I've done a search on the website ... on the web ...etc. and I can't seem to find the answers, so here it goes...
I've got a feature project that's in pre production. Over 75% of the script is in a bar. We want to shoot on a soundstage because there is so much dialogue. Question is how we are going to deal with the front windows. Most of the shots are night interiors, only a few will be sunset, none are day interiors.
>The bar is supposed to be in a city like Chicago. The director wants to be able to shoot towards the front window on several shots, and see buses, taxicabs, pedestrians, etc. passing by. I can handle the pedestrians; It's the buses and cabs I'm concerned about. I'm thinking of using rear or front screen projection, but I'm worried about the size of the image, and how that's going to read. The other option I considering is a huge translight (for the building across the street) and pulling cars though the stage. (I don't want to drive them- I like my brian cells the way they are.)
>There are three bifold windows 8' tall x 4' wide across the bar front, and they open inwards. (they will be open for some of the shots). There is a glass front door on one side as well.
>I am using mostly practicals for the bar set. Fortunately, the practicals hung from the ceiling are lekos hung on truss (just like the actual bar the story is based on). I'll probably be shooting 5279 super 35 composed for 2:35-1. I will be shooting at an average of 20-25 footcandles. Our look will be shallow DoF, so the exterior will not be sharp, except when someone enters the bar.
>My concern is getting enough image size on the projection to cover the entire front of the bar. I will be doing a test in February, and I don't want to waste a ton of money only to find out the concept is a total waste of time. I've only done process photography on cars, so this is an entirely different scale.
>I've never tried something like this, so any recommendations will be greatly appreciated.
>Timothy Norman Huber DP/ Atlanta
>Question is how we are going to deal with the front windows. Most of the shots are >night interiors, only a few will be sunset
>Not knowing everything about your situation, it seems that rear projection is not likely to be a practical option, but more importantly, not really necessary. I did rear projection last year and it was a big deal. Necessary on that project but not something to be engaged in lightly.
>As for the buildings across the street: as much as I like translites, you may find it is cheaper and more flexible to build some basic set pieces "across the street" especially if it's not supposed to be a wide boulevard. Some set items and a good scenic painter can do wonders and then you can light for sunset or night, change the lighting, put lights in windows, have people go in and out of doors, etc.
>Moving cars across the set isn't as hard as it sounds. I've done things as big as a Greyhound bus pulling up to a diner on a foggy night and it worked great; and that was on a fairly small sound stage.
>Also, since its night and not in focus you can do a tremendous amount with lights. Recently did a Las Vegas hotel room on a set. We rented a few flashing neon signs (big ones but cheap) and put small fairy lights into black foamcore in the shapes of buildings. Some flashed, some moved. We had fog and translucent sheers on the windows so it's not exactly the same, but the idea is that some animated lights can be a big part of selling the illusion.
>Frankly, as great as they are, translights can be a little confining, they are never quite big enough and especially if you are doing a lot of scenes on that set (as you are) they never change. I always enjoy spotting translites on big movies - no matter how good they are, if they don't change it bothers me. (The standard device these days seems to be to add some CG birds.)
>Also, for sunset, you will have to deal with the issue of changing the lighting for the translite (which can be done, if not always convincingly) and keeping the "midground" lighting completely clean off the translite. Wouldn't it be so much nicer to be able to rake some warm streaks of sunset across the facade of the buildings? Depending on your situation, basic set building like this isn't nearly as expensive as you might think - especially opposed to a very large translite rental. I'm not saying it is necessarily cheaper but it may be.
>A final thought: and I'd love to hear other opinions on this: instead of rear projection - how about putting greenscreen outside the windows. Not as real or as good as a real set, but I think probably easier and cheapr than rear or front projection.
>I did rear projection last year and it was a big deal. Necessary on that project but not >something to be engaged in lightly.
>That's a why you ask first, spend second. I don't want to enter into this one until I know the facts. My nightmare is having to do rear or any kind of projection on this one, but within the confines of the studio, and my own personal experience, I got to admit, I don't know to pull this one off. Gotta learn this sometime.
