Shoot It On Video
The Panavision/Sony 24P
Camera
An interview with Nolan Murdock, Vice
President of
Product Management, Panavision (Part 2)
by Michael Buday, Stephanie Argy and Steve
Cohen
Go to Interview Part 1
Over the last several years, Panavision and Sony have been working together to modify Sony’s new 24-frame-per-second, high-definition video camera and optimize it for motion picture and television production. The most well-known picture to use it so far is Star Wars: Episode II, which was shot entirely in 24P video. In this, the second part of our interview, Nolan Murdock talks about how the new cameras are being used, how they might influence post-production and what the future looks like.
|
|
 |
|
with the Panavision/Sony 24P Camera. Photo by Steve Cohen |
|
|
Based on what you’ve seen, would you say that moviegoers, when they watch Episode II on film, will be aware that it was shot using a different medium?
I’m sure LucasFilm is going to do a huge marketing campaign, but if you just took somebody who didn’t know and popped him in the theater, he would look at it and say, "This looks a little different," but he’d never figure out why.
Did Lucas have to do anything in terms of changing his style?
From what we could gather, he moved the camera more. He was a little more dynamic in the way he approached things. He wound up using Technocranes a lot, and a lot of Steadicam – a lot of movement. Maybe because the camera was all self-contained and had 40 minutes worth of load, so that he felt he could run it more.
Can the camera run without an umbilical?
It can be totally on its own, if you choose.
What does it weigh?
In Steadicam mode, depending on which lens you have on it, it’s going to be in about the 20 to 22-pound range. In hand-held mode it’s going to be 23 to 26 pounds, with lens and tape.
How does that compare to other Panavision cameras?
A Millennium XL in Steadicam mode is about the same. In hand-held mode, a standard Panaflex is 3 to 7 pounds heavier.
|
|
|
 |
|
|
|
|
|
|
|
I’ve seen a lot. Obviously Star Wars is going to be a film output. We’ve taken a lot to film ourselves. We have discovered that there are two stages that we don’t have a lot of control over – first, how do you get the HD data converted into something that can be dealt with either in the computer or by the film output device, and second, which film output device do you use, how is it calibrated and set up, and how are you doing color-space conversions and those sorts of things. What we discovered is that you can send the material to five or six houses and get five or six completely different results. So we are encouraging people to go to very specific facilities to do I/O and conversion.
Can you mention which facility?
We are very pleased with EFilm. EFilm has a real understanding of the color-space conversion, and they are in the process of getting their I/O up and running so you can simply give them the HDCAM material, and they will do the color-space conversion, as well as the data conversion.
Normally one would deliver an onlined 24P master?
Correct. Cut Master.
When you say I/O, what do you mean by that?
Unfortunately, right now you can’t just pop a tape into a VTR and have it play right into a laser recorder. You have to do some data conversion. There’s a whole series of things that have to happen before you get it into a film recorder, and that’s where a lot of things go wrong, particularly in color-space conversion where you’re going from color differencing to RGB. It’s pretty tricky.
How are people recording sound? Are they using the on-board sound of the camera?
One hundred percent of the projects we’ve done have recorded sound onto the camera. Out of that, probably 20 percent used it as production audio. Only one show has not done a backup; every other show has had a backup along with the sound on the camera.
There’s no confidence head on the camera, so you can’t monitor audio, right?
Not really. You can get audio back, but you can only listen to the recorded track in playback.
I know that the HD signal supports ancillary data. Can filmmakers include additional information on the tape for later reference?
All of our lenses have very high-resolution encoders built into them. The new lenses all have that data coming out on a Universal Serial Bus port. So we can collect focus, T-stop and zoom – all of the parameters that you can set up in the camera – along with timecode and sync pulses and anything else. We can actually collect up to 16 channels of data.
And this goes onto the tape?
Yes, onto the tape. We use an audio track for it.
The user gives up an audio track?
Yes. We can use a serial digital adapter to provide two extra audio tracks, and we can also use that same adapter to encode ancillary data. We think it’s going to be very helpful when you get into big, match-moving, large-effects features, those sorts of things, where you know where the camera is in space and what it was focused on. For instance, we have already encoded the Panahead so we can get pan and tilt data. If you get Technocranes from Panavision Remote Systems, those are all encoded. So we can get boom information. We can get all of that – 16 channels of information.
How often does it sample?
It samples a lot. We actually only write it once a frame. But the sampling rate is fairly high.
How close is the relationship between Panavision and Sony?
