Russell Scott

OK, so instead of just telling me I am wrong, how far apart SHOULD they put two cameras?

 

And I definitely read that in a book on 3D photography published for the United States Navy, so I guess there are a lot of clueless sailors out there who are "misinformed" too :unsure:

depth is determined by the angular separation your eyes perceive. so if you capture 75mm separated stereo but double the projection area -without changing your distance from screen you effectively change your angular separation - i.e. you perceive more depth. or less depth when you halve the area by projecting it on a monitor.

the Navy probably use 75mm projections because they are trying to get human style space relations, i.e. identical to reality. That's a bad idea for movie making. I can't tell you what separation to use because that is determined by a whole bunch of factors. Factors like, how big its the screen being projected on, how far away will the audience be, what frame rate, how much is in front of/behind the screen, how fast are they moving, what came before it, whats your FOV...etcetc

As I said, filmmaking requires less than 75mm which is exactly why the cameras are a hassle (think about how wide a lens is)

Well, even with a true 3-D setup, they exagerated the effect by spacing the camera lenses farther apart than 3-in. (~75mm), which would produce a 3-D effect the same as our eyes produce with their spacing.

 

By placing the cameras farther apart, like 6-in. (~150mm) you get the gimmicky ball-jumps-off-the-screen-and-is-coming-right -at-you effect that made people duck in the theatre while watching the movie.

not to get off topic but that is wrong...

75mm is the average eye separation of a human, but it does not necessarily follow that you get the same depth effect. and for film work you generally want much less than 75mm

nor does camera separation define position-in-depth of objects...

carry on.... :ph34r:

I am going to be shooting a lo budget feature on 3D HD with a home made double EX3 rig and beamsplitter. Any suggestions or tips out there for things to worry about, or problems someone has experienced dealing with a similar capture medium.

 

Thanks.

 

David McGrory

make sure you can adjust the cameras position and rotation on all axis. check the stereo result prior to shoot - your cameras might be parallel but the chips can be offset, damaging the result.

make sure you adjust for the refraction displacement of the mirror.

Watch your pan speed, its not very forgiving in stereo. reduce the stereo for fast motion *and* reduce your shutter speed. don't let it strobe. Compose with it - don't gimmick the stereo...

bring two laser pointers...

don't use f/30 if you're displaying on a big screen...

if in doubt use less stereo - if you can't reshoot later then use less. a broken stereo shot will damage the rest of your film due to the eyestrain in the audience (once the eyes are strained, they won't recover for the rest of the film)

I'm due to film for the first time in the desert on the Red One.

With regard specifically to controlling light and contrast, what do I need to know before going in?

I'm comfortable with the camera so aware of the limitations in lattitude.

 

I imagine the light is very harsh and high contrast. What's the best way to cope with this? Silks to diffuse and relectors for soft fill? Filtration? - ND grads and low-con or diffusion filters?

 

Any advice gratefully received

depends what desert & what time actually. if the desert is high-altitude you'll find the nd grads will kill the rich blue of the sky (as would polarised filters) but your sunsets will last an hour or more and can be quite gentle light, if its white sand deserts you'll likely have problems with the sand more than the sky and if you're filming people the glare from the desert could be a problem - lots of squinting & ambient light from the ground up...

can i ask what desert?

With respect, I completely fail to see your point. We're talking about a system that currently offers worse quality than existing technologies. (Presumably, the quality could be raised to the same level with 2k DCI – but not significantly higher – by changing the mastering process). And the ticket prices are 33% higher. Surely you cannot regard that as a good, long-sighted business model?

And why exactly should filmmakers be "encouraged into the arena" when the facts of the case look like this? They seem to have been encouraged enough towards traditional IMAX as it is, due to The Dark Knight's success... but IMAX's recent actions could very well be discouraging them again for good.

the point is the Dark Night was filmed in 70mm IMAX because of Nolan's love of IMAX format. and even with that huge budget he could only afford 20 minutes worth. It isn't an affordable format, so the company is trying to make it relevant. You're talking as if this will send IMAX bankrupt, but it already was heading in that direction.

