[Mark Suszko]

Yes, Alan, if the projectors all run the same image.

He’s talking, as far as I can tell, about three independent views, say, front and side views from a flight simulator pilot’s cockpit perspective, or some kind of semi-wrap-around virtual reality setup, say, for simulated target practice with light guns, where each screen is running a different picture. If you want to put one image up over three screens, say to make it really large, you can build a huge comp in something like AfterEffects, assign carfefully cropped versions of each section from that master comp to each screen, and use screen-blending boxes to throw that huge composite image across the three screens acting as one. But I’m thinking the usual graphics cards in a mac pro tower can only do up to two such screens. To make three independent AND interactive feeds, you are going to need at least two video cards, or more likely, multiple computers networked toegther with one of the screen-control presentation systems like AV stumpfl Wings or something. But that’s just my guess, the AV guys in the stage forum will have a better grasp of this because they do it every day.

[Mark Suszko]

I don’t know the actual method for this case, you *could* just shoot the real thing and motion-stabilize the entire frame. This is hard to light well yet make the screen really readable. Once the footage of the unit is tracked, you can put anything you want on that screen, though, even an exact but cleaner copy of the same exact screen. The hand/finger would need to be roto’ed first. I think if I was shooting this I would light it primarily for the overall machine and hand and shoot my footage, so the finger actually moves and interacts with correct timing. I’d then stabilize the frame and motion-track the screen, and probably replace it with an identical screen that had been pre-comped to look as darned good as possible. You can also do this in reverse; i.e. precomp all the web pages and such, then just rehearse the moves tha hand model makes until it looks like the finger drives the display and not vice-versa.

Keeping the screen on a separate layer also lets you have more flexibility in color grading and exposure control. I’m guessing though that if you do it this way you have to put a tiny legal disclaimer caption on the frame somewhere to say the images are simulated. If the spot had one I didn’t notice it. So maybe it’s all really as you see it, and no special tricks except good lighting and motion stabilizing.

[Mark Suszko]

Your question is actually how to get three different pictures into three projectors off of one computer video card. I think the best experts to help you with this question are in the live/stage forum, this kind of thing is their bread and butter. I’m pretty sure you’re going to need more than one computer though.

[Mark Suszko]

In a sense, there ain’t no such animal. If they want to put it on the air, it is a “broadcast format” even if it is poop. See Bob Saget’s show on ABC. Every station has its own mix of equipment on different standards, the last “lingua franca” you used to be able to count on was analog betacamSP and maybe digital betacam. I’m kind of hoping the next “standard” will be an HD signal on BluRay myself, because players and mass dubbers will be cheap, cheaper than tape equivalents, the media is cheap and durable, and it will be readable on multiple platforms, from game systems to stand-alone players to computer ROM drives. We get the benefit of all that if the consumers get excited and buy BD in mass quantities. Right now they are not, so BD may wind up like Mini-Disc: a lot going for it but not enough user base to keep it alive. C’mon, Apple, get on the BD bandwagon and help me out here!

You can look up CCIR601 and other standards in the SMPTE web pages, but that doesn’t get you closer to what I *think* you’re asking for: the “magic decoder ring” format choice that is universally playable at any station in the USA. MPEG2 might come close today, since worst-case you can buy a DVD player for $30 at the corner Walgreens and find a way to hook it up analog composite to the station’s racks, or ingest it into a computer and cross-convert that file to the local preference. We picked DVCPro 25 when it came out ahead of Sony’s DVCam because it was a shipping product and we needed it NOW, couldn’t wait for Sony to catch up, others in our area used it, and it could play consumer DV mini-tapes as well as DVCam tapes; this was as “cross-format” as we could afford at the time, and it has served us well up to today. But your market may be really diferent so you can’t go and judge by our success with the format. And that’s just SD. We’re looking at DVCPro HD decks to go with our Panasonic cameras and FCP workstations because that all plays well together, but how many outside users will be able to take HD off a DVCProHD tape, we can’t yet know. I’m hoping to dodge the whole issue by using FTP to move the files to those stations that can handle that, and BluRay for the rest, and for deep offline archival.

I’ve been trying to find a standard for our shop for several years, thru annual polls and surveys of stations across the state, and IMO there just has not been enough time for one really “universal” standard to shake out and become the “next betacam”. (This year I got a request for some stock footage from a major network and their producer still preferred a betaSP dub over DVC Pro or DVD, go figure. The installed user base remains strong and hard to overcome, like old habits.

Due to fragmentation of the markets, it may never happen like that again anyway. On top of that, many stations will keep the analog gear they use today and just run it into a magic box just ahead of the transmitter to be digital-legal after February. It takes time to replace years worth of physical plant, and you know the money men like to keep old stuff working until it melts, as long as it has already been amortized and still gets a job done… so my own sense of it is that many stations will still carry a wierd mix of internal standards for a while yet, until they can afford to tear out the pre-2008 stuff and start from scratch with the latest thing. Perhaps a safe strategy is to pick the codec that is easiest to transcode or render into another format while retaining quality (If you know what that is, please write me, I promise, I’ll keep your secret).

I think you’re going to have to do it the hard way and call every station you want to send to, ask for the Engineering or Traffic desk and talk to the actual people that actually handle incoming video about what they prefer and what they’ll take if it’s important enough. Don’t ask the sales department because as long as you pay they’ll take grease smears on acetate or old quadruplex, they really don’t care about the technical details. Other station departments may also not really know, and give you a generic answer like “quicktime” which, as you point out, is a wrapper that can hold any kind of codec inside. That’s no help.

