Home of Professional Cinematography since 1996

Lighting for Blue/Green Screen

>Hi everyone.

>I have had some experience in the past in lighting a blue room, footage that was shot on 16mm and was manipulated later on with "Henry". The thing is that I couldn't make many conclusions from that shoot.

>Now, I'm facing two shoots, one is 35mm shooting a room and thinking of hanging blue/green screen on the out side (Studio) and replacing it later with something else.

>The other one is Video (DVPRO, Beta SP or DigiBeta) and in that case it is more comprehensive work in terms of different locations, full environment blue/green room.

>My questions are these:

>What are the lighting implications on the objects? How do you deal with BackLighting the object (if there is anything special there in terms of separation from the background)? What would be the best way to light the blue/green screen out side the window? What would be the way to deal with the full blue/green room (lighting wise)? Which lighting units do you prefer for lighting the screen. How do you deal with Blue/green spills on the objects. What is the ideal stop ratio between the key object light to the blue/green screen? Is there any preference to use Blue or Green aside from avoiding color components that exist in the object. What is the deal with Marker points on the green/blue screen for "Post Area Replacing"

>Answer's/comments to any of these will be much appreciated.

>Thanks a Lot

>Yariv


I didn't quite understand weather outside the window meant in a studio or outside a studio.The Important thing for blue screen is that it should be as flatly and as evenly lit as possible. So it depends on your location and the weather conditions. If you have the money and the room go for a mini/maxi brute that would give you punch and evenly spread light. You might not even need light if you have the right weather conditions. Just remember to have the blue screen about a stop over exposed compared to your key light on the subject to give as much separation as possible(opinions would probably differ on this).

>>What would be the way to deal with the full blue/green room (lighting wise)?. I generally use sky pans about 5 ft apart and about 5ft away from the screen. How do you deal with Blue/green spills on the objects. Keep your subject as far away from the screen as possible to avoid light bouncing back off the screen usually appears as halo or back/rim light. This can be awkward for the post op!

>Is there any preference to use Blue or Green aside from avoiding color? I don't think so but check on this point. What is the deal with Marker points on the green/blue screen for "Post Area Replacing" These points are for the post operator to track the screen subject to the live action, if there are any camera moves. Best tracking markers are white camera tape crosses inside black gaffer tape crosses.

Hope this helps and is not too late.

>Good Luck

>Matthew Woolf


>Opinions absolutely vary on this.

>IMHO this is the worst advice anyone could be given.

>Go to the CML web-site and look at the frame grabs of blue & green screen tests, they range from -2/3 to -3 stops.

>Then look at the blue & green density values.

>At -2/3 the blue layer is reading 255.

>Geoff


>I would second Geoff on this. When I shoot blue/green screens I try to light it as little as the system will allow.

Daniel Villeneuve, c.s.c.

Directeur-Photo, Director of Photography

Montréal, Canada

http://www.aei.ca/~davil


>The key word here is "system." Different systems work differently and pull mattes optimally at different light levels. Talk to the guys who will be doing the effects ahead of time. Since they do this every day they do know what stop the screen should be lit to. Also be careful with advice. Like Oscar Wilde said "Advice isn't good for much. That's why everyone is always giving it away." I think it was Wilde. Tom "Don't forget to separate the actor and the screen" Jensen >http://home.earthlink.net/~tomjensen/


>I prefer the screen to be in the range of one stop down for green and at stop for blue. This assumes everything at tungsten or daylight. If screen lights are gelled, or colored fluorescents are used, discussion of screen stop is less relevant because now we're dealing with saturation, and saturation is very difficult to measure accurately on normal meters. The single largest mistake I see is *way* overlit screens when using colored lights or heavy gels. If you're using narrow spectrum blue Kenos and bring them up to stop using your Minolta spot meter, you're in trouble. There will be blue wrap on objects two stages away.

>By far the most important aspect is that the screen be as evenly lit as possible (no more than 1/4 stop variation), and that it be all at the same temperature (no gelled light spilling over, no mixed temperatures at all). This gives the post person the best chance to pull fine detail out of the shot.