It also bodes well for the Imax Corporation, which two years ago saw its stock plummet after an SEC inquiry into its accounting practices. The company has since bounced back, signing deals with theater chains AMC and Regal to expand beyond its current network of 300 theaters in 40 countries.

<snip>

Add to that the fact that IMAX film is more than three times as expensive as 35-mm, that there's only one lab in the world able to process it, and that the cameras have to be reloaded after three minutes of shooting.

Well, yeah, obviously the digital "IMAX" cinemas are equipped to show 3D as well. But projecting 2D films on a dual-projector system like this was new to me – they're not doing that in other digital-based 3D cinemas, or are they..?

this I don't know, sounds like a mistake to me, you can never align projectors well enough (especially over time) to do this effectively. Even side by side projection needs edge blending to mask the projection differences. Its possible the poor quality of your viewing was driven by the overlapping projectors destroying the clarity of the picture. perhaps the one projector was supposed to be switched off for your screening but wasn't. if it is their plan to double project it won't last long....

Hi All.

there is a lot of bagging going on in this thread but I'm going to suggest that this was not a decision just thrown up for corporate greed.

The IMAX mainstay was education, most IMAX films have that flavour. I see a lot of indie films here that have budgets 1/3 of the cost to simply print an IMAX run. If you think it is tough to get funding for a typical film, imagine the difficulty of educational film funding. A lot of IMAX film making has relied of the philanthropy of the producers, i.e. no profit, no worries because we're doing something significant. Likewise, the IMAX theatres have had to hope that the 5 movies they get per year have been good ones. If they weren't its a pretty nasty risk for the operators. That is a poor business model to expect people to gravitate towards. making it more mainstream gives the theatre operators more stability and encourages more filmmakers into the arena

Antti, it sounds to me like you were in a stereo theatre (without the stereo movie), hence the 2 projectors...

I suppose that most businessmen would think that OMNIMAX must be just a gimmick and so these bean counters probably think that the future of cinema will simply be 4K digital video projcected up to 60 frames per second on a flat screen with a 2 to 1 aspect ratio will be all the market requires or at least all that they will be willing to pay for but I dare say that these businessmen cannot call themselves artists because they are just in the business for the money.

Dome has its place, in fact, it is unbeatable at certain styles of stories/cinema but after having the artist beaten out of me while making 8k stereo dome (2 x 8k) I can tell you that the bean counters are wise to be counting. dome gets real expensive real quick. Incidentally, not that anyone asked, a 4k camera does not equal a 4k dome camera...

part of the problem with these formats, is the clarity of story is difficult to impart. you lose the directional cues of traditional movies so narrative becomes difficult. Immersivity is good in some scenarios, but maintaining audience focus is hard enough on a typical film where the norms are well established, and with dome you are giving them the choice to look around, potentially at key moments when you don't want them to look around. As a result most narratives on domes tend to adhere to traditional cinema concepts, only to use the dome as a gimmick in certain scenes (as you said) but given the immense technical difficulty of achieving dome shots properly and given the high likelihood that the immersive screen will actually dilute the intent of shots, it begs the question why bother on such projects.

In terms of the future of cinema, its all about risk, why reduce the cinemas you film could play in, increase the cost of production, negate DVD sales, all for a film that doesn't use the dome effectively anyway? Worse there are competing dome standards, so what you make for one dome, might not work in another...

That's kind of like asking a sculptor to embrace 3D Studio Max..... ;)

this is perhaps a great analogy. Clay was the first "creature special effects tool", digital came along and it definitely wasn't better - look at pixars first shorts cute, technically amazing for the time, but pretty crap; claymation was definitely better. The medium matured and now the medium has matured so much that much claymation is done digitally! The lessons of sculpting have been applied digitally, with mudbox and force feedback digital sculpting tools (and with 3d printing). Sculpting still has a place, its shifted, but its still highly relevant. But from a movie art perspective these tools have been a massive boon. Look at Craig Mullins or Feng Zhu or Dylan cole (etcetc) these guys have defined the very look of some films and they are 'merely' photoshop artists.

notwithstanding the romance of dealing with film, the digital medium will only get better. Complaining about flesh tones seems premature when comparing it to a medium that has been established for decades (and when it first went colour, the fleshtones were terrible too!)

Film will be around for a while yet, but when digital is ready, its quality will surpass that of film (a CCD can capture both more light and more colour than film can so theoretically, once the CCDs & electronics are advanced enough, you can have a "make it look like this filmstock" button.

I should note too that right now, I generally dislike the look of most digital films, but am obviously not so pessimistic about its future.

Has tape gotten that much better while I wasn't watching? I remember having to shuttle every tape on the shelf to prevent magnetic bleeding. Even though we did this once a year, the 3/4" Umatic tape's images were degrading after only about five or six years. Maybe tape has better integrity, now. Is that the case?

no you're right, you're looking about about 10 years in the correct conditions (magnetic bleeding is a reduced problem). but being digital it is in a WSIWYG format, so there is no subtle degradation over time as with film.

At first, when I watch footage from an HD shoot, I think "Wow, that looks pretty good." Then after a while, I realize that I'm just looking at ones and zeroes, a rigidly defined image that really be no more than that.

 

I think film will be around because it is such a solid recording medium. I may be wrong, but don't they burn Movies/TV shows onto flim for archival purposes? I was always taught that with digital archiving, you constantly have to be upgrading your hardware and making sure to migrate your data regularly. With film, you keep it in a can and in a cool place.

:P

isn't that the real issue though? its the idea of a non-physical medium that you don't like not the actual images(not trying to get into a flame war here)?

as to storing digital films you can back it up to tape drives, which are just as stable as film reels probably better too, because a tape could store multiple films, allowing you to make multiple copies . ie, instead of having 20 individual backups of 20 individual films, you could have 20 tapes each with 20 backups... your physical footprint is much smaller too...

at a guess I'd say you're using a cmos camera with an incorrect frame rate.

basically a CMOS camera will warp the image if you don't have the frame synced properly with the movement, hence the jitter varying with height (the parallax ground to the plane will result in different apparent motion) ..

if you aren't then ignore what I've said

I'm not so sure. First, getting a movie theater to take your film is to do with a lot more than having it on a format they can project. The main concern is what limitations of use the studios put on equipment they effectively own.

I can certainly see your point in the near future. however in the longer term I don't see it a problem, I don't believe that digital projectors will remain so expensive that cinemas will remain in limiting agreements.

from the point of view of the cinemas, restricting themselves to a certain type of content for the sake of a $15-30k outlay will not be an attractive prospect. Admittedly I am picking numbers randomly - a current long throw digital cinema projector costs $110k and a short throw costs $30k but I expect these to drop to a certain baseline price. I really can't see a long term business case for the cinemas locking themselves into restrictive contracts.

additionally, in the US as I understand it, there are strong laws preventing such anti-competitive behaviour, though obviously its no guarantee of proper conduct...

"Every person who shall monopolize, or attempt to monopolize, or combine or conspire with any other person or persons, to monopolize any part of the trade or commerce among the several States, or with foreign nations, shall be deemed guilty of a felony, and, on conviction thereof, shall be punished by fine"

I'd like to hear folk's ideas on market turf with the change-over to digital projection? Will this give indie producers better access to screens or will it, somehow, help Hollywood protect these screens for itself?

 

I am also concerned if film manufacturers will even bother to make negative stocks when their true money makers- print stocks- dry up.

 

For me, this all has a vague feeling like the death of a friend.

I'm a bit surprised by all this to be honest. first of all this will most definitely give better access to screens. There might be deals struck initially, but digital projection will both improve and drop in price to a point where such deals will be untenable for theatres to continue with. If anything digital projection will *increase* the number of arthouse cinemas, if they can fit out a small theatre with good projectors for $15k (probably 2 years away) it can only help.

likewise:

I don't know how a cinematographer can call him self one if he shoots digital. It's about art and passion and creativity

if its about art and passion and creativity, then don't stress about the technical details. As the product matures you might find chip size is adjusted resulting in more filmic DOF, better lenses, better digital 'film response' approximation etc. you are all talking about a product in its infancy, and talking cross purpose too, digital projection is not digital capture. Digital projection simply lets you cut vast quantities from your films budget. If you want to use 'real' film to capture nothing is stopping you... plus the comments have fundamentally ignored DI as a process that many films are undergoing. does that cheapen the whole thing? what about going from DI to film again isn't that a bit nonsensical?

I'm left scratching my head a bit here. Lower distribution cost can only be a good thing and most large budget films that can easily afford distribution costs are already digitized.

Hello,

 

thanks for your reply.

I have found some trouble with the 750...different color temperature, one of them was a bit higher the other one, unbelievable!!So, different vertical parallax!Big problem...we can adjust it anyway.

I have already tried the D21 and they are better.The only problem, the interpupillary distance is a lot and it avoid me to keep the subject closer.

Using the parallel system is the best way, anyway.Less trouble.

One question: which rules do you use to calculate the distance to the subject?

I have found something but it doesn't include the lens and the format used(65mm,HD,35mm,16mm). And especially where I am going to project it(dimension of the screen).

Do someone have some advice?

 

Thanks again and all the best

 

Michele

not sure if you got my original message, so I'll reply here too.

a vague guide is cam_sep = f/30 is the easiest formula to use (f= dist to screen). it is what I would call the minimum guide for stereo. working your specs backwards, with a 750 being 13cm or so across, your screen depth will be 3.9m away.

this doesn't include lens either, but the simple rule of thumb is the more the zoom, the less the i/o needs to be. So effectively 3.9m will be the minimum zero parallax you can achieve.

the d21 is no good, far too wide. you will encounter differences between any cameras you use, it depends too on the rig you are using (in terms of vertical shift etc).

the funniest I know about is from LOTR.

a drunken extra was not noticed as they filmed a long tracking scene (I forget exactly where). he stumbled down stairs, through ranks of soldiers and generally causing havok. Somehow this was missed until the entire set - an entire town - had been destroyed.

The solution was to track it all into 3d, build the entire set again digitally and matte him out. I seem to remember it cost $3million to fix, but not sure.

It was told in jest but hooo boy was there a lasting was a bitterness

:lol:

I don't it will. I think the standard will always be 24p. I hate the look of 30p. 30p is soap opera and anything faster will take away the dramatic look to it.

A debate on some other forum with members practically dancing around the idea that 30p will be next gen frame rate for some reason or another.

who knows, if James Cameron gets his way, maybe 48fps will be the next digital standard!

May I ask what it is that you intend on shooting and where will it be shown? you will have a lot of trouble using unmodified 750s side by side for anything other than distant objects

oops, it wouldn't let me edit, so to clarify,

assuming a lens with FOV of 45 deg (24mm) with the default i/o for a 750 (looking at the specs) you're looking at a comfortable zero parallax of around 4m, you can push this closer, but depending on screen size and length of film, it isn't a great idea. There is a limit to which you can adjust stereo before pain sets in.

the i/o calc is done assuming using the primes. I don't have the specs on the lenses but it appears to me that they would affect the achievable i/o. certainly stock matteboxes will be out of the question...

I am still reading the book and it is very interesting. It is plenty of rules and details.

Because I am going to use a HD cameras side by side(Sony 750), and because the format is 2/3'', I think the rules about interocular distance/subject distance are different.

What do you think?

And I have found different speech about convergence.

Do someone knows something about that?

Which kind of criteria do you follow when you are shooting?

 

Michele

May I ask what it is that you intend on shooting and where will it be shown? you will have a lot of trouble using unmodified 750s side by side for anything other than distant objects.

The convergence 'speak' is a friendly disagreement between the players. The maths says don't converge your cameras, keep them parallel otherwise vertical parallax will be introduced but if you keep them parallel there is more work in post to achieve correct stereo. overall it is more flexible but convergence is more instant for the director. problem is, when you're toe-ing in, if you don't separate the lens from the electronics you haven't got a hope of getting a usable i/o, so parallel cameras are likely to be your only option (this is actually good!).

your intended display size makes a big difference too...

hope that helps