Home of Professional Cinematography since 1996
>I just got back from the monthly ASC meeting here in Hollywood (I was the guest of a member) I saw "Kodak's gray card / TEC lineup film system" demonstrated there. It purports to give the DP "printer lights" from his negative. It requires you shoot a special gray card "in the illumination of the scene" (whatever that means!!!)
>Have you heard of this system and what do you think of it? I haven't used it, but the problem I see with it is that it depends on the "proper" exposure of the gray card. What if the scene is heavily backlit? What if your main characters are illuminated by bounce light coming off the white tablecloth on the table in front of them.
>The Hazeltine film color timing machine did not depend upon a gray card's presence in the frame. I think a system has to be devised for telecine that also doesn't require a gray card.
>In theory it's a great idea, unfortunately, as you say, it depends on the insertion of a grey scale and that has to be quite large in frame. Now I don't mind including a grey scale and lighting it to a level I think is correct, it's how I try to get the video dailies I want now. At each new set-up I try to shoot a grey scale in controlled light and get them to grade to that, hopefully the scene will then look as I intended!
>There is an Aaton variation to the Kodak system that looks very interesting, you can have a very small area of grey in the scene and they'll be able to put frame grabs of each scene together with your printer points onto a web-site.
>I read Bill Bennett's note on the Cinematographers site. This system was designed to help cinematographers to determine their exposure when a Hazeltine or other color analyzer is not available (the case today with much of the film being directly transferred throughout the world in telecine houses without an adjoining print lab). Also, this should help starting cinematographers who have never seen their work printed (unfortunately quite common now).
>Bill is right-on when he says that the gray card needs to be exposed properly( BTW ANY 18% gray card will work with the system, ours just has a good low reflective surface as well as a specific black and white patch for consistent tone scale reproduction in telecine transfer). In the instructions we have described in detail five different uses of the card besides the exposure determination. Specifically, in the cases he sited, a meter reading would be taken in the normal manner and then the gray card would be lit to 3200K and exposed as per your meter (all OUTSIDE of the scene). This provides a reference gray, evenly illuminated, for exposure determination AND subsequent grading. The actual scene may not be 3200K for your artistic work, however, if the telecine operator is told to grade the gray card to "neutral gray" then the scene's color will be maintained. This is important for unsupervised transfers.
>The problem with the telecine is that there are too many knobs (degrees of freedom) to attain a "look". Also, there are a few telecine manufacturers, each with a different slant on the hardware. We tried to provide a system that uses the telecine as an expensive densitometer which will be consistent from LA to London to Japan. As far as the Aaton system (they call it GrayLink which links into their KeyLink system), we have worked with Jean-Pierre Beauviala of Aaton on this. Actually, their system uses our calibration film as well as our calibration tables, they added hardware with excellent software to speed the exposure determination. So essentially, the systems are the same (that is the answer you get should be the same). Cintel also has a system the TKG system which again is based on our calibration film and tables. The Aaton and Cintel systems work a bit differently and it is up to personal choice as to which is preferred.
>I've had great luck doing the following: I shoot a gray scale full frame, "properly" exposed for whatever ASA I've been shooting at. Then I pull back to include the set, set the aperture to my shooting stop, and put the gray scale in the corner of the frame lit to the shooting stop. In essence you are saying to the colorist, "Okay, here's the chart full frame: tune up on this. Now I'm showing you the chart exposed the same way with a properly exposed set in the background; this is to show you that I'm not crazy and that I actually want it to look that way." Nearly every time I've done this dailies have come back exactly the way I wanted.
>Kodak Telecine Printer Light System- I was involved in earlier BSC demos of this system 18 months ago and again at Madrid Imagen this year.. Sure there are problems. its just another tool it doesn't cure cancer. Your exposure meter has to be turned this way and that for you to evaluate scenes other than pure front lit frames. At least its a system that gives feedback about exposure when telecine is the only way the negative is being viewed.
>The Aaton system is more sophisticated. jean Pierre's system allows DP's to log on to the lab site and download thumbnails of each scene and pseudo printer light info with it. If you want a Hazeltine system then telecines will have to be stopped and started which slows down the dailies process as telecines work in real time. Rotary printers work at much faster speeds thus allowing labs to turn out mass dailies overnight faster than any t/c house.