>Back, top, and side light should be based on the background image that will be matted into the screen. Remember that the back, top, and side light you see by eye which results from the natural bounce off the screen will be gone when the image is composited, because one of the functions of compositing is to remove color flare (which is what you see as back/side light by eye). It sometimes helps to roll in 12x12 black behind the foreground object (or just turn off the screen lights) to evaluate just how much fill you really have. What you see with the screen off is what's really going to be there after the composite is done.

>Remember that dark areas and shadows will fill up with the screen color, and make matting more difficult.

>Blue will give better flesh tones than green, but green requires less light, and so has become a producer favorite.

>The quality of the screen or paint is very important if you're not shooting with colored instruments, less so if you are. The best stuff is made by Jonathan Erland at Composite Components. He is the inventor of Digital Green and Digital Blue, and more info is available at www.digitalgreenscreen.com. One of the great things about his stuff is that his screens match his paint match his tape match his body suits. Do not allow prop or wardrobe people to use just anything of color lying around to cover areas that need to be matted out. They have a tendency to believe that green is green is green. It is not.

>Keep in mind that a lot of people are trying to peddle indescribable garbage as being "that digital green stuff." It is not. It is ripoff technology which will cause you no end of grief later. Spend a little more on the screen and paint and avoid a lot of problems.

>As a disclaimer, I have no financial interest in Composite Components, but his stuff sure has made my life easier.

>And finally, remember that no matter how bad it is, it can be fixed later if a sufficient amount of money is available to throw at the problem. I no longer tell people "that won't work." I say "that's going to cost a LOT of money to fix."

>Bob Kertesz BlueScreen LLC


>I agree with Daniel on keeping the matte at the lowest possible level thus avoiding contamination from the screen. If the system can do it, then why not?

>The problem with talking to the guys pulling the matte is that some know and some don't know the correct answer. I've seen this more than once when the post person recommends a level and then it's wrong. You may just ask the wrong person and get bad advice. Some of the people giving advice on lighting the matte have very little set experience but very strong opinions. I'm curious if the post people on the list have a consensous on suggesting matte levels with regard to T stop?

>The quality of the matte material is also of consequence. Hard cyc vs. soft material. Watching the angle of reflectance of the matte material in regards to lighting and field of view. And tungsten light contamination. If you have a matte lit with tungsten and then lit with gelled tungsten you will get different results with the same exposure relationship to the matte. And various stocks and formats affect the outcome.

>In my opinion, either have an FX supervisor who is responsible for the mattes on the set and follows through in the post house or shoot a test and base your stop on that. It becomes a matter of who is ultimately responsible for the image, whether succesful or not. That way one responsible person deals with the problems and not a consensous of ideas from a committee or person not on the set.

>I've shot tests and know at what range I'm comfortable with in getting mattes pulled. All the exposures are 1/2 to 3 stops under T stop, not over. But that's the way it tested that day a while ago with Kinos in NY. Perhaps I saw this on a system with lots of latitude but I know I'm not going to try 3 stops under before I test but if I can't test, I'm happy with 1 - 1 1/2 stops under (reflective) on both green and blue using a Minolta F.

>BYW - Just what does it mean when it's said that they had a hard time pulling a matte? Does it mean that hours were spent in an expensive post suite? Days of Rotoscoping? Or a few more buttons needed to be tweaked? I'm sure it varies but I'm very curious and uncynical with these questions as I've heard this said but never understood what people actually meant after all the technical discussion occurs on the shooting side of things.

>Just my 2 cents.

>Best Regards,

>Jim Sofranko NY/DP


>BYW - Just what does it mean when it's said that they had a hard time pulling a >matte? Does it mean that hours were spent in an expensive post suite? Days of >Rotoscoping? Or a few more buttons needed to be tweaked?

>It can mean any and all those things. When the screen is not lit well, or there is too much spill on the subject, or there is fine flyaway hair detail which is being lost due to no light on the subject other than the key, or the entire left side of that car bumper is blue, or the talent's tie has the same color green as the screen and no one noticed, then it is difficult to make a proper separation, even if the person doing the work is highly competent.

>Then it may be necessary to go in and clean things up frame by frame, which could take hours or days, or go to a suite with better tools, or a more knowledgable operator, or some combination of the above.

>Sometimes these things are difficult because the person doing them has no idea what they are doing, but has the confidence of the person paying the bills, and so is allowed to take 18 hours to do a one hour job.