No, you need to talk to someone who still knows what a “greenie” is, what a vectorscope is, and what you do with it. They are somewhere behind an overstuffed patching rack in the engineering department. They’ll tell you what the station can really work with.

[Mark Suszko]

There are many systems for this that basically use a local hard driveor flash drive/video card in a tiny box to feed a display. The hard drive can be updated by an internet or ethernet connection. Typically these do their loading of new info overnight, so even a slow connection is not a problem.

Sony makes some of these, simplesign and BKM-FW50 come to mind, I think Panasonic might, look for NMstage by them, there’s a player called mantis, and I remember a maker called DoReMi, Aplhavideo and DVEO and adtec digital soloist that offer a solution, there are dozens more you might find thru a careful googling of digital signage, video servers, and iptv as search terms.

[Mark Suszko]

Don’t over-do the magenta, and don’t do it at all if specifically using Ultimatte. It screws up Ultimatte’s special version of the process, and Ultimatte AdvantEdge has it’s own artifical spill suppression built in.

If you have enough distance between the green and the talent, and didn;t over-light the screen, you shouldn’t actually need any magenta, it’s only needed where green reflection spills onto the talent. If you overdo it, you’ll have the devil’s own time trying to get rid of it later.

Been there, done that, got the t-shirt.

[Mark Suszko]

They cover it pretty well witha mini-how-to documentary in the DVD bonus materials, you can rent the DVD for like 2 bucks. In short, particle effects and a lot of compositing, BAMPF!

[Mark Suszko]

Point number four of Terry’s reminded me of a tutorial I saw in a college film class, the example was some race scenes in the Robert Redford film “Downhill Racer” that illustrate what Todd was getting at perfectly. You sometimes have to “lead” the eye to create an anticipated motion and the cut actually works better that way than if it was more “scientific” in the matching action.

I don’t have too many more tips to add, just that tape is cheap unless its blank when you’re done, so in a situation like this I would roll nonstop just in case something magic happened by happy accident. Meaning, roll on ALL the rehearsals, even if you know the first couple of run-thrus are going to be rough.

They give you practice in positinoing the camera and anticipating moves, they also help you fine-tune the blocking of the shot, looking out for things in the background that should not be seen, etc.

They give you a chance to grab some good closeup reaction shots of each person, adequate room tone, extra and alternate versions of dialog that could save you later, etc.

You might shoot the rehearsals with the red tally light disabled to trick the talent into relaxing because this is “just practice”. I often get the best material in such faux-candid moments.

Pre-blocking the action on paper, and knowing where you are using what camera lets you speed things up and not waste time on camera and lighting setups that won’t end up in the final cut.

You start to get an eye for continuity over time; having OCD or a tendency to be overly critical of others in this case is a plus. One thing you might want to try to help you get all you need is to add some columns or boxes in your paper script where you can check off multiple takes of each scene as you capture the wide coverage, the tight coverage, the reaction or cut-away shots, the room tone, shots of a color bar chart (not camera bars but actual footage of bar charts under your lighting conditions).

In the bustle of production it is easy to skip a scene you thought you already captured, but really only rehearsed, or only got from one angle, so as things in my script are shot I like to run a sharpie or highlighter thru each paragraph or block of lines in my copy of the script to denote it got covered. You can go with multicolor highlighters and perhaps use one color for each character or each camera angle needed. This gives you a very plain illustration by day’s end that you got everything you needed, if the shooting script has been completely lined-thru in every page. Anything still pure white is missing or deliberately deleted.

Also don’t be afraid to slate takes, especially end slates. If you don’t have a slate, get a cheap $5 student locker white board and erasable marker, and write the page number and scene number and any other key data on that board like audio assignments or camera settings, and hold that up at the beginning of the take, or upside-down at the end of a take, (called “end-slating”) if you forgot to slate the beginning or came up with something extra to write.

This can be a huge help when digitizing later. Of course more high-end and high-budget producers have better and more technological means to capture this information as metadata and put it into spreadsheets and databases but even simple slating on-camera is still useful to everybody at any level.

[Mark Suszko]

Oh man, this is a hot one, optically speaking. The first thing I would want to think about is to use a lens that most closely approximates the field of view and other characteristics of an average human eye. No zooms, and the biggest issue perhaps for this case is the mechanical lens and the human perception of what the eye give it are radically different in terms of depth-of-field perception. If you can’t get these two issues locked down in a scientific manner, the other lawyer will easily get your footage thrown out as inadmissible.

[Mark Suszko]

Well, you might have one other trick available, if you shoot HD but deliver in SD, the HD wide shots can be tightened-up a lot in post and still work for SD. I would say you could zoom into a hi-rez HD shot about up to 25% and get away with an “acceptable” (YMMV) SD master, but that would be pushing it.

I have faced a similar situation when I have to shoot non-pro acting talent in skits and the like. I shoot that multi-cam with a locked-down wide cover shot and the close-up cam placed as close to the wide cam as possible. I jam synch the time codes so both shots have the same code. If the talent gets a take right, which is rare, O have my wide and my tight already, and can move on. When the talent is untalented, or the action hard to reproduce in some other way with precision, double-coverage is cheaper in the long run over trying to single-cam it. The time saved in shooting and in post pays for the second camera.