>Is this questioned better served to my production designer, who has yet to be picked? This does seem to be a set issue. Maybe FX related? Maybe. Maybe I'll sleep on it and have an epiphany (or a cow if I don't figure it out. )
>I do know that we need to do this on a stage. There's too many long dialogue scenes to just put up with the noise of aircraft, barking dogs, lawnmowers, deranged lunatics, and other inconveniences. It was hard enough on the last show. The other issue is how I want to shoot. I want to be able to pop the walls out and have the control you have in the studio. It's hard to impossible to pull off a big crane shot in a location like a long tall narrow bar.
>Timothy Norman Huber I just need a cocktail
>It's the buses and cabs I'm concerned about. I'm thinking of using rear or front >screen projection, but I'm worried about the size of the image, and how that's going >to read My concern is getting enough image size on the projection to cover the entire >front of the bar.
>If you use Vista Vision background plates, you can easily fill a 20x24 screen , but I agree with Blain that you may be going after a difficult way to get where you need to go. I go out as an RP supervisor with the Motion Process Vista projectors, which sync really easily to the shooting camera and are relatively production-friendly... ... but you are still dealing with shooting a lot of plate material and then constantly re-racking it and playing it. Even the best RP brings with it certain constraints on camera angle with respect to the screen and you need a lot of room on the other side of the screen to back up the projector - unless you work with a long projection lens, you are plagued with "hot-spotting" whereby the center of the image is hotter than the edges due to the difference in incident angle from the projector. It is true that RP is generally less expensive than compositing - no scanning etc and the composite is done on your neg, but it still requires equipt rental and crew and often bumps up the size of the stage, and you have to chase light off the screen, especially on something like this where you don't have a moving vehicle in which to hide the bits and pieces that are less than perfect.
>Building set pieces that give you the street and winching vehicles back and forth may actually be cheaper...you might even be able to build a bus facade that plays OK. Depending where in Chicago, you may or may not see sky, and putting in a blue sky drop that you can play sky changes on is a cheap way to get different looks out of the set.
>If you are interested in pricing RP, you can contact us by calling Bob Shepherd at Chandler Group 310 305-7431 here in the LA area. Ours is, I believe, the only Vista Vision RP system with which the operator and videotap see the composited image - projectors which are based on Mitchell camera movements can't do this
>You could also check out Hansard Vision - they do have a website.
>Bear in mind that you need pin registered projectors and steady cameras to shoot the plates - images weaving or jittering behind a window will be VERY distracting. That would leave out Background Engineers as their projectors are not pin registered.
>All in all, I would suggest that you go digital if you can afford it or build a practical set...but RP might make sense - it is hard to figure the economics without knowing how many shots etc etc. One obvious advantage of blue or green screen is that you can change the background around as you edit - but this is capability comes at a price. Good luck Mark Disclaimer - I am involved in RP and specifically with the Motion Process Linear Loop RP systems that won their inventor a Plaque from AMPAS... ...but I am also involved in lots of practical sets and blue, green and red screens... SO THERE!
>I forgot to mention that I totally agree with everything that Blain said about translights...with the caveat that Rosco does translights that have a day look and a night look (front light/back light) which is pretty trick and I think won them a certificate from AMPAS this year
Good luck with your project.
>CML is a terrific resource for this kind of thing.
>Mine is just one opinion, I'm sure you'll hear many others. There are so many interactive factors (budget, size of stage, experience of production designer, director's wishes) that someone who is not there can't just give you the one right answer.
>I'm pretty sure, however, that you'll find projection is not the way to go. It's a big solution to a small problem. I know that when I did rear projection, I was told (no way of knowing for sure) that the machine we used and the guys who ran it were some of the last people here who did it - and that's in Hollywood.
>Solving problems like this is a big part of the fun of being a DP - enjoy. And thank your lucky stars that something like CML exists.
>Mark Weingartner wrote:
>My concern is getting enough image size on the projection to cover the entire front of >the bar.
>I plan on shooting longish lenses for most of this. Think the "Insider" for lens choice.
>...but you are still dealing with shooting a lot of plate material and then constantly re->racking it and playing it.
>How's that compare to a fleet of taxicabs, buses and street sweepers?
>Even the best RP brings with it certain constraints on camera angle with respect to >the screen and you need a lot of room
>The set is about 60 feet long. The stages I'm looking at are in the 120-150 foot long range. Is that enough room?
>and you have to chase light off the screen, especially on something like this where >you don't have a moving vehicle in which to hide the bits and pieces that are less >than perfect.
>All of the light would be inside the set. Does that help?
>Building set pieces that give you the street and winching vehicles back and forth >may actually be cheaper...
>I'm chasing down that cost for comparison.
>Depending where in Chicago, you may or may not see sky,
>The lens selection would pretty much take care of the won't see the sky issue.
>If you are interested in pricing RP, you can contact us by calling Bob Shepherd at >Chandler Group 310 305-7431
>I'll call tomorrow. Promise.
>Thanks Mark, Great post.
>Timothy Norman Huber
>Ours is, I believe, the only Vista Vision RP system with which the operator and >videotap see the composited image - projectorswhich are based on Mitchell camera >movements can't do this...
>Interesting - how is that accomplished?
>Timothy Norman Huber wrote :
> My nightmare is having to do rear or any kind of projection on this one, but within the >confines of the studio, and my own personal experience, I got to admit, I don't know >to pull this one off. Gotta learn this sometime.
>Sounds like an interesting challenge. I haven't done a projection in such a large environment but I think that it may add nearly as many problems as it may solve. The idea of using real set pieces or a backdrop sounds considerably cheaper. It may help to have a specific design and look for each of these shots out the windows depending on the time of day and exterior activity. Lighting gags, moving vehicles, extras...I also would strongly recommend a test to consider how much you would like to diffuse the windows and/or possibly add a second netting material, in shot, to soften the background. This would add depth and help a lot in being able to cheat with the exterior FX.
>Consider the exterior as a strong element that can add considerable dramatic strength to the action and use that to your advantage.
>Whatever you end up doing remember to give yourself as much room as possible outside the windows to create the effects. There never seems to be enough.
>Hope this helps somehow.
>Jim Sofranko NY/DP
>I absolutely agree with building and dressing up a set. Much more freedom and realism. And like Jim wrote using a slight diffusion between the windows and the built set (single white nets, scrims... whatever) ehances the illusion of depth.
>I do know that we need to do this on a stage. There's too many long dialogue >scenes to just put up with the noise of aircraft, barking dogs, lawnmowers, >deranged lunatics, and other inconveniences.
>Maybe because my perspective is from low-budget features, but it sounds like your director wants it ALL: convenience of shooting on a stage, but he wants to see a moving background as if he's on location, and he wants things to go faster and more efficiently.
>At some point, something has to give. How do you know that the time spent cueing up the RP background won't cut into the time you're saving by not filming on location? How do you know it won't actually LIMIT your camera angles & movement compared to shooting on a real location? You might have more space to work than a real bar, but there might be certainly types of arcing moves that would be hard to make in front of the RP screen with a single perspective.
>You might consider a hybrid approach. First of all, break up the view through the windows with blinds & neon signs, etc. so you barely see through them. For most shots, just put some painted flats & moving lights outside the windows to suggest a background -- especially if the windows are in the far background and are out-of-focus. Then use rear-projection for scenes playing close to the door & windows, treating those shots differently than the rest of the film. Because I think that you are going to be slowed down if EVERYTIME you see a window, it has to have a rear-projected background behind it.
>I don't know about greenscreen, since most of the time, the windows will be out-of-focus, and having real lights drifting behind blinds & signs, all out-of-focus, would look more real & interactive with the setif done live. Having a window with stuff on it, like blinds let's say, and a greenscreen behind it, and having it extremely out-of-focus, might make it difficult to pull mattes. And again, you have to consider how you are going to handle camera movement.
>Greenscreen might be good for a cutaway shot, let's say, of a man coming through a door where the street becomes clearly visible behind him.