>Sometimes a post house does not have the right tool to do the job because the right tool gets things done in an hour instead of six hours, and when you're billing hourly, where's the financial sense in that?

>The people paying the bills tend to go where either the hourly rate is lowest (or even flat), or where they feel most comfortable. This may not be the right place to do mattes, however, and they have a hard time making things work.

>And of course there is always that I'm-going-to-be-a-hero golden oldie "well, this is pretty screwed up, but I think I can save you..."

>--Bob

>Bob Kertesz BlueScreen LLC


>Well, if you overexpose it then you're going to increase your spill for one. And two it's not going to be as saturated, which is what you want.

>I lit a poor man's backdrop once for an EPK (read lo-budget) and we stretched a sheet behind the artist, I lit it with a couple of blondes (unfortunately they weren't naked!), stuck some diff and colors on them and overexposed to take the wrinkles out. Worked a treat. But I wasn't worried about spill or pulling a matte!

>Regards,

Shangara.

http://graffiti.virgin.net/eye.eye


>Bob Kertesz wrote :

>It can mean any and all those things...

>Yes, of course, and in hindsight I realize how broad and silly my question really is, however...

>The people paying the bills tend to go where either the hourly rate is lowest... This >may not be the right place to do mattes, however, and they have a hard time making >things work.

>This is all too often the case and probably leads to the most frustration. A technique that worked well on one job and not as well on another can be misleading.

>After being burned once on an FX driven job that went to a cheap house and was too cheap to hire an FX supervisor but took lots of misplaced (free)advice from the off-set post FX editor, I now often insist on an FX supervisor on-set for FX-intensive jobs. I'm not talking about simple blue/green screen shots but more difficult composites, etc. It's difficult to take direction from an off-set post editor when many don't really know the on-set 'nuts and bolts' issues. I recall the days when R.Greenberg was producing commercials in NY and insisted on sending out the FX supervisor for simple matte shots on every job. It was a pain but we never had a matte problem. He would constantly be the liason between the post and the on-set FX. He monitored the lab and the densities of the key every day, leaving the more creative elements to the DP. Because mattes have become so much easier to accomplish more recently, the FX super is often left out on smaller commercials, most often with no problem. But it is critical for a DP to realize when one is needed and to unabashedly recommend it.

>Best Regards,

>Jim Sofranko NY/DP


>When the meter reading are mentioned, are they all reflective? And if not, how do you deal with the different bounce angels of the blue/green screen

>Yariv.


>I agree with Daniel on keeping the matte at the lowest possible level thus avoiding contamination from the screen. If the system can do it, then why not?

>The problem with talking to the guys pulling the matte is that some know and some don't know the correct answer. I've seen this more than once when the post person recommends a level and then it's wrong. You may just ask the wrong person and get bad advice. Some of the people giving advice on lighting the matte have very little set experience but very strong opinions. I'm curious if the post people on the list have a consensous on suggesting matte levels with regard to T stop?

>The quality of the matte material is also of consequence. Hard cyc vs. soft material. Watching the angle of reflectance of the matte material in regards to lighting and field of view. And tungsten light contamination. If you have a matte lit with tungsten and then lit with gelled tungsten you will get different results with the same exposure relationship to the matte. And various stocks and formats affect the outcome.

>In my opinion, either have an FX supervisor who is responsible for the mattes on the set and follows through in the post house or shoot a test and base your stop on that. It becomes a matter of who is ultimately responsible for the image, whether succesful or not. That way one responsible person deals with the problems and not a consensous of ideas from a committee or person not on the set.

>I've shot tests and know at what range I'm comfortable with in getting mattes pulled. All the exposures are 1/2 to 3 stops under T stop, not over. But that's the way it tested that day a while ago with Kinos in NY. Perhaps I saw this on a system with lots of latitude but I know I'm not going to try 3 stops under before I test but if I can't test, I'm happy with 1 - 1 1/2 stops under (reflective) on both green and blue using a Minolta F.

>BYW - Just what does it mean when it's said that they had a hard time pulling a matte? Does it mean that hours were spent in an expensive post suite? Days of Rotoscoping? Or a few more buttons needed to be tweaked? I'm sure it varies but I'm very curious and uncynical with these questions as I've heard this said but never understood what people actually meant after all the technical discussion occurs on the shooting side of things.

>Just my 2 cents.

>Best Regards,

>Jim Sofranko NY/DP


>The problem with talking to the guys pulling the matte is that some know and some >don't know the correct answer. I've seen this more than once when the post person >recommends a level and then it's wrong. You may just ask the wrong person and get >bad advice.

I stand semi-corrected and agree, almost completely. On the last big effects show I worked on, we had seven months of FX. As I was driving to work this morning I remembered that in the beginning of the season we had been given the light levels on the green screen. Our Visual FX supervisor/second unit DP said he thought the level was about half stop too bright. After a little test we found that one stop under was appropriate. So, indeed, the guys at the shop were in fact wrong. But,at least now, they know.

>Simplified, pulling a matte means separating the foreground(the actor) from the background (the green/blue screen) by making them color-contrasty. You can pull a matte from any primary or secondary color. Blue background is desirable because it creates a blue matte line which can easier to time out of an actors face and it blends better if you were using sky as your background plate. Also, I think, when the matte guy makes his high con image,the blue becomes clear making the matte line cleaner (somebody who knows could help me out here as to the why or why not). Red screen is bad on actors because it is harder to get red out skin tones which, hopefully, have little blue in them. We all know bad mattes when we see them. It looks like someone took scissors and cut around the actor and pasted it on some background shot. Watch for hair, fingers, smoke and foot shadows. I'm no expert on the subject but I have found that key to a successful FX shot is that everyone involved has to be on the same page and in constant communication. One guy makes the call and the rest of us do what he says. On the same show, the art director picked the color for the green screen for some reason. When the FX guys came in they couldn't use that particular color green on their system. All that money wasted for nothing.

>Talk about starting off on the wrong foot. All the most basic questions need to be answered ahead of time by the right guy. What color green do we use? Shouldn't we be using a blue screen? Can we use this tarp from home depot? How close to the screen should we place the actors? Should we show their feet? What size screens do we need? Do I use green tennis balls or blue handballs for tracking points? Are we going to film or tape? Will rim-lighting the actors with magenta help the guy pulling the matte? If you don't have your act together you will be spending all your post time and money on cleaning up mistakes as opposed to doing cool FX shots that are supposed to make your show interesting instead of laughable. Pre-production is time and money well spent.

>Tom "Why do you think they're called FX supervisors?" Jensen

>or Tom "Why is the Weather Lady turning green when she points to Encino?" Jensen

>http://home.earthlink.net/~tomjensen/


>Thanks a lot Bob, for the great advice.

>One more thing.... What lighting units would you use to get the most even lighting spread on the screen and were would you place them?

>Thanks

>Yariv


>One more thing.... What lighting units would you use to get the most even lighting >spread on the screen and were would you place them?

>That is a "how high is up?" question. Remember in the following that I am not a lighting person, but I do like to have pre pro input on how, in general, things will be lit. And, the following is my IDEAL situation. Budgets, time, numbers of crew people, available instruments, stage height etc. will sometimes result in compromise :-). Also remember that the spacing I suggest below will vary wildly with different stock, stop, stage size etc., and is just a vague suggestion of where things might be placed.

>First rule: NO DIMMERS on the lights hitting the screen. A constant temperature must be maintained. It is very tempting to even out a large screen by dimming, but don't do it.

>In no particular order, the rest depends on a lot of things. The big question is, do you see the floor?

>If you don't see the floor and you're just lighting a vertical surface, anything on the truck that you like and has enough output should work fine. If the money is there, my personal favorite is always 3200 or color Kenos. Low power, cool running, easy to set, wide soft dispersion, easy to get very even (especially the color ones). Six to twelve feet from the screen on all four sides.

>If you see the floor, and the set is not on an insert stage (ie. has some size to it), I have had very good results with space lights. Setting them up and trimming them in is sometimes time intensive, but you get a nice even illumination on walls and floor. Twelve to fifteen feet from the back wall, six to fifteen feet separation. Have I mentioned no dimmers?

>All color temp hitting the screen should be the same. Don't gel the lights hitting the back wall if you're going to see the floor, because you can't gel the floor without getting the gel color on the subject, and that will be very expensive to fix.

>I have also had good luck for years with the brute force approach of sky pans or cyc strips on the back wall, and coops with diffusion for the overhead floor light.

>Your mileage will vary, objects in the mirror are closer than they appear, and always keep hands and feet inside moving vehicles. Have I mentioned no dimmers on the screen lights?

>Bob Kertesz BlueScreen LLC


>Respectfully, Bob, I feel that this is such a broad question that answering it specifically can inevitably be misleading. The first question I always ask regarding the lighting of a blue/green screen is if the floor is seen, as you mentioned. But the second most important question is how much depth of field and stop will I need. I want to light the screen to accomodate the necessary stop and, of course, this will determine which lighting units I plan to use. I always use the least amount of screen light to accomodate the key but to not contaminate the subject. The key level of the matte is directly related to the T stop on the lens. Once this relationship is determined it should be maintained, whatever the stop needs to be. This, to me, is real important as well as choosing the correct instruments with which to light in regards to whether I need to change the level of the key during the shoot.

>Bob Kertesz wrote :

>>First rule: NO DIMMERS on the lights hitting the screen. A constant temperature must >be maintained.

>As a gaffer in a previous life, I would certainly disagree with this statement but for different reasons than you imply. The reason to utilize dimmers is if you're using a deep green or blue gel for your key, it saves time and money to dim the lights between shots. This keeps you from having to change gels as they burn and fade during the shoot. If your not using gels, as in a 'seeing the floor situation' it's usually good to dim the lamps so that talent and the set will stay cool. Space lights, skypans, etc...can heat up a stage very fast and make talent very uncomfortable. Some people use a disconnect to turn off the lights but I find that the heavy load going on and off will more often than not blow out lamps and such. So I think dimmers are our friends in this situation but perhaps not for balancing the lighting. Although, WalterNY has made references to using dimmers to correct color saturation of mattes. I wonder if you're gelled for green or blue if you could make tweaks with dimmers that would not be noticable in the matte.

>A question that I come up with occasinally is how much latitude does one have in the subject lighting compared to the key level and T-stop in matte work? In other words, let's say I'm in a situation in front of a green screen and I need to, unexpectantly, gain a 1-1/2 stops for depth of field. I add 1-1/2 stops to the green, stop down 1-1/2 more stops on the lense and underexpose my subject in the foreground by 1 1/2 stops due to lighting constraints. Can I correct the foreground subject 1-1/2 stops up later in post without compromising the matte? Hmmmm...

>Thanks.

>Jim Sofranko NY/DP


>Respectfully, Bob, I feel that this is such a broad question that answering it >specifically can inevitably be misleading.

>Absolutely agree. That's why I put in so many disclaimers. What I posted was more to give the person asking an idea of what I have found to be the general and generic overall requirements for a couple of different sorts of setups. I tried to stick to basic setups.

>First rule : NO DIMMERS on the lights hitting the screen. A constant temperature >must be maintained.

>As a gaffer in a previous life, I would certainly disagree with this statement but for >different reasons than you imply.

>Yes, I have seen dimmers used for this purpose, and I understand what you mean. But I have also found that people sometimes find it hard to resist temptation, and every once in a while, someone will use them just to see if I notice....

>Although, WalterNY has made references to using dimmers to correct color >saturation of mattes.

>If you were heavily gelled, and made minor corrections on dimmer, probably not. It's not so much that spotty color temp will totally ruin the matte as it is that it just makes things harder (more time consuming and hence expensive) later on. Dimming green is less horrible than dimming blue.

>A question that I come up with occasinally is how much latitude does one have in the >subject lighting compared to the key level and T-stop in matte work?

>If you have the capability to do a multiple pass transfer, and the footage is reasonably steady, it should be relatively straightforward to do a pass for the screen and a pass for the subject. If you're in a facility where they have real tools for this sort of work (ie. a digital Ultimatte in a Spirit telecine suite), it is a breeze. SMA in N.Y. can do this sort of thing really well, if you're looking for someplace local. If it's a feature digital finish, multiple passes shouldn't be a problem either.

>Disclaimer: A good friend is one of the principals at SMA Video (but truly one of the most gifted Ultimatte operators there is).

>Bob Kertesz BlueScreen LLC


>First rule : NO DIMMERS on the lights hitting the screen. A constant temperature >must be maintained.

