|
|
Author
|
Topic: IMAX 3D ghosting
|
|
Julio Roberto
Jedi Master Film Handler
Posts: 938
From: Madrid, Madrid, Spain
Registered: Oct 2008
|
posted 02-11-2009 08:55 PM
I've experience plenty of crosstalk in Imax. The reasons are many.
First of ALL. The finest polarized performance theoretically ever attainable today is linear and something like 99.998% efficient. That means that you still will always, even under best ever theoretical conditions, would be watching with your "wrong eye" 0.002% of the "wrong image" (a double image).
That may seem like little, but the eye is a remarkable photon catcher So this amount of ghosting is the very least theoretical for a polarized (linear, circular is quite worse, Dolby is even better) system.
But, of course, such performance in real life is not usual. Most linear systems have a practical performance of about 250:1 contrast ratio. That means that, say in a spot on the screen shared by the left and right image, the right image is 251 times brighter than the same part in the other eye, you would start to appreciate the "ghost" spot of light. (Circular is about 100:1)
DCinema has a contrast ratio of 2000:1, so plenty of oportunities to see plenty of light from the wrong eye enter the wrong filter on the glasses while not being overshadowed by the right amount of light from the right image.
The main reasons for excessive ghosting in ANY movie presented polarized are: -Bad screens. The surface has to be metallic for a reason. Dust is usually not metallic. -Bad filters. Most effective polarizers are also the ones the degrade fastest. They need to be changed "often". -Bad alignment or theater geometry. Filters (projector) needs to be perfectly leveled for linear systems, as need to be filters in the glasses. The projector should maintain a good angle for the incident light to bounce off the screen to the audience in a good geometry. -Audience tilting their heads (in ear-shoulder direction). Circular is mosly inmune to this. In practice, circular also benefits from straight heads for reasons of residual analyzing polarization. -Fingerprints or oils on the glasses, or otherwise degraded glasses filters.
That affects all films, making ghosting worse than it needs to be.
But the "practical" reason for a lot of ghosting is bad cinematography on some movies. They are shot with so much parallax (i.e. the left and right views so far apart in many parts of the image), that fusion of both images is far more difficult, leading to all sort of problems and worsening of any inherent ghosting.
Also, movies with inherent lots of contrast (i.e. a white star field in a black space background) and lots of parallax (i.e. things very far into or out of the screen), make the limitation of each projection setup more evident.
I saw Space Station 3D in Imax and don't recall it being particularly ghost-trouble film. But I've seen so many 3D films my memory is but a big blur.
A lot of Imaxes were shot with their "small" integrated 3D camera which allowed for very little error. I used it once myself during a training seminar in LA, and it was quite a little piece of equipment (that I would never considering using for real-world projects other than Imax, of course).
RealD systems are now trying to avoid these pitfalls by pre-processing the movie to get rid of contrast above their perception threshold in parts of the scenes with large parallax (lots of "3D"). From IEEE "Cancellation of image crosstalk in time-sequential displays of stereoscopic video".
quote: In this paper, we propose a method to reduce such crosstalk. We present a simple model for intensity leak, we assess model parameters for a time-sequential LCS/CRT system and we propose a computationally efficient algorithm to eliminate the crosstalk. Since the full crosstalk elimination implies an unacceptable image degradation (reduction of contrast), we study the tradeoff between crosstalk elimination and image contrast. We describe experiments on synthetic and natural stereoscopic images and we discuss informal subjective viewing of processed images. Overall, the viewer response has been very positive; 3-D perception of many objects became either much easier or even effortless. Since the proposed algorithm can be easily implemented in real time (only linear scaling and table look-up are needed), we believe that it can be successfully used today in various stereoscopic applications suffering from image crosstalk.
That paper is from the year 2000 when LCS shutters were thought to be the future and none of this DCI stuff existed, but it applies somewhat to the same problem.
Other random quotes that may apply:
quote: In the Real D system, the encoding takes place at the projector using an electronically controlled polarizer, which Real D calls the "Zscreen™." Images are decoded when the audience wears complimentary decoding polarized glasses. To allow head movement without upsetting the decoding quality of the glasses, Real D uses only circularly polarized filters in its system. Polarization alone, however, does not offer sufficient protection from crosstalk, also referred to as extinction ratio, with stereoscopic images. The audience experiences such crosstalk as a ghost in the motion picture. To enhance the ability of its polarization method to reject ghosting, Real D employs a "ghost busting" technique, which requires pre-processing of the images prior to projection. In its early systems, Real D's ghost-busting is applied prior to distribution. In future systems, ghost-busting will be applied in real-time by means of a processing box in the playback system.
quote: While Real- D is very popular with audiences and gaining acceptance, the silver screen used in the process can cause a ghosting highlight from one eye to the other that reads like a second offset limit. To solve this, a "ghost-busting' pass is done for the REAL D version, which mathematically subtracts the bright highlights of the primary eye from the secondary eye. This cancels out the effect, but leaves the left eye with strangely dark hot spots. For this and reasons of stereoscopic adjustment and minor softening from beam splitters/mirrors, the untreated primary eye is used as the standard DVD or TV version.
You don't want ghosting? Tell your theater to order an Imax Infitec (Dolby) color-corrected print, replace their current polarizer filters with Infitec's, double the light output to compensate for light loss, and buy Infitec glasses instead.
[Edit: for many technical reasons, it may help in many cases, as a temporary measure to reduce ghosting, to sit as centered and as far back in the theater as you can and to wear two pair of 3D glasses, one on top of the other] [ 02-12-2009, 01:33 PM: Message edited by: Julio Roberto ]
| IP: Logged
|
|
|
|
|
|
|
|
|
|
Julio Roberto
Jedi Master Film Handler
Posts: 938
From: Madrid, Madrid, Spain
Registered: Oct 2008
|
posted 02-27-2009 03:52 AM
It's hard to say without seeing it first hand.
Do you mean a fade-to-black or they actually superimpossed the end of one scene fading-out over the beggining of the next scene fading in?
If it was a fade to black, well, that usually means there is a lot of areas of relative high contrast (i.e. large parts of the screen go to black or very dark) while others still remain lit. If the image at that moment had significant parallax (i.e. "lots of depth"), then ghosting is quite unavoidable on real-life polarized system. In practice, most polarized systems allow certain amount of light leakage to the "wrong eye". When the "right image for a given eye" is bright, it "overshadows" the dimm (leaked) wrong image. But if most/everything in both images is pretty dark, you can make out the bright parts of the "wrong" image over the pitch-black dark parts of the "right" image.
If what you saw where supperimposed fades, well, they sometimes work, sometimes don't. The eyes need to "concentrate" on fusing the two images currently on the screen and finding a confortable point of convergence (i.e. where both your eyes point to). If you show a new set of two images over the first ones, if the point of preferred convergence doesn't match (bad editing practice in 3D), your eyes are going to have to decide on which set of images to converge to, and "in the periphery" will notice that the other image now looks doubled.
We kind-of see this everyday, but never pay attention to it. Put your thumb in front of you about 1/3 way between you and the monitor and "concentrate" in your thumb. If you look "in the periphery" while starring straight into your thumb, you'll see your monitor and everything behind your thumb looks "doubled". With a gigantic Imax image these effects are pronounced.
The short answer: multiple image fades may not always work right in any 3D system. The larger the image and the brighter the image, the more obvious. It can be done and has been done successfully, but it's not something that usually works without putting thought and care into it.
In real nature it only occurs when looking through something like a half-way mirror, where you partially see one reflected "scene" and a different one through the glass. If one of them has a subject of interest at different distances from the half-way mirror, whichever one you are not "concentrating" on will look doubled.
It may also be something else, I don't know. Having seen that film yet.
| IP: Logged
|
|
|
All times are Central (GMT -6:00)
|
|
Powered by Infopop Corporation
UBB.classicTM
6.3.1.2
The Film-Tech Forums are designed for various members related to the cinema industry to express their opinions, viewpoints and testimonials on various products, services and events based upon speculation, personal knowledge and factual information through use, therefore all views represented here allow no liability upon the publishers of this web site and the owners of said views assume no liability for any ill will resulting from these postings. The posts made here are for educational as well as entertainment purposes and as such anyone viewing this portion of the website must accept these views as statements of the author of that opinion
and agrees to release the authors from any and all liability.
|