>In low-budget land, I'd either go with a real location with a great view of the street and really take advantage of the view (shoot on fast stock, wide-open, open all the shades, etc.), or I'd go for stage that had windows that were mostly closed-up (like most bars do) with just some moving lights now & then to suggest a street background. Most bars are black, windowless pits anyway...
>David Mullen Cinematographer / L.A.
>As for the buildings across the street: as much as I like translites, you may find it is >cheaper and more flexible to build some basic set pieces "across the street" >especially if it's not supposed to be a wide boulevard.
>One trick I've seen used to good effect is to stretch a curtain of very sheer "scrim" material in the middle background, between the set elements that Blain describes and the windows of the foreground set. This adds the effect of distance and brings in some "atmosphere" to the air of the background, making it feel as though you are looking at a real city street, complete with air pollution.
>Part of Blain's further comments :
>....and a good scenic painter can do wonders
>I do hope you are banging heads with the Production Designer on a regular basis! If not, you might be missing out on a powerful ally in your corner. Discuss all these possibilities (sets, translights, rear screen) with that person, well in advance of any testing you do, and be sure to include the designer in on those tests. If, for no other reason, it will promote a spirit of cooperation and they might just come up with some ingenious ideas which you will be able to take credit for, later ;-)
>Best of Luck
>Joe Di Gennaro, SOC DP/Camera Operator Sherman Oaks, California
>Wow. I really did not expect the response to this subject. I did the night owl, obsession thing last night, and decided that rear screen would probably be more hassle than it's worth.
>Knowing that a full description usually garners more detailed response, here's a little more info on what we are doing.
>The bar is long, narrow and tall. 60 ft long by 30 feet wide with 14 foot ceilings. The front of the bar sits on a busy two lane Chicago street right beside an El train. From certain angles close to the front windows, the El structure is visible, but not the train itself. Directly across the street is a row of Brownstones, one or two may be storefronts. The street is one way. (I decided that one). Cars park on the opposite side of the street.
>We would love to use the real location (this is a story about a real bar), but the noise would be intolerable. Taxi drivers in Chicago tend to use the horn as a...well I never did figure out why they honk so much.
>The bar is actually a two room bar. The front room is the one that is 60 feet long. The back bar is a 35 foot square, with a hallway and stairs connecting the two. I'd say 35-40% of the action takes place in the front bar, with maybe 15% of the shots aimed towards the front .
>Here's the real kicker. This project is part of a two picture deal. The second story is a European suspense thing, and the producers want to shoot in Prague. That way we can get a deal on everything from cameras to crew and locations. Works great for the second script, but it makes it kind of tricky to deal with the Americaness of the bar story.
>I know that it will be tough to pull of a Chicago story in Prague. We are not locked in yet on this one (the bar story shooting in Prague, that is), but I'm trying to find a way to make it work.
>So, late last night I thought it would be interesting to make life size prints of a bus, two or three different taxi cabs, and a bunch of cars using a large format inkjet printer. Maybe a tranlight or two for the windows on the bus, as an extra touch (& expense). For the bus, I could see getting a golf-cart-like electric vehicle and driving it past the windows. The cars could just run on a cable, pulled along behind the bus, or just by themselves. The rest of the street would be set facades, something I think the Czechs do very well (from what I'm told).
>Does this logic sound doable? Any tricks to pulling off this illusion?
>Thanks, guys. I'm having fun going through this process.
>Timothy Norman Huber DP Atlanta
>Timothy Norman Huber wrote :
>"So, late last night I thought it would be interesting to make life size prints of a bus, >two or three different taxi cabs, and a bunch of cars using a large format inkjet >printer.
>I'm not at all sure you could get away with this. Vehicles have THICKNESS, and relief, and you'd expect to see the front when the bus/taxi is approaching from one side of the window, and the back when it is moving away, at least to some degree. I think using the method you suggest there is a very real risk it will look exactly like what it is....something as flat as a pancake.