>Great idea. Boy have we all been there before, with the colorist trying their best to "save" the shot. Communication is everything. Also at the ASC meeting last Monday (I was the guest of a member), there was a member demonstrating the lengths he went to describe his "look" to the lab. It went like this>
>He has two Mac PowerPC computers. Each with the same type color monitor. One he has with him on location or at home wherever he is. The other he takes to the telecine facility he is going to use. His has a video frame grabber card and a CD ROM recorder on the computer with him. The one at the lab has a CD ROM player. Both have Adobe Photoshop installed. Both computers have the "ColorTron" monitor alignment tool that matches the two monitors -exactly-. (It is a probe device that suction cups to the front of the monitor and a software module that reads the probe and tweaks the monitor.) You need to set up the monitors with this tool every day. This way he is assured that his computer monitor on location looks exactly like the monitor at the transfer facility.
>Before starting production, he takes his daily transfers of tests of the set, wardrobe, actors, etc. He "grabs" frames from each setup. He uses Photoshop to manipulate the color, density, contrast of the images. He works with the Director and Designer at this phase. When everything looks correct, he records the images on to a blank CD ROM. (Each blank CD ROM holds more than 600mb and costs US$6 each) He sends the CD ROM to the transfer facility with the next day's dailies. There they put it in their computer and look at the images on their calibrated computer monitor and can easily see exactly what he wants.
>This is a lot of effort to communicate EXACTLY how he wants it to look, but he goes to this extreme to get the look he wants. Each day, he grabs frames from the dailies that he feels that could be done better, tweaks them with Photoshop and sends them with the film to the transfer facility, constantly giving the colorist feedback. In the future the time will come where the images can be quickly be sent via high speed modem to and from the lab. Aaton is working on a system the uses a web site to do this. What he is taking advantage of here is the digital computer Desktop Publishing developed tools to match computer monitors. This is something that can not be done in the NTSC world, at least not easily or cheaply with the sort of monitors you or I would want to buy. Otherwise why would NTSC stand for "Never The Same Color" !! :-) Because there are no standards restraints on computer display systems, computer hardware and software developers have far exceeded the display capabilities of the NTSC television system. He is taking advantage of that capability to communicate his "look" to the colorist.
>"Marc, all I can say to you about this is that it will, in time happen. After all, I am talking about the future. " That'll happen on the day when you can shoot 200 set-ups in a day. Good work takes time, and that's all there is to it. I still say that a telecine room is not a sausage factory, and you're wrong to expect it to be treated as such.
>Marc, I don't think any of us want to treat a TK room as a sausage factory. As far as final grade goes I always try to be there and to spend as much time as possible with the colourist, talking to him before & during the shoot as well. The problem we all seem to have is with video dailies that are often run off by trainees who don't seem to have spent a lot of time looking at pictures, probably spent the time searching for their white stick.
>The traditional film analogies have to be rush prints and grading the answer print. I'm happy with the TK equivalent of an answer print, happier than I ever was with the film process in fact. It's the dailies that give us grief.
>I have shot gray scales and had them ignored. My biggest problem is "how" to expose the gray scale in many of the lighting setups I do. How do you do it in a backlit cafe where all the people at the counter are mostly silhouette with a tiny bit of detail visible in their shadowed faces and the outside world blown out to where only a little bit of detail is visible there in the highlights? How to shoot a color chart when you have the 800mm lens up shooting a tractor trailer truck rounding a corner in the desert at sunrise a quarter mile away? In the old days, I always got printer lights for all of these tricky scenes that really represented where the exposure was on the negative for every foot of film.
>Yuri Neyman devised a "system" described in last March's issue of American Cinematographer in an article I wrote. The system is based on a gray scale/color chart with IRE numbers on the gray chips and numbers under the color fields. There are also film strips that are made at a lab to exact densities in every available negative stock, with a variation of the same gray scale/color chart.
>The idea is that you have the tele-cine colorist roll up the film strip chart and set the machine to the IRE numbers on the wave-form monitor and the red reference on the vector scope. This should "zero-out" the tele-cine to your stocks particular bias. The colorist then rolls to your chart which you shoot at the head of each roll, or scene, or day, depending on your paranoia, in "white" light correctly exposed. Ideally the charts should match, but whatever the colorist does to make them match gives you accurate and useful info about exposure and color shifts you get from your lenses, the lab, or the film stock.
>If you want, you could forbid the colorist from straying too far from your set-up, making sure you control the look without it getting "corrected" out. Or you can use the numbers on the chart to discuss what the colorist is doing with more precision than before. It's a pretty simple way to deal with video dailies, and you could devise your own system based on the same principles. Glad to join this forum.
>>The problem we all seem to have is with video dailies that are often run off by >trainees who don't seem to have spent a lot of time looking at pictures, probably >spent the time searching for their white stick.