>I believe that although not ideal, the use of dimmers could be applied keeping in mind the following: 1) that you are aware of the dimmer curve, the sliders on a dimming board are not linear and vary from manufacturer to manufacturer and how the board is set, this will allow you to gauge how much dimming could be achieved on the slider without too much temperature drift. Of course I am not talknig large amounts, probably in the range of a bees dick or a gnats nasty.

>The above should also take into consideration the difference in relation to colours and tones of the foreground subject in relationship to the rear screen. For example, when I key a subject, the post op will be choosing a window of chroma screen variation, our job is to keep this window as small as possible to preserve the fine edges in post and make sure our subject is nowhere near the window. If however the subject is a long way from the screen colour then the dimming technique can be used judisciously as the post op can expand the keying window and our subject should not fall within it.

>Although, WalterNY has made references to using dimmers to correct color >saturation of mattes. I wonder if you're gelled for green or blue if you could make >tweaks with dimmers that would not be noticable in the matte.

>Because a gel of a deep blue is in fact transmitting deep blue only, in theory then I would think that you could pobably get away with murder (after testing of course) the colour temp change would not be linear at all, the warm spectum should not be transmitted. I think the amount of light should deminish faster than normal on the dim because not only are you dimming but you are dimishing the amount of the blue side of the spectrum and thus the total amount of light transmitted thought your deep blue gel (strictly theoretical).

>A question that I come up with occasinally is how much latitude does one have in the >subject lighting compared to the key level and T-stop in matte work? In other words, >let's say I'm in a situation in front of a green screen and I need to, unexpectantly, >gain a 1-1/2 stops for depth of field. I add 1-1/2 stops to the green, stop down 1-1/2 >more stops on the lense and underexpose my subject in the foreground by 1 1/2 >stops due to lighting constraints.

>This is probably fantastic as far as extracting the matte shape but what of the bottom end detail in your subject and the noise created from your underexposure and the fact that this has to be comped with a nicely exposed background plate. You'll probably end up with mud in front of lovely backgrounds.

>Nick Paton Director of Photography T/F 61-7-3862 3994 Mob. 0411 596 581


> Bob Kertesz wrote :

>If you have the capability to do a multiple pass transfer, and the footage is >reasonably steady, it should be relatively straightforward to do a pass for the screen >and a pass for the subject.

>Would that would mean that it is possible only to do in a lock off or motion control situation with two passes? The foreground can't be color or density corrected seperately from the background (in one pass)?

>SMA in N.Y. can do this sort of thing really well...

>Yes, I know they've got the goods. I've had some things done there but haven't had the pleasure of visiting myself. I hear it's a beautiful facility. I remember Dave Satin (the S in SMA) from days long ago when he went out on sets as the FX super. He certainly knew his stuff then.

>Thanks for the info.

>Best Regards,

>Jim Sofranko NY/DP


>Nick Paton wrote :

>This is probably fantastic as far as extracting the matte shape but what of the >bottom end detail in your subject and the noise created from your underexposure >and the fact that this has to be comped with a nicely exposed background plate. >You'll probably end up with mud in front of lovely backgrounds.

>I guess it would depend on the subject. I know that a 1 1/2 stop underexposure can be timed up ok in telecine but the grain and mudiness would probably depend on the contrast and tonality of the underexposed subject.

>Jim Sofranko NY/DP


>If you have the capability to do a multiple pass transfer, and the footage is >reasonably steady, it should be relatively straightforward to do a pass for the screen >and a pass for the subject.

>Not at all. Regardless of the scene content, while in telecine you lay down a scene correcting for the screen and ignoring the foreground, then lay it down again correcting the foreground and ignoring the screen. The pass which was for the screen is then later used to make your holdback matte (or hicon, or alpha channel, or whatever you want to call it).

>The procedure is GREATLY simplified if there is an Ultimatte in telecine, because the hicon can be generated and checked immediately during the screen pass, after which a processed and corrected foreground pass may be done.

>The foreground can't be color or density corrected seperately from the background >(in one pass)?

>Some places can, some cannot. This is dependent on your colorist and the tools available to them.

>SMA in N.Y. can do this sort of thing really well...

>>Yes, I know they've got the goods. I've had some things done there but haven't had >the pleasure of visiting myself. I hear it's a beautiful facility.