Panavision’s relationship with Sony actually goes back to the mid-1980’s. We looked at the HDC100 way, way, back, and then when the HDC300 came out, we designed an optical viewfinder for it. We had a few in rental, along with some HDD1000 one-inch digital decks. So we’ve had an excellent relationship with Sony for quite a long time. It became more formalized, obviously, when we formed the partnership that we are currently enjoying. We work really closely with Sony. We have a direct line into the camera group, and it has proven to be extremely beneficial. They’re great guys to work with.
Are there any other high-definition-specific tools that are going to come from this collaboration?
Our major contribution will be continuing to do lens development. We have some new lenses that we’re working on that will be a collaborative effort.
Do you have competition? Are there other companies renting high-definition cameras?
Sure, there are other rental houses.
And are they modifying their cameras as well?
No. We’re the only ones that we think have approached it as a true cine system, and we’re the only ones with the optical expertise to do the lens design that we think is critical to make the camera really work. Canon and Fujinon have to build a lens to a price point because they have to sell it. And since we don’t, we can design in a lot of things that need to be there.
|
|
 |
|
Murdock stands in front of an automated 3D milling machine, holding a blank block of metal in one hand and the final machined part, produced by the machine, in the other. Photo by Steve Cohen |
|
|
How much of the lens do you actually make in-house?
The only thing we don’t do is grind the glass.
But you specify the glass.
We spec the glass, we spec the coatings and we do all the prescriptive work. We just don’t physically grind the glass.
Do you have a sense of the economics? If a show – Titus, for example – converts from film to tape, are they sharing with you how much money they’ve saved by doing that?
We’ve heard everything from $10,000 to $18,000 an episode. That has come almost exclusively from film savings, stock, processing and telecine.
Is the camera rental less expensive than it would be for a film camera?
More.
A lot more?
It is somewhat more. You also need HD monitors, waveform monitors and other assorted boxes that are not supplied on a film show.
So the economic advantage is you’re trading off telecine…
And stock and processing.
For higher rental fees. Is that basically the formula?
And I think the post-production costs right now are also slightly higher in HD. So, you ultimately have to balance how much stock, processing and telecine time, versus extra cost on camera rental and post-production.
Now that you’ve got a few shows under your belt with the camera, what’s your sense of the future? Do you see these cameras giving film a run for its money downstream? Do you see this eclipsing film at some point or will they just continue to co-exist?
I think there’s going to be a great co-existence for quite a while. I see episodic certainly as being the leader in making changes. In episodic television, there are compelling economic reasons to make a change, and the quality is very, very close now. So it becomes a balancing act – economics versus aesthetics. If you get a large producer, who is making multiple hours of television programming, that’s even more compelling than somebody who’s just got one show – even if you’re only saving $10,000 an episode. I think feature films are further off. I think, and we’ve already seen, that there’s going to be a co-existence of a lot of film in a feature and some HD. For instance, we just finished a show, Simone, which is primarily film, but which also has a fair amount of digital work in it.
Why are they shooting with digital on a film show – because it makes effects easier?
They’re shooting digital for a couple of reasons. Part of it is the effects, and part of it is the look. And there are some things that the camera does well that, truthfully, a film camera doesn’t do, such as shutter angles wider than 180 degrees, to produce motion blur. You can set the camera for a 30th of a second integration time, which is equivalent to a 220 or 230-degree shutter angle.
A lot of motion blur.
A lot of motion blur. And you can do that in-camera.
So you think you’ll see shows that mix and match film and digital video?
Right now we only have one exclusively digital feature shooting – it’s starting tomorrow actually – and we’ve done, as I say, several others previous to this. But mainstream, ‘A’ features, I think, are still a ways away.
What’s going to keep film alive?
I think there’s certainly an aesthetic for film that you can’t get yet in hi-def. There’s the organic feel –how do you define the look of film? It’s soft, but it’s also a very high-resolution medium. And certainly I think Kodak will try to raise the bar.
Because these high-definition cameras are stuck at ASA 320?
Yes, that’s right. Now, obviously you can do things with gain and a few other odds and ends that make the camera fairly sensitive, but there’s not a lot of speed without other compromises – either noise or motion blur.
How do you see film companies like Kodak adjusting to this?
Kodak certainly is going to continue doing film stock development, but you also see them making moves into the digital business even more heavily. Cinesite is obviously gearing up directly for the digital intermediary process. Again this goes back to the I/O. The I/O is so critical – how you get this material in, how you do the color-space conversion and how you record it out. I think Kodak sees a huge potential for that. They’re going to be opening an Image Technology Center over at Cinesite that deals with film technologies as well as I/O capabilities. It will be a learning lab, which I think is pretty exciting.