>I really do understand your feeling that way, but let me explain the situation here in Hollywood regarding dailies colorists. Now keep in mind that this relates primarily to those transferring television material, primarily American network television programs.
>The "normal" route to these positions is through videotape operations/telecine assist. Because these are positions that already require full time involvement, there is not much of a chance to apply real training for the colorist position. Therefore, those who truly want to move up must donate quite a bit of their own time to learn the skills required. Admirable, yes, but unfortunately most of the "available time" on the rather expensive equipment that populates telecine rooms is only during those hours, particularly weekends, when master colorists are not available to provide the supervision these employees seek.
>The pricing structure in telecine in this town dictates that the rooms themselves must be heavily utilized in order to remain financially viable. Now you might say that this should spur some of us on to create a training system, possibly modeled after the tried and true union apprenticeship system, that would create a better trained group of potential dailies colorists who would not have to undergo so much "on the job" training. And if we had a union controlled system in the video post houses, this might very well have happened. But it didn't, and although I'm not crying that we're between "a rock and a hard place," that pretty much is the issue.
>It's all a very significant catch-22: you can't have well trained daily colorists if there's no mechanism available to assure their training, and you can't have that mechanism without available equipment and personnel. Part of my job is to find some sort of answer to these dilemmas, but so far I'm having a very hard time. In editorial, and to some degree in camera, it has always been less financially stressful to provide the training this industry requires. Not to mention the fact that in camera, for instance, there is some degree of crossover between the positions you normally move through: camera assistant, operator, and lighting cameraman. There is little of this crossover between videotape operations and telecine. If anyone else has had success with training of daily colorists, I'd really like to know some of the steps that were taken to achieve that success.
>In the meantime, I'm trying my best to pay attention to the work we turn out and make helpful suggestions and provide some degree of instruction when I can. But it **is** difficult, and the "graveyard" shift hours required for daily work doesn't make it any easier.
>This is all well and good but I'd just like to get true ONE LIGHT video dailies. I'm finding with both of the labs I use that, unless I include ridiculously repetitious notes with my dailies, the "colorist" will almost always attempt to "correct" the look on-the-fly in mid-shot. This, despite the fact that I'll often put a nice long grey card at the head of each new shooting scenario. I realize the colorist must assume that we'll be going back to re-transfer later but I'm also convinced that this practice subliminally stacks the deck against the take in which it occurs. Does anyone else encounter this and has anyone found a way to avoid it?
>I think you may already have done this, but... I've been playing around with grabbing frames from my ARRI color video tap onto a laptop with a "Snappy," and printing it out on a Fujix color portable mini printer. If the colors or contrast are off, we play around with it in Adobe Photoshop. Admittedly this takes a little time-but the advantage is it can be done after wrap, and hopefully I finish before the film gets sent to the lab. If we're lucky, the video assist operator helps out, and does the printouts during the day. When we're really lucky, we're working with an on-set AVID, and printing the frame grabs is really easy. The results are not bad, and certainly better for the colorist than the black & white Polaroid 667's I have been sending in.
>On the feature I did this past summer I used the new Gamma Density Chart http://www.loop.com/~gamma1/ and I was pleasantly surprised with the quality of the ¾ inch video rushes. I had told the telecine place to adjust only on the chart. The AC also labeled very boldly the camera reports telling them to adjust on chart.
>I meticulously shot charts on almost every roll so they would have no excuse not finding one. I would say that my rushes were like I had planed the shots about 90% of the time. I don't think it's only the chart but the fact that I made it very clear to only adjust on the chart.
>I'm not trying to defend the labs or transfer houses, but this issue ("one light dailies") is not really as cut and dried as it seems. Unlike film, video is a very high contrast medium in which minor exposure differences are highly magnified. Consider, for a moment, the fact that on a projected print, slightly mismatched blacks are easily acceptable to the viewer's eye, whereas on video, any minor differences in black balance are highly visible.
>Also consider that, particularly in the case of commercial dailies, the look of the dailies is considered to be of paramount importance, and commercial daily colorists are, by and large, encouraged to make the dailies as close to the final color correction as possible. Now, I'm not claiming that all commercial producers demand "perfect" dailies. But I think enough of them do that many of you become concerned when they don't come back that way. I know that some of you are able to attend some of your daily transfer sessions, but I really do believe that if, as a group, you were able to provide more direct supervision, you would find that even given gray cards, it is next to impossible to maintain a quality look without applying some degree of color correction on new setups, even given film negative from which one light FILM dailies would be entirely acceptable.
