                 
|
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 various 24p Options
A lot of people use the term "digital" and the format "DV" (aka "mini-DV") more or less interchangeably. But the digital-camcorder universe is far wider than just this consumer or prosumer format. There are about a dozen distinct digital formats available these days, all with different pros and cons, levels of quality, and, of course, price ranges. Here are your basic choices, divided into three main groupings representing progressive steps in overall quality and price: (1) Low-end, (2) Mid-level, and (3) High-end. The Digital Formats:(1) Low-end: Your options start with mini-DV, DVCAM (Sony) and DVCPRO 25 (Panasonic) at the lowest end of the digital scale. These formats are all related, and are very similar in terms of quality (actually pretty much identical, because the technologically is fundamentally the same). They have the fundamental advantage of being exceptionally cheap with a lot of bang for the buck, relatively speaking. (2) Mid-level: The mid-level formats, in roughly ascending order of quality, are DVCPRO 50 (Panasonic), and D-9 (a.k.a. Digital-S from JVC) which are also closely related technologies and very similar (more or less identical) in quality. And the extremely interesting variant from Panasonic called DVCPRO 50 Progressive. Then there is Betacam SX (from Sony) using a MPEG-2 compression scheme, and Digital Betacam (Sony) which is the highest quality in the Standard Definition digital formats (SDTV). (3) High-end: Best of all are the High Definition formats, now including the HDCAM and DVCPRO HD formats (and soon D-9 HD). The HD formats all have an aspect ratio of 16:9, whereas all the Standard Definition formats are usually 4:3 (but can be 16:9 as well). The most popular choice for High Definition shooting are the Sony HDW 700/A or "HDW-F900 CineAlta" camcorders. But Panasonic has now released a 24p multi-frame-rate camcorder model AJ-HDC27V, and more choices will become available soon (including D-9 HD from JVC). The higher up the scale you can get, the better the quality ....... and therefore the better the quality (especially) when blowing up to film, or for large-screen digital projection. The 24p Options:Up to recently, there are only two 24p camcorder options, namely: (1) Panasonic model AJ-HDC27V/V-M1/F multi-framerate camcorder and (2) the Sony HDW-F900 24p CineAlta camcorder But things are rapidly changing on the 24p front. In recent years, Panasonic in particular, has been very innovative in the digital camera market. Panasonic was the first to release a camcorder shooting 24p in the mini-DV format (2002) . And Panasonic now has 24p camcorders at all levels of digital production i.e. low-end, mid level, and high-end. In 2003, Panasonic released the AJ-SDX900 24P DVCPRO Cinema(tm) Camera, a 2/3-inch, 3-CCD, 16:9, DVCPRO50 format camcorder that can acquire images at 24-frames-per-second (price is US$35,000 approx.). The quality will be roughly equivalent to Digital Betacam. And Panasonic already has a 24p High Definition model on the market, their model AJ-HDC27V/V-M1/F multi-frame-rate HD camcorder (mentioned above). And just by the way, this special standard-definition format from Panasonic called DVCPRO 50 Progressive is probably the best option for a "poor-man's" HD camcorder. So if you want to get HD-like images, but at a SDTV price, then this is perhaps the format for you. It upconverts to HD very well ..... the company demos look very impressive indeed. But this format is probably a nightmare in post production, because of the lack of essential technical support and infrastructure for this somewhat non-standard format. It seems you need to upconvert to D5 or HDCAM to facilitate the normal post production procedures. Panasonic AJ-HDC27V, AJ-HDC27V-M1, AJ-HDC27FThe new Panasonic AJ -HD27V camcorder is a very interesting alternative to the Sony HDCAMs. It has about half the number of pixels per CCD compared to the Sony 24p camcorder. This is because the Panasonic pixels are more or less physically twice the size compared to the CineAlta, and therefore twice as sensitive. So this fact leads to some interesting pros and cons. Perhaps the biggest advantage is that the Panasonic CCD literally shaves perhaps $15,000 off the price of a new camera compared to the Sony HDCAM's. And you get another advantage, namely that the AJ-HDC27V is at least twice as sensitive to light (similar to having a higher ASA film stock in your camera). More about the sensitivity later (which, like I say, is just like ASA). On the downside with the Panasonic, is the image resolution perhaps? Although High Definition, the Panasonic has less resolution than the Sony CineAlta in absolute terms. Both in terms of pixels, and also in terms of the 720p structure as opposed to a 1080p "line" structure. So there are less "pixels" overall, but this is balanced out by Panasonic giving you more color depth compared to the Sony CineAlta. So that is the tradeoff ..... Panasonic gives you less "pixels" but compensates for this by giving you more color depth, or color resolution (i.e. it is true 4:2:2). The reality is, that both camcorders produce stunning images, and it is not the case that you see one clearly lacking in relation the other in a subjective sense. Only in the case where you saw both images projected onto a huge cinema screen, might you begin to see the differences in terms of overall resolution. Or at least, this is my guess anyway, as I haven't seen this for myself yet. But what I can see with my own eyes, is that the Panasonic images look great. But I don't have enough experience with the Panasonic to give an absolute comparison at this stage, unfortunately. Of course, you will tend to get greater depth of field with the Panasonic for the same lighting conditions. But this is either an advantage or disadvantage depending on your point of view. So this may or may not be desirable, although you usually want less DOF to give more of a (so called) "film look". In this case, you could add a lot of neutral density filters to cut down the light, forcing you to open up the lens aperture to give the desired amount of DOF. My biggest fear with this Panasonic format, called DVCPRO HD by the way, is for the post-production side of things. Last time I checked, it was a little bit up in the air. It is certainly a non-standard format right now (but it might catch up ...... eventually?). The camcorder always shoots at 60p, and the camera simply "tags" the relevant frames to give all other frame rates. So you need special equipment in post to interpret what you have shot, so to speak i.e. to extract the 24p you need from the 60p recording, or whatever you set your frame rate at. Panasonic is still trying to get these sorts of facilities in place in the post environment (and also in co-operation with other manufacturers of post equipment). So in these terms, it is relatively early days yet. The Panasonic reputation in the professional arena is also a point to consider. They have never had the reputation of the Sony's, nor as good technical support and backup (or at least, not in the past). So in a renting environment, that might put you at a disadvantage with this particular camcorder, because the industry "demands" Sony gear more or less? I personally think the AJ-HDC27V is in all likelihood, probably a very good camera, but this stigma is part and parcel of the deal in some (many?) people's eyes. Birns and Sawyer in Hollywood have AJ -HD27V for rent by the way. Bottom line ...... a great camera at a lower price (so great bang for the buck, so to speak). The format is coming late to the market, so it is difficult to say how established it will become in the long term, but I assume it will eventually. And it is the only "Video" camera that can shoot true variable frame rates (in a very similar way to a cine cameras). In this sense, it is unique. In terms of conversion to 35mm film ......... well, my guess it probably has enough resolution to look really good on a big screen. But this is an educated guess, rather than from my own direct experience. Panasonic told me their model AJ-HDC27V camcorder was at least 6db (or at least one stop) more sensitive than the Sony HDW-F900. They gave me an equivalent film-speed rating of E.I. 1000 ("A.S.A.") Tungsten, with clear filter, 0db, and 1/48th second shutter. I can imagine this is about right, as there are about half the number of pixels on the Panasonic CCD. So each pixel will be approximately twice as big as the pixels on the Sony chip. Twice the size, means twice the light-gathering capacity, so twice the sensitivity compared to the HDW-F900......well, this is just speaking in broad general terms assuming everything else being equal. Actually, E.I. 1000 is about a stop and a half more sensitive, but it is pretty much in the ball park more or less. And the equivalent film-speed rating is a bit subjective anyway, so some D.P's might consider E.I. 640 closer to the mark. It would probably be more appropriate for Panasonic to give an equivalent film-speed rating of E.I. 640 to 1000 ("A.S.A.") Tungsten, with a clear filter, 0db, and 1/48th second (180 degree) shutter. For comparison, the CineAlta at 24 PsF, 0dB, Clear Filter, 1/48th shutter (180 degree) is an equivalent rating of about E.I. 320 Tungsten when compared to a film stock. Compared to film emulsions, the sensitivity of the older Sony HDW 700/A camcorder is roughly equivalent to 400 E.I. (ASA) for Tungsten balanced film (3200°K) with your light meter set to 30 frames per second or 1/60th of a second. And compared to film emulsions, the sensitivity of the older Sony HDW 700/A camcorder is roughly equivalent to 250 E.I. (ASA) for Daylight balanced film (5600°K) with your light meter set to 30 frames per second or 1/60th of a second. The sensitivity of the new HDW-F900 CineAlta camcorder is slightly better than the older model HDW-700A. It is f10 at 2000 lux (89.9% reflectivity) with camcorder running at 60i. This is the equivalent of E.I. 640 (ASA) Tungsten, at 60i (and this is about E.I. 400 (ASA) for Daylight). This sounds rather fast compared to the normal film stocks that most people use. But it seems faster than it really is because the rating is at 60i (60 fields interlaced). You get a better comparison when the HDCAM is running at 24p (24 fps, progressive scan) like film. At 24p, the sensitivity of the CineAlta is equivalent to: E.I. 320 (ASA) Tungsten E.I. 200 (ASA) Daylight (All the above rating assume a clear filter, and a 0db gain setting, and normal 180 degree shutter of course) The shutter speed at 24p is the equivalent of 1/48th of a second (....so if using a light meter, set it to 1/48th of a second for the HDW-F900. But if shooting 60i, then set your light meter to 1/60th of a second (i.e. for the HDW-700/A, and also for the CineAlta if shooting at 60i). A ball park for the exposure latitude is 6 1/2 stops under & 2 1/2 stops over. So roughly the same exposure latitude as Vision stocks (more or less). Film is still a little better in the highlights, but performs a little less well in the shadows perhaps? And you can tweak the HDCAMS to pull even more latitude (or apparent latitude) than this. 24p, or not 24pEverything else being equal, there is no real disadvantage shooting 24p (except if the camera package may cost significantly more money perhaps). And movement tends to have a "stuttering" or skipping effect becuase of the relatively slow sampling rate i.e. 24fps. As for the advantages, then that depends on what you want to do with the final product. If it is to be converted to 35mm film for release, then it is much better to shoot 24 and Progressive, as you will get significantly less image artifacts in the conversion to 35mm. If you need to show the movie in Europe on video or broadcast on television, then shooting 24p is still a good idea, because of a relatively easy conversion to analog PAL 50i. Basically, if you shoot at 24p, you can convert that to any other film or video format relatively easily and relatively cheaply. If staying on tape, and the US is the main market, then shooting 1080/60i is quite O.K. (especially if this saves you money at the same time). But if converting to film, I'd definitely shoot 24p if you can. Choice of format is always a difficult question because there are so many important variables on so many levels.......if you tell me more about what is likely to happen to your project in post, and for later distribution and exhibition in a world-wide sense.......then I can give you better advise. Image EnhancementI've mentioned this before ....... but it is important to turn off all image enhancement (or "detail") in the camera when blowing up to film later!!!!!! Or have the enhancement circuits turned on, but with a very, very subtle setting i.e. turned way, way down. You'll want this type of low enhancement baseline setting for the whole movie anyway, as the degree of enhancement has a big bearing on how a big-screen digital projection will look, and how an eventual blow up to 35mm film will look later.If in the slightest doubt, simply turn off all image enhancement (a.k.a. "detail" or "sharpness") circuits. Or if you think somebody might change your setting behind your back, it is safer to have it well and truly turned off. You can always add enhancement later in post if you decide you need some later. Once recorded, you can't get rid of it. So don't it on your camera original tapes. But you can add it at any time later i.e. prints or sub-masters etc. Stuff that stays on tape for broadcast etc. will probably look slightly better with a little enhancement added later. Image Enhancement adds an artificial edge to make the overall image appear sharper (especially on a monitor or TV screen). This artificial edging can greatly improve the apparent resolution of inherently lower-resolution, small-screen displays like T.V.s and video monitors. But the effects tends to be objectionable with the large amount of magnification needed for big-screen displays, like digital projection or converting to 35mm film for theatrical projection. Overuse of Enhancement can lead to a “cookie cutter” effect, especially in areas where there is sharp contrast between elements in the image (like an actor framed against an open sky). Also, overuse of Enhancement makes it harder (or even impossible) to pull clean mattes in blue and green screen work. So special care needs to be taken if converting to film, or doing green/blue screen work. If in doubt, it is better to turn all enhancement or detail OFF. Detail, or enhancement, is amplitude driven – the stronger the signal, the higher the level of enhancement. The added edging can be either a white line or a black line, and usually some combination of a white and a black line. There are a lot of menu controls to set the level and type of image enhancement in the 24p CineAlta (in Paint Menus #7 and #8). For example: DETAIL LEVEL: Increasing the detail level increases the amount of overshoot on the detail edge (upwards for white, and downwards for black as seen on the waveform monitor) Suggest –35 as a starting point. DETAIL LIMITER: The Detail Limiter will prevent edge halos due to over enhancement The limiters limit the amount of edge enhancement, to prevent objectionable detail edging, or haloes, and helps to prevent the “cookie cutter” effect, especially around objects that stand out sharply from their background in terms of contrast differences. Detail is amplitude driven, so the stronger the contrast difference, the stronger the degree of enhancement added. As you increase the Limiter setting with a positive number, you are actually lowering the point at which it takes effect. Bigger positive numbers reduces the degree of black and white detail edging. [Note: Below software version 1.24, the Limiter settings are not stored in the scene files] FREQUENCY: Effects the width of artificial edging. The higher the number, the higher the frequency, the finer the edge. +99 is virtually invisible. High numbers are usually desirable for conversion to film or for big-screen presentations like digital projection. A +99 setting rolls out the black edge completely so you can't see it at all. And the white edge up-shoot is about 3.5 IRE, which is extremely fine. A very fine edge is way beyond the frequency capability of a downconverter, so new enhancement is normally added during a downconversion. CRISPENING is a (“coring”) filter that takes the noise out of the baseline (don’t use this without a scope to set it up properly). It can counter the effect of adding enhancement, especially high frequency edges. So if you are not careful, crispening can filter out the degree of enhancement you might be trying to add to an image. H/V RATIO tracks pretty well in the 0 to +35 and the 0 to - 35 range (otherwise you may need to adjust it). The HDW-F900 has Limiters rather than Clippers for limiting enhancement towards the extremes (in the strongest signals). A Clipper would make the enhancement edge wider, while a Limiter rolls it off. So using a detail Clipper, rather than a detail Limiter, is likely to make the enhancement edge wider. This is the opposite effect of what you are trying to achieve. A detail Limiter rolls off this edge, and provides a much better, or cleaner, control of the edging. The new limiters in the F900 are much better than the old detail clippers. Remember, if your are not 100% sure about this any of this stuff, better to play it safe if converting to film (or for green/blue screen work), and turn image enhancement OFF completely. That is safe, and you can't go wrong that way. Overall ResolutionAfter the conversion to 35mm film, you will see the image very greatly magnified. So instead of seeing it on a 20-inch diagonal monitor, you might be looking at the same image on a screen 20 or 30 feet across. I just roughly calculated that this is about 280 times magnified for an average-size cinema screen compared to your on-set monitor. So magnifying the image means you see both, more of the subtleties of the images (making it more impressive), but it also makes more obvious the inadequacies, artifacts, or other faults in the image in as far as they exist. Remember, any subtle problems, while being hardly discernable on your monitor, may be more obvious on the big screen? Of course, I'm only talking in generalizations, and it is hard to be more specific without actually being on a set to see what's going on (....like the doctor diagnosing the patient over the telephone). This may or may not help, but you could try a crazy idea of mine that just sprung into my head. Get hold of a magnifying glass (a reasonably powerful one) and hold it over your monitor and try taking a "closer" look at details in the image. This might help you detect almost indiscernible problems perhaps, or give you a better impression of what you might see on a big screen. 23.976 fpsIf you are in a NTSC environment, I definitely recommend you shoot 23.976p frames per second, and not true 24p, unless you have a very, very good reason to do otherwise. This rather strange-looking frame rate will make later conversions to analog NTSC video must simpler, easier, and cheaper. Converting 24p to 59.94i is very difficult technically speaking, but converting 23.976p to 59.94i (NTSC) is relatively simple and comparatively cheap. Most people shoot High Definition, not at 24p, but at 23.976p for precisely this reason. If you live in a PAL-based country, then shooting true 24p is perfectly viable. In either case, conversions to 24 fps film is a straight forward process. This only applies to conversions to analog video. So even if you are producing in a PAL country, this consideration may help with your world-wide distribution plans, if they include NTSC markets. Copyright © Peter Gray (2001, updated 2002)
Peter Gray
|