Are there questions that you would want to ask editors?
The biggest issue in our mind is frame rate. From an editorial standpoint, when you go to lay back audio and that sort of thing, how you would like to see this material originated? The camera can run at either 24 or 23.98 fps. If we record audio at 29.97, is that a problem?
My thought is if you’re going to mix and match film and video at the same time, you want everything running the same way. As long as everybody does the same thing – the sound is pulled down, the video is pulled down – you’ll be okay. Everybody should be in the same world – either the full rate world or the pulled-down world. Film, video and sound. If all the rates get adjusted together – it shouldn’t matter and everything will run in sync.
Our argument is we should be shooting film at 23.98, because it gets pulled down in telecine anyway. The second you throw a piece of film on a telecine, it’s pulled down to 23.98.
You’d have to think about what the Avid would do in that case.
That’s the thing we don’t know about from Avid. They’ve announced 24-frame HD editing devices, we understand that they delivered a prototype, but I haven’t seen one working yet.
In terms of our workflow, that wouldn’t really matter. Whether you were onlining conventionally, the way we’ve been doing it, or doing a non-linear online through a fancy, future version of the Symphony, it wouldn’t change our workflow in the cutting room, because for the foreseeable future we’re not going to be cutting full-res material anyway. But do you really want to go out there and advocate that everybody start shooting film at 23.98 instead of 24? Do you guys have that much clout?
That’s why we haven’t made a huge push. But we sit around a table like this and as far as we can see, there’s no downside to shooting film at 23.98. If you do that, you must record audio code at 29.97, so everything is running pulled down. It would be even better if we could record audio code at 23.98. Some sound gear won’t do this and some of the post-production equipment won’t work with it, either. But if we could switch to 23.98 across the board everything would be a lot simpler.
What about down-conversion? You need to create normal NTSC tapes for offline editing and viewing. That’s normally done in a transfer facility.
Yes, but it can be done during production as well. You can feed the down-converted signal live into an NTSC deck and that becomes the editorial tape. The problem with transferring from the camera to NTSC is timecode. The monitoring down-converters don’t do proper 3:2 pull-down. But that’s changing. Evertz makes an HD version of the Afterburner, which will add appropriate 3:2 pull-down, track the 24-frame timecode and generate 30. We have some of them, and that’s exactly what Lucas used.
While you were shooting?
Yes, absolutely.
In other words, you avoid dubbing and at the end of the day, you have not only your HD master, but also an NTSC tape for cutting.
That’s right. It goes back, though, to this whole issue of frame rate and what you do with audio. There’s a picture delay through the Afterburner and we haven’t been able to find a digital audio delay unit that can accommodate it.
The timecode will be off, too.
Right.
This has been a big problem not only for the camera, but for the HDW-F500 VTRs, as well. I guess the camera does not have a built-in down-converter chip or anything.
No, it doesn’t. That’s one of the big problems. And the other problem, of course, with the F500’s is that the SDTI card didn’t exist for a long time. It’s only fairly recent. [The Serial Digital Transport Interface is used to directly input, output and transfer the native, compressed HD signal without degradation.] This is another thing that Lucas used. They were down-converting in real-time, but the second they got a tape out of the camera, they also cloned it for protection.
The HD tape?
The HD tape. We set them up with SDTI to SDTI between two decks. So they were actually making cloned masters immediately.
You would never do that with film, because you couldn’t. But is there a sense that tape is more fragile? Or it’s simply, ‘we can so we should’?
You can so you should. And you know what? There was a completely unexpected savings in camera negative insurance because they had multiple masters. They saved $100,000. All of a sudden the underwriter said, "Oh, you’ve got how many copies?" No one had thought about that.
So if we came back here a year from now, what do you think you’d be saying different?
I think I would be saying that we’re doing a lot more episodic television shows and we’ll be doing more and more features. But I don’t think you’re going to see any more major changes in camera technology for another couple of years.
Michael Buday is an offline and online editor
as well as a consultant to Sony Broadcast.
His HD online credits include 'Dane', 'Family Law',
'Judging Amy', 'James Brown Live at the House of Blues'
and 'Touched by an Angel'. He can be reached via email
She can be reached via email
the publisher of the Guild Magazine. He recently completed
'15 Minutes' for New Line. He can be reached via email
Reprinted from
The Motion Picture Editors Guild Magazine
Vol. 22, No. 2 - May/June 2001
Guild Home | Magazine Home | Top of Page
Copyright © 2001, All Rights Reserved by The Motion Picture Editors Guild, IATSE Local 700