2004 Proceedings
2004 Photos
|
|
Stereoscopic Displays and Applications XV (2004)
Proceedings of the SPIE Volume 5291
Discussion Forum
"3D & 4D attractions: the new stereoscopic cinema"
(Summary)
Tuesday, 20th January 2004
Introduction
by Lenny Lipton
To the layperson the term 3D is well understood - it means stereoscopic. It's a term used to
describe stereoscopic movies at theme parks, or anaglyphic comic books. The waters get
murky when the term 3D is used for computer graphics because here it was first used to
differentiate vector from raster graphics two decades ago. As a consequence, for computer
graphics, the term 3D usually does not mean stereoscopic.
What then are we to make of the term '4D'? In the context of the cinema one might think that
the reference is to Einstein's space-time continuum and that the fourth dimension is time. But
that's not what the term means when applied to theme park venues. Here 4D refers not to the
dimension of time, in addition to the three spatial dimensions, but rather to practical special
effects used to heighten the theatrical experience. On my family's yearly hegira to Southern
California's theme parks, I have experienced, in stereo theatres: droplets of water sprayed in
the face, vibrating seats, laser beams projected around the theatre, and virtual mice brushing
against the legs.
Taken to its logical extension certain dramatic situations, such as the drowning of Pharaoh's
minions in the Red Sea, would call for flooding the theatre with the resultant loss of life of
paying customers. The key term here is 'paying customers', who if drowned cannot continue
to purchase tickets, making the application of the 4D flood unlikely on economic grounds.
Considering the possibilities of the medium, it was a genuine pleasure to have some the world's
great experts in the field of the stereoscopic cinema critique the medium in what, for me, was a
pleasurable discussion that invited spirited audience participation.
We had intended to record and reproduce the full discussion in these proceedings.
Unfortunately, while all of the complex stereoscopic projection systems worked perfectly
throughout the conference, the simple sound recording system failed to do its job. Fortunately
Neil Dodgson took extensive notes during the panel discussion which we have reproduced below
in an edited form. While not a verbatim record of the panel, they provide both a summary and
a flavour of the lively discussion.
The panel members:
Lenny Lipton, StereoGraphics Corp., (Moderator)
Lenny wrote "Foundations of the Stereoscopic Cinema", which has inspired many people and is
available on the SD&A website <www.stereoscopic.org>. He currently runs StereoGraphics
Corporation.
Max Penner, Paradise FX Corp.: Started out in 1969 in cinematography; did his first stereo
work around 1989 with Disney on "Muppet Vision 3D"; has since done about 30 pieces in
various formats. He has developed stereoscopic camera rigs with variable interaxial and
variable convergence.
Jason Goodman, 21 st Century 3D: First tried to make stereo movies in college in the early
'90s. Set up and runs a company making 3D content and cameras.
Sam Zhou, IMAX Corp.: Joined IMAX in 1994. Has a PhD in digital image processing. First
stereo work was the film "T-Rex - Back to the Cretaceous". He has since done the 3D
Nutcracker, Space Station 3D and other stereoscopic IMAX movies.
Chris Ward, Lightspeed Design Group: Came into stereoscopy from laser light shows. He
moved into combining stereographic projection with stereo-laser content. In last 3 years the
company has moved into all-digital media, which has resulted in lower costs.
The panel discussion:
[Panel members are identified by their initials. 'Q' indicates a question from the floor; if the
questioner's name was recorded in our notes, he or she is named in square brackets. "Panel"
indicates a comment from multiple members of the panel. "Comment" indicates a comment
from the floor.]
LL: | Before the late '80s there was no real continuity in the stereo cinema. Since then it
has established itself in a niche market: IMAX and theme parks. Before the late
'80s, there were only sporadic bursts of stereo production; we have now had a
sustained 15 years of effort.
|
Q: | Why does it seem that digital makes it easier to do 3D production?
|
JG: | Digital is good, easier to use, but not yet up to the quality of IMAX. Having said
that, the physical media used by IMAX (and other analogue techniques) have
problems; e.g. the two films used to produce stereo get different physical jitters
after a few showings (sprocket wear, etc) whereas digital has no such problems.
|
MP: | There is an immense amount of knowledge and experience in producing stereo.
The big issue now is to ensure that post-production is good. This is much easier
and quicker in digital media compared with physical media.
|
LL/MP: | It used to take a long time to get from shooting 3D to viewing it (shoot the film,
get it developed, get it synchronised, get the projectors set up, show it). It is now
down to showing on the same day as shooting, so directors and cameramen are
now much more able to understand what is going on.
|
CW: | Digital post-processing and projection allow much better quality stereo, causing
less eyestrain for viewers: you can get essentially perfect stereo. Indeed, they
shoot higher resolution than they need to project in order that they can throw
away data in the quest for perfection.
|
SZ: | [Changed the subject to the economics of 3D] IMAX went into 3D by accident. An
IMAX cinema covers almost all of a human's field of view, plus you can make actors
real size on the screen. The image is very sharp, which helps the 3D illusion. IMAX
is very bright which also helps (the pupil shrinks to its minimum size). The problem
with 3D is economics: a 40-45 minute documentary costs $15-20 million. IMAX
has about 250 theatres around the world and about half of those theatres are
stereo-capable.
|
LL/SZ: | James Cameron's ‘Ghosts of the Abyss' was shot in digital HD and then blown up to
IMAX, but it involved compromises.
|
JG: | Almost any animation out there is now done in 3D software and it should be
relatively straightforward to convert it to stereo, simply by shooting a second
camera.
|
SZ: | Every time we have tried to get a production cost estimate from studios for a 3D
CG film, the cost estimates that have come back have been prohibitive (over $20
million).
|
Q: | Why is this the case?
|
Panel: | It may be because of the CPU time involved. For example ‘Toy Story' took 100
years of CPU time to render (on a render-farm). We now have faster computers,
but we also have artists who put more stuff in, so the frames still take as long to
render.
|
[Another reason is that some CG effects used in high quality CG films do not work
when converted to stereo due to animation short-cuts. In order to produce good
stereoscopy, those effects would need to be re-authored with proper depth
information which could be expensive.]
|
[There was widespread disbelief in the audience that the cost could really be as
high as that quoted by SZ, but no one else could come up with any concrete
figures for how much they thought it might really cost.]
|
Q: | How much does it cost to install stereo in a neighbourhood cinema?
|
LL: | Not that much, the real costs come in maintaining the equipment and ensuring that
things stay aligned and in sync. The equipment is cheap compared to the salary of
a trained projectionist who is able to keep everything working.
|
Comment: | Digital projection may solve this problem.
|
MP: | The cost of the digital projector and the screen is high compared with a non-digital
projector and a non-polarisation-preserving screen. But digital cinema is now
creeping into cinemas via commercials, so it just needs someone to add a
polarisation-preserving screen.
|
Q: | [Mike Weissman] Is HD of good enough quality?
|
MP: | It works surprisingly well. I've done some experiments and there are colour-depth
and resolution issues in blowing it up onto an 80 foot screen. To get to high enough
quality you currently need to record uncompressed data, which uses ludicrous
amounts of disc space.
|
Q: | [referring back to an earlier comment by CW] Is 3D really perfect? How many
people walk out of IMAX cinemas feeling nauseous?
|
SZ: | IMAX has done extensive research; the majority of people prefer 3D, but a minority
do not like it and some do find it makes them feel ill.
|
LL: | CGI is totally controllable so it's theoretically possible to create a "perfect" stereo
movie in CGI; whereas real-world cinematography is more problematic.
|
CW: | It may be that the best way forward would be to have digital cinema that is 90%
2D and 10% 3D, e.g. a movie like ‘Shrek' could incorporate 8-12 minutes of 3D.
|
Q: | How long can people watch 3D for?
|
SZ: | 45 minutes of 3D is OK, because that's what IMAX does.
|
MP: | Extending that to 90 minutes shouldn't be a problem, so long as the
cinematography is done well.
|
Comment: | [John Merritt] So long as the screen is over 2m away, people should not have a
problem; so long as we don't pull things too far out of the screen.
|
Comment: | Properly done 3D should not cause eye fatigue. In my opinion, ‘Ghosts of the
Abyss' and ‘Spy Kids 3' both violated rules of good 3D and therefore caused eye
fatigue in a large number of audience members.
|
Comment: | Good post-production tools are needed and projection technology needs to be
properly aligned and synchronised.
|
SZ: | IMAX has dedicated projectionists but most other cinemas are much more
automated and therefore you need something that will work and will continue to
work with little maintenance or attention.
|
LL: | The only real solution is a single projector doing both views.
|
Q: | You need to do post-production differently for the different stereo media (screen
size, etc): how do you cope with this? And what do you put in the archive?
|
CW: | [explaining how Lightspeed / DepthQ do it] We currently zero-out our master
content (infinity is at the screen) and then punch in numbers for the size of the
screen to push the content back appropriately.
|
Summaries from the panellists
|
MP: | I am optimistic about digital stereography: it gives more control; it is faster and
cheaper then film; it could solve many of the problems we've been seeing; people
will learn how to control stereo better.
|
JG: | I agree with MP and look forward to advances in technology.
|
CW: | I agree with the other two and look forward to seeing more and better content.
|
SZ: | IMAX is committed to 3D; we are currently installing the next generation of 3D
projectors; we want to make production of 3D easier: provide better technology for
shooting 3D.
|
LL: | There is a lot of hope. People are going to figure out how to do stereo properly,
now that we have the enabling technology.
|
|