>He still does. That is the person to whom I was referring in my original post.

>--Bob

>Bob Kertesz BlueScreen LLC


>I believe I have read all the comments on this subject so far. I am curious if anyone has an opinion on depth of field. Better to have everything behind your subject go soft to eliminate dust problems? etc... Or better to go for better saturation and a crisp blue screen?

>Dan C.


>Dan C. wrote :

>I am curious if anyone has an opinion on depth of field. Better to have everything >behind your subject go soft to eliminate dust problems? etc... Or better to go for >better saturation and a crisp blue screen? I don't see this as an "either/or" set of >options.

As long as the subject is sharp, the subjects edges will be sharp, and a (hypothetical) uniformly blue or green background will look exactly the same whether it is in focus or out of focus. I would suggest that as long as the screen is at an appropriate stop and uniformly lit, you wouldn't even be able to tell whether it was in focus or not... The most obvious exception to this rule is if you have put tracking markers on the background for camera motion tracking in post. It is worth getting the digital folk to look through the camera at shooting stop so that they can actually see their tracking markers (if you can get them down to the set) as they will often ask for a specific type, shape, or size of marker based on success on a previous job which may or may not bear any resemblance to the one YOU are shooting now. The subject of tracking markers might almost be worth a thread of its own...but only almost.

>Mark H. Weingartner Lighting and VFX for Motion Pictures


>One more thing....What lighting units would you use to get the most even lighting >spread on the screen and were would you place them?

>Nutshell: 4 ft. 4 bank narrow band fluoros. spaced with 1-2 ft. gaps. Along the top... overlap off sides if you need to use entire screen, and along bottom unless you can lay mylar or mirrored plexi down. Don't hang vertical fluoros on sides as it will create dark hole in center. Treat it like cyc strips. Lighting the screen with this method works for 4 out of 5 shoots... very simple... has become rather routine (just test for saturation that approaches 255 blue, although on green 200 can be plenty). It's the fg lighting that's interesting and fun and is actually what creates a convincing comp.

>Mark Doering-Powell


>Mark,

>Maybe I missed something here, but what is the 255 and 200 reference?

>Thanks.

>Jim Sofranko NY/DP


> I am curious if anyone has an opinion on depth of field.

>>As long as the subject is sharp, the subjects edges will be sharp, and a >(hypothetical) uniformly blue or green background will look exactly the same >whether it is in focus or out of focus.

>In the non Hollywood world or for El Cheapo productions the studio's hired often have scuffed and tired painted backdrops and cycs. I find a shallow depth of field very handy in smoothing off the imperfections.

-Tom Gleeson


>Flame/Inferno (most digital comp software) can quantify colors as 0-255 Red, 0-255 Green, 0-255 Blue.

>0-0-0 is black 255-255-255 is white

>Let's say you're shooting blue-screen: it's the difference between the blue record and the other 2 that gives you a saturated hue... and therefore a better matte assuming your subject does not contain too much of that hue.

>A near perfect blue screen might be 10 Red, 10 Green, and 255 Blue. Most Compositors are very pleased with the Red/Green contamination as high as, say 50 or so, but the purer the better. The higher those 2 "contamination numbers", the paler the screen as it approaches white.

>Green tends to be less saturated since the screen is more "luminant" and accepts tungsten spill (does not absorb other colors as well as a deep blue would). It also puts a slightly more spill on the subject, but that's another thread.

>Yesterday I photographed a deer that was entirely too close to the blue screen but was trying to minimize the roto job on the hoofs (had to be on black rubber matt... we're hoping for some luminance matte). Night/ext lighting... dark! I am hoping not to hear from the Compositor about clean up.

>Mark "hand roto fur & motionblur... or suppression?" Doering-Powell


> lighting... dark! I am hoping not to hear from the Compositor about clean up.

>The following are unassailable Facts About Production and Post Production :

>1. Production will always be under the gun to get what they need to get in as little time as possible, amid a sea of compromise and chaos ("lead actor just spilled coffee on his jacket -- OK, that's a half hour..").

2. Editors will always complain about the footage they get. ("why didn't they just....")

3. Compositors will always complain about the elements they get ("why didn't they just....")

>These are facts. They cannot change. They will not change.

>Mike Most