Visual Effects Supervisor
Ocean Warrior Moving Picture Company,
>Paddy Eason wrote:
>I think using the method you suggest there is a very real risk it will look exactly like >what it is....something as flat as a pancake.
>Ouch. We don't want that. What if we added a front and rear panel to our mock up. Try to give it a little thickness in the x and z plane.
>I've been building a virtual set in Cinema 4D for previz, and was thinking longer lenses from the other end of the bar would give me that flattening effect to conceal our pancakes. What do you think?
>Right now, I'm just trying to get a handle on budget. Once we get fully funded, I'll be able to bet a production designer on board.
>Paddy, you're the compositing wizard. How hard and or expensive would this be in post?
>Timothy Norman Huber
>Timothy Norman Huber wrote:
>Paddy, you're the compositing wizard. How hard and or expensive would this be in >post?
>Heh. Well I think someone else said it pretty well in this thread before - it wouldn't be the easiest set of compositing shots in the world.
>Glass, with reflections, dirt etc should actually work fine with the right compositing software (I personally rate the "Keylight" plug-in that you get free with the Shake system, and can buy for Cineon and Discreet). You just need a good saturated green or bluescreen out behind the windows, and you obviously (though people often forget this) need to provide the appropriate backlighting from outside the windows (eg, if you're going to have a sunlit snow scene out there in place of your blue/greenscreen, you'd better have a lot of white light bouncing around out there, hitting the glass and window frames from all angles, and shining into the bar). Same if it's a neon-lit night scene out there.
>The nice thing about comping stuff through a dirty set of windows is that you actually don't need to worry about pulling good, clean mattes. Nothing should be 100% transparent - any dirt and imperfections (with the exception of film grain) that you get in your key off the blue/greenscreen should be religiously retained!
>A large blue/greenscreen set well back from the windows is far preferable to a small one close. The closer it is the more blue/green spill light you will get encroaching into and bouncing around the set. So do make all effort to get the largest screen you can light. One about a mile across and half a mile away would be great! Seriously - the perfect position would be at the distance of the far side of the street.
>Then you can address your other problem - tracking. I'm guessing that this isn't all locked off shots? So we need to be able to track camera movement, at least for pan, tilt, zoom and slight rotations. For that you need to have at least two tracking markers in frame at any one time. These should ideally be at the same kind of distance from camera as the scene you are looking to add in place of the blue/greenscreen. So - another reason to set up that big screen on the other side of the street (or at that kind of distance). Put your tracking markers on the screen. Move them as you move from setup to setup, so that you always have at least two well separated markers in frame, preferably clear of complex matte edge areas (eg hair, dirt on the glass etc). If the screen is going to be well out of focus, then the ideal tracking markers would be self illuminated red dots - I'm thinking something based on a watch battery and a bright red LED. This should provide something good to lock onto for the track, and not too hard for the compositor to remove.
>Talking of focus - make sure your compositor understands how to get a good digital out-of-focus effect on the back plate. It's really easy to get this totally wrong, and end up with an ugly digital smudge that look nothing like real photography. I continue to be amazed at how insensitive some people are to this.
>Don't expect to shoot on fast stock (above 200) with bluescreen (less of a problem with green, but still an issue).
>But for all these things, the bottom line is always - get a vfx supervisor on board, and talk to your post house NOW!!!
>You know it makes sense.
>Money-wise - well, you have scanning and recording overheads - I think this kind of thing runs at about 60c per frame nowadays, in and out? For the compositing, well, it's a one guy/one workstation job, and I think a good compositor should be able to turn out maybe one 100 frame shot a day of the kind of work you describe. Compositing (along with all the stuff you get free from a good facility - vfx editorial help, vfx production help etc) is charged at something like $2000 or $3000 a day, I think (more at some places, less at others, it depends if you are going for big name houses). If you find the right person, and aren't in a hurry, you might be able to find someone with a G4 Mac and a copy of After Effects (or a Linux machine and a copy of Shake) who will do it for you for far less, but you need to make really really sure they know what they are doing. Make them do some serious tests.
>This is very rough, off the top of my head, so don't go quoting these prices back to me or anyone!