes, no matter how close to the screen they are sitting. You can go outside of that, and be fine - but that's the zone of total safety for that screen size. I seem to recall Legends of The Guardian usually had about 2% positive.

You can get more complicated with your math than that, too, to enter more factors in.


Keep in mind, too, that the bigger the screen, the deeper it looks. The perceived depth scales with the screen size.

 

[Pedro Guimaraes]

1% neg 2% possitive is the same for Sky over here in the Uuuukaaaaaay.

BTW for some reason when they say negative they mean out of the screen, and positive they mean into the screen. You can also go beyond that for certain shots etc etc, yada yada yada

yeah, sky just went with the discovery specs......They are trying to keep things mild since they hope that people will be watching 3dtv for hours and hours at a time....instead of the usual theatrical 2 or so hours....

 

[Jason Goodman]

I'd be very surprised to see anyone from these networks actually measuring the parallax on screen. They would look at a given show and either want to air it or not. Do you think any network would refuse to show U23D because Bono's hand comes too far off screen? I'm thinking no. And that is a compelling moment in the film.

Aspiring stereographers can rely on numbers, formulas and guidelines, but to really shoot 3D properly you have to shoot and post A LOT of 3D. And then you have to look at a lot of 3D, in a lot of different ways. After gauging the numerous subtleties that make a particular image or sequence work or not work, show the material to a lot of people. Listen to their feedback. Eventually, you will start to know what the right amount of parallax for a given shot will be. You will learn how much you can push things or how much you need to hold back in a given situation.

People seem to want to calculate 3D settings in the way an AC measures his distance to focus, and then just dial that in somehow. Its not the same thing. There is no formula that tells you how to compose a striking photograph. 3D is not about following rules and doing math. Surely they come into play, but it is not connect the dots. There is no secret recipe, just shoot a lot.

 

[jonnycom]

It's important for the guidelines to state what their definition of positive and negative is. It seems that about 50% of the people have them reversed. To me it seems logical that negative means going deeper into the screen and positive is coming at you.

 

[Patrick Tresch]

3D is not about following rules and doing math. Surely they come into play, but it is not connect the dots. There is no secret recipe, just shoot a lot.

Your probably right by shooting a lot you'll always gain experience and nothing is more valuable as experience.

But start following the rules and learn faster than trying to reinvent the wheel and keep doing the same mistakes as all newbie.

A friend of mine did this app, not only taking in consideration screen size but also all other measurable factors (sensor size, lens, distance FG, BG, screen plane... etc.)

http://www.cine3d.ch/cine3d/outils/41-stereographer-une-application-iphoneipod/

Experience will definitly help in judging scene contrast, movements... that will help the 3D friendly experience.

Patrick

 

[Bruce Allen]

I agree with Jason and Pedro.

May I add: 1% - 2% is a stupid, simplistic restrictive spec.

A much better answer would be:
1. Most shots should minimize eyestrain. For the majority of shots, what the eye is supposed to be focusing on should be converged VERY close to screen plane (within 0.5% or ideally at screen plane - Vince Pace method .
2. Maximum stereo budget should be 2% into the screen, 1% out of the screen. For occasional longer shots designed for ease of integration, 2.5% in, 1.5% out can be approved.
3. The focal point should not move more than 1% from cut to cut
4. EXCEPTIONS - you can go up to 5% out of the screen if the content:
- is centered in the screen and does not break horizontal edges
OR
- it is flying in or out of camera (breaking horizontal edges for maximum of 0.25 sec)
- it is very dark

What do you guys think?

Bruce Allen
www.boacinema.com

 

[Jason Goodman]

I'm not one to agree or disagree with paint by numbers guidelines. Watch a lot of 3D, figure out what looks good, pop the glasses off, watch it again. Shoot good 3D.

The one thing that is not up for debate is the definition of positive and negative parallax. Like it or not, negative parallax describes items coming off the screen. It may be easier to remember if you consider that your eyes need to cross to observe items with negative parallax. The act of crossing could be negative eye rotation, while rotating back out towards parallel might be positive rotation. Spottiswoods or somebody like that defined this like 100 years ago or something.

 

[Charles Angus]

I agree with Jason and Pedro.

May I add: 1% - 2% is a stupid, simplistic restrictive spec.

A much better answer would be:
1. Most shots should minimize eyestrain. For the majority of shots, what the eye is supposed to be focusing on should be converged VERY close to screen plane (within 0.5% or ideally at screen plane - Vince Pace method .
2. Maximum stereo budget should be 2% into the screen, 1% out of the screen. For occasional longer shots designed for ease of integration, 2.5% in, 1.5% out can be approved.
3. The focal point should not move more than 1% from cut to cut
4. EXCEPTIONS - you can go up to 5% out of the screen if the content:
- is centered in the screen and does not break horizontal edges
OR
- it is flying in or out of camera (breaking horizontal edges for maximum of 0.25 sec)
- it is very dark

What do you guys think?

Bruce Allen
www.boacinema.com

These are pretty reasonable guidelines.

I personally disagree slightly with #1, and disagree strongly with #4.

Regarding #4, I am personally opposed to objects EVER being in theatre space. ESPECIALLY things flying into or out of camera. Unless you're making theme park attractions, but that's another story.

For me, S3D movies create an illusion that I am looking at the real world through a window. Things coming through the window seem to conflict with that metaphor. And not just because of edge violations, although those are also a problem without a solution (no, floating windows aren't a solution - they're a low-grade band-aid).

Because of that, I disagree slightly with #1. Yes, having the object of interest being close to screen is good. But that shouldn't come at the expense of keeping things inside the screen. It also is often good to have the object of interest be the closest or one of the closest objects in the shot (storytelling reasons can modify this, obviously, but barring some story reason this works for me).

But on the whole, Bruce's suggestions jibe with the conventional wisdom on the subject. I only disagree because I have a more narrow idea of what constitutes good S3D than most people.

It's important for the guidelines to state what their definition of positive and negative is. It seems that about 50% of the people have them reversed. To me it seems logical that negative means going deeper into the screen and positive is coming at you.

Might seem logical, but it's not the case.