|
director of photography, peter gray, dp, cinematography, dop, cinematographers, lighting cameraman, videographers, dv, high definition, 24p, digital films, HDW-F900, CineAlta, Varicam, AJ-HDC27F, 70mm, independent films, lighting directors, filmmakers, filmmaking, HDW-700AThe Big Issues in Choosing a Digital Format when Converting to Film
There are several important issues to consider when deciding on the best digital format to use in your production, especially if converting to film. Format choice is always a complex question, as there are many variables to consider in terms of quality, price, availability, flexibility, relative production value etc. Before we delve into these issues, please review my summary of what digital formats are available to us these days. Actually, the first big question to resolve is ...... are you staying on tape, or do you want to release on film? If the latter, this section is for you. NTSC or PAL:The PAL system has more scanning lines, and will give you more vertical resolution. It is a bit of a hassle to shoot PAL in a NTSC country, but it is possible. One of the main things to figure out is how to have a working PAL monitor on your set, or at your location (need multisystem). Everything else being equal, the PAL blow up should look a bit better quality compared to a NTSC blowup. But so many factors come into play, like how you shoot and light your project in the first place, and which camcorder model you choose, and which lab, and so forth. I'd only go to the trouble and expense if optimum quality is important to you and the project. I definitely wouldn't bother if you are staying on tape (i.e. not converting to 35mm film). In other words, I'd only go to the trouble if you are definitely blowing up to 35mm film and therefore need all the resolution you can get as a basis for post processing. WIDESCREEN ISSUES:If intending to releasing on film, then it is important to shoot with a 16x9, or "wide-screen" camcorder if possible. In other words, to shoot an aspect ratio of 16:9 using a wider CCD chip (which is equivalent to 1.78 : 1). People in Europe are lucky, as16:9 is pretty close to the shape, or aspect ratio, used for cinema projection across Europe i.e. 1.66:1 ...... in fact the digital camera is just a tad wider. But of course, digital camcorders with the proper wide-screen chip are more expensive, and therefore are more expensive to rent or buy. Also the choice of models is much more limited. Unfortunately, most of the low-end (and even mid-level) digital formats don't have a wide-screen CCD. Nearly all record an almost square image, or 4x3, like the shape of a standard T.V. screen (this is equivalent to 1.33:1). The usual way to get a wide-screen shape out of this type of camera is to crop the image top and bottom, and throw those bits away. So when you blow up what is left in the middle to match the shape of the cinema screen, it will be about 25% reduced in overall quality i.e. in vertical resolution. So it will be a bit softer looking with lower reproduction detail. This approach is to be avoided if at all possible. You have one thing in your favour if shooting with the PAL system. With PAL you have more vertical resolution to start with, and the cameras shoot 25 frames per second (or 50 interlaced fields per second) which is very close to film at 24 f.p.s. NTSC cameras run at 30 frames per second (or 60 interlaced fields per second) and have lower overall resolution (less lines) than their PAL equivalents. So it is nice if you are shooting and posting in a PAL country, or can find a way to shoot PAL in a NTSC country (otherwise be aware of NTSC-to-film conversion issues, especially motion artifacts and 3:2 pulldown problems). So to get a film print from a low-end, standard definition digital camera, you will have to degrade the image about 25% roughly speaking. You wont really notice this on a monitor perhaps, until you blow it up to film and project it onto a big screen in a cinema. Then you see everything the particular format has to offer, and perhaps more importantly, what it doesn't have to offer. But never fear, I've seen some memorable films shot on the DV format and released on 35mm film ........ so if you are considering that, it certainly can be done. Make a good, powerful, gripping narrative (or documentary) with great characters and story, and the means of production, or the "tools" you use, becomes less significant ...... and perhaps virtually unnoticed by the audience. I have another way of explaining this with numbers, that might be easier to follow. NTSC is a 525 line system ....... meaning 525 lines of vertical resolution. But you never see all those lines on the screen. We actually see only 480 of those lines comprising the image on the screen with mini-DV/DVCAM. To get a 16 x 9 widescreen image, you need to cut off 60 lines from the top and 60 lines form the bottom, leaving only 360 lines for the image that you will eventually blown up to film. That is effectively a 25% loss of quality (because of the loss of lines of resolution). To get American wide-screen for the cinema (1.85:1) you need to cut off even more of your available vertical resolution, namely 67 lines from the top and 67 lines form the bottom, leaving only 346 lines for the image that will eventually be blown up to film. This is a 28% loss of quality. Things are better with PAL, which is a 625 line system. You actually see about 575 of those lines on the screen (about a 20% improvement compared to NTSC). To get European wide-screen for the cinema (1.66:1), you need to cut off about 71 lines from the top and bottom. This leaves about 432 vertical lines for the image that you can eventually blown up to film. For 16 x 9 you need to drop 76 lines from the top and bottom. This leaves about 323 vertical lines for the image that you can eventually blown up to film. For 1.85:1 you need to drop 93 lines from the top and bottom. This leaves about 389 vertical lines for the image that you can eventually blown up to film. The improvement for all aspect ratios is approximately 45 additional lines compared to NTSC. This represents a general improvement of about 13% in vertical resolution (after cropping the image) compared to NTSC. ANAMORPHIC APPROACH:There is also a way to solve the aspect ratio problem with an anamorphic lens attachment on camcorders with a 4:3 CCD chip. This solves the loss-of-image-quality problem (or loss-of-vertical-resolution problem), but you need more care and precision shooting with such a lens attachment. I have a lot of experience shooting anamorphic (on 35mm i.e. CinemaScope), so I know it can be done. But you need the lens attachment (which costs about $750 here in the U.S. for DV-type camcorders). But it is much better to choose a camera that will do wide-screen without any modification at all. It is definitely the way to go, if you possibly can. Don't be fooled by cameras that have a so-called 16:9 switch. This switch does the cropping I've been describing above (and often "re-expands" the cropped image in the viewfinder artificially ...... to help with focus and so forth). But it is nothing like having a proper 16x9 CCD however. And there is a major operating problem with the cropping-the-image technique, as you have to try and mask off everything (including the T.V. monitor on the set) so you have a rough idea of the 16:9 frame, or the 1.85:1 frame etc., depending on how you decide to frame it. I've done this exact same thing with the Canon XL-1 on a dramatic feature ...... and to cut a long story, it is a real pain. You need to take special care with parallax-error problems, as the viewfinder glass is not on the same plane as the image. This means that as you move your eye in relation to the viewfinder, then the precise framing also changes. Same with video monitors and television screens (if you don't believe me, then stick a horizontal strip of tape on a T.V. screen or monitor, and move you head up and down and see if the "frameline" remains in the same place. Now ask yourself, is that mic or barndoor just in, or just out of the frame.) It should be a little bit easier with camcorders using a flip-out LCD screen, as you should be able to mask that carefully with strips of tape i.e. much easier than trying to mask a viewfinder. I think in most cases, the protective glass is thinner, thus reducing the parallax error. Of course, shooting the normal full frame with any standard definition camcorder is easy ....... but I assume most people will need some sort of wide-screen - but not always. People staying on tape will be happy with 1.33:1 (or 4x3) shaped images (....... until the advent of digital television eventually renders this aspect ratio obsolete). In the above discussion, I'm really talking about how to squeeze every last little bit of quality out of the chosen format, in order to make the best quality blowup possible. All the little bits count ....... and I think you will very likely see the difference in the final result. HIGH DEFINITION:My general advise is to go for the best quality format you can afford (or beg, borrow, or steal). This will give you more flexibility, both at the time of shooting, and later in distribution and exhibition. So for example, things are much, much better in the High Definition world. To start with there is a significant jump, or rather a quantum leap, in quality. Hi Def cameras rival the quality of 35mm film, and indeed surpass it in some respects (in sensitivity and latitude). Up to now, there is virtually only one High Definition camcorder available in the world made by Sony (the model HDW-700A). I've shot several projects with this camera, and I'm very impressed with it. Like analog NTSC, it also runs at 30 frames per second (or 60 interlaced fields per second), but Europe doesn't have a choice right now in a portable Hi Def camcorder (although this is changing as I speak, as the new 24p/25p/30p progressive scan camera from Sony are finally being delivered). But you have so much quality in Hi Def, that even if you lose a little converting to film, or any other video format, it is hardly noticeable. This is essentially the camera Panavision is modifying with their lenses, for George Lucas to shoot the next Star Wars prequel in Australia. Actually it is very similar in many respects to the older HDW 700A camcorder, especially the 60i part of it ..... but significantly, it does offer a choice of frame rates and scanning modes. Anyway, this Hi Def format converts to 35mm film like a dream, and to an untrained eye, you can hardly tell the difference (especially if you light and shoot it like film in the first place.). So apart from the camera hire, you can shoot with this particular Hi Def camera for about the same cost as any digital or video camera available out there ....... well, there is another factor to consider - the tape stock costs about $80 for 40 minutes (compared to DVCAM which is roughly $15 for 40 minutes). So not a big difference in overall cost at the shooting stage (maybe an additional 10%, say, as a ball park figure?), but a really, really huge difference in quality. In terms of budget, High Definition tends to catch up with you in post production however. Firstly you down convert it to any cheap video format you like (including mini-DV), so you can load it into a computer for editing purposes. And yes, you can edit HD in Final Cut Pro and Premiere based NLE systems. Then you produce an EDL (Editing Decision List) so you can do an online assembly and a color correction of the Hi Def originals. And finally the conversion to film ..... the most expensive part, by far. To give you an idea, this costs at least $65,000 for a 90 minute feature for a basic conversion with no frills at one of the the best (and most expensive) facility in the world right now (which is the Sony Pictures High Definition Center at the Sony Pictures Studios at Culver City). And there are rumours that their prices will increase soon. But other labs are slowly cropping up to compete with the Sony Hi Def Center in the price-verses-quality domain. I think we will soon seen an economic print-to-film process, more in the $20,000 sort of price range. I have colleagues in Santa Monica with a fully equipped High Definition Studio (called ADMIT ONE PICTURES), who can handle all the post for a really super competitive price compared to most other places I know about. They have full HD camera packages too ..... not one, but two 24p HDW-F900 camcorders. In fact, you could get the whole deal, with one-month camera rental, and all post to final film print stage for approximately $25,000 for everything (but excluding post sound). So this particular package deal would include a Hi Def camcorder for 4 weeks, the downconversion for editing, the online assembly with color correction (but not the print to 35mm negative as they don't offer this service - yet?). I can send more details to anyone interested in this approach. You can still do virtually all your post anywhere in the world, as you more or less only need to do the very last steps of making the masters at a facility like the one in Santa Monica i.e. after the editing and sound mix are completed. IMAGE ENHANCEMENT:And there is another important issue to do with image enhancement (or "sharpness" control). It is best if you can turn off , or turn down, the artificial image enhancement if converting to film. This "image enhancement" feature puts (black or white) outlines around things according to contrast differences in your image. This enhancement makes your images look sharper on a T.V. screen, but it looks terrible blown up to film, or otherwise seen on a big screen ........... as these artificial lines become apparent then!!! It can make objects (especially subjects/actors) appear like cardboard cutouts on the cinema screen. So any camcorder that you can control the degree of image enhancement is certainly the superior choice. So if you can get rid of (or turn down) artificial enhancement (sometimes called "sharpness") ...... the print to film will look just that much better. It is the same issue with digital projection of that image onto a big screen. You need to establish if you want the option to convert to film eventually? Or will you always just stay on tape, and release on tape? By the way, your D.P. should be able to exploit a lot of little tricks with lighting and lens craft and so forth, to make the blowup to film look just as good as it possibly can. The more factors you can exploit and control in this regard, the better it will look on a big screen. RACK FOCUS OR PULLING FOCUS:This brings me to another point regarding the smaller camera, concerning Rack Focus or Pulling Focus. All the lenses on the low-to-medium-priced DV and DVCAM-format camera, are incapable of "pulling focus". As you know, this is a standard technique used in all drama production (i.e. the job of the so called, Focus Puller). This means that when your actors approach the camera, or the camera approaches them (i.e. a dolly in .... or out) with anything other than a very wide lens, then the corresponding focus adjustment can not be made. The Focus Puller normally marks the various positions of the actor/s on the lens barrel, and pulls between these points as the actor/s move within the shot. This can be highly pre-arranged with extreme precision with so called "marks", or can be more open and spontaneous to varying degrees. The skill of the Focus Puller is to do with anticipation, timing and speed of the adjustments themselves (takes a lot of practice to do it very well). Anyway, none of the "cheaper" DV camera can do this, because the lenses are "electrical" or servo controlled, rather than mechanical (and in the servo world, there is no absolute beginning and end points). So the implication is that you have to learn to shoot without focus pulls whatsoever (especially a problem with drama production!!!). I had to do this on the DV feature I shot here in L.A. It impacts your storyboard and the types of shot that are possible. Basically, you have to work around this limitation ...... which means using much wider shots in general, and avoiding to-and-from actor/camera movements. So clearly, this can impact "production value" to some extent, and your cinematic structuring. In a nutshell, the upshot is you tend to shoot your movies pretty wide, and you have to avoid push in and pull out type movements, and/or avoid your actors approaching, or walking away from the camera (unless the shots are very wide). BEST mini-DV CAMCORDERS FOR "BLOW UP" TO FILMHere is the advise I got from Patrick Lindenmaier from the Swiss Effects Lab in Zurich in Switzerland. Patrick is a well-known authority about converting the DV format to 35mm film. His precise recommendation (in late 2000) were as follows: (1) Sony VX1000 (note 1000, not 2000), then Other cameras like the Sony VX2000 will do a pretty good job as well, although the PD150 has the edge in a good conversions to film for camcorders of that particular size, price and type. Bottom line if considering a blow up to film. You need to choose a lab and shoot some tests. Everything I'm saying above is pretty abstract until you see it with your own eyes. The choice of lab is critical to this process. I'd even start with the lab before choosing the camera. I don't mean to baffle anyone with all this techo talk ...... some people tune into this stuff and others not. But as you can see, there are several issues to sort out which could be to the overall benefit (or detriment) of your production, especially if converting to film, or considering large-screen digital projection. Happy Shooting!!!!!! Copyright © Peter Gray (updated February, 2002)
Peter Gray |