>>I know that some of you are able to attend some of your daily transfer sessions, but I >really do believe that if, as a group, you were able to provide more direct >supervision, you would find that even given gray cards, it is next to impossible to >maintain a quality look without applying some degree of color correction on new >setups, even given film negative from which one light FILM dailies would be entirely >acceptable.
>Okay, then tell me this: what, in your opinion, will get us the closest to the look we want? Gray card? Chip chart? The gamma/density chart? Macbeth chart? And how should we go about telling the colorist what we want them to do and how far we should go? What language should we use? Should we write notes on the back of the slate and photograph them? Send telegrams with boxes of candy? What do you suggest? And if our colorist is about to get us fired because he/she won't cooperate and the director/producers don't like what "we" are giving them... what then? Inquiring minds want to know.
>On film, minor mis-matches from shot to shot or day to day are tolerated because in the end we print from the original material....the negative. The big problem with video is that, more and more, we make the final corrections from what is quite often a poorly timed set of dailies....which have now become our original....given the reluctance of Producers and in many cases (due to tight schedules) the inability, to go back to the original negative. The problem...which has not been addressed here..is that transfer houses are assigning less skilled people to do Dailies and highly skilled people to do the final tape to tape. In fact it would be much better the other way around as the Dailies guys are, in effect, redefining our Original material.
>Rob Draper ACS
>>The results are not bad, and certainly better for the colorist than the black & white >Polaroid 667's I have been sending in.
>I wonder how many of you also do that. When I shoot B&W 667's on those heavily backlit cafe scenes (Bill Bennet ) and send them to the colorist, the dailies are usually what I intended. Obviously not great for color rendition, but a few additional notes help. This video dailies for eventual film release thing is KILLING me! There has got to be a way to control it!!
>>Okay, then tell me this: what, in your opinion, will get us the closest to the look we >want? Gray card? Chip chart? The gamma/density chart? Macbeth chart? And how >should we go about telling the colorist what we want them to do and how far we >should go? What language should we use? Should we write notes on the back of >the slate and photograph them? Send telegrams with boxes of candy? What do you >suggest?
>All I'm saying is that there is no blanket answer. I think that ALL cards are a help. Simple gray cards are good, although the one that Kodak's about to distribute, with 18% gray as well as reference black and white, is better. I would avoid charts with color chips, simply because most colorists, if they use a reference chart, will likely set up to the gray chips only. So I would suggest avoiding the Macbeth or any other chart with color. The gamma/density chart is fine, although I personally don't feel it provides anything revolutionary. I have heard from some colorists that they like its ability to help set a proper gamma level.
>As far as language, much of the recent discussion here has only confirmed the rather sad fact that daily colorists, in general, are not coming from a film background, but a video tape operations perspective. Because of that, they will more often than not fail to understand much of the technical language a cameraman is likely to use, such as filter types, lens lengths, stock ratings, or the concept of exposure stops, all of which you would expect a lab timer to fully comprehend. This is extremely unfortunate, and I've tried various methods of education, but most new colorists see the world in a totally electronic way.
>Because of this, I would suggest that in talking to a daily colorist, one should attempt to keep the language less technical and more artistic. Whenever you say "I underexposed this scene by half a stop," many daily colorists will interpret this to mean "I'll need to open this up." If, instead of referring to the exposure, you say something like "this scene will come up a bit denser --- that's intentional, leave it that way," this will likely be more easily understood and properly interpreted. Instead of saying "the windows are almost 5 stops over," you might want to say "the windows should be somewhat blown out-don't try to save them." I know all of this sounds rather simplistic, but you're dealing with personnel who have not worked in production, and have likely not been transferring film for more than 1 or 2 years, tops.
>So KEEP IT SIMPLE!!!!!
>Now, a disclaimer: I won't guarantee that following the above advice will get you what you're after all the time, but I do think that everyone needs to realize that we live in a less perfect world than we did on film, a world in which the average experience level is considerably lower than it used to be (lots of reasons for this, I could go on for hours but won't), and a world in which what you do is not well understood by those in video houses who must help to interpret it. Use charts that avoid color, and avoid technical terms in speaking to your daily colorist, and I really do think you'll get more consistent results. Oh, and BTW, photographs are helpful too.
>Mike, The more I think about your "Interpositive" idea the more I like it. I would post the details here and not worry about self promotion. We are all here in an attempt to improve upon the current situation and this seems to be the most viable of all solutions offered to date. To tell you the truth I have not had a serious problem with video dailies for several years...shooting 5-7 TV Movies a year...and I am hyper-critical of what comes out of those dark little rooms.
>I have had four colorists in LA over that period and they have all been relatively new to the game and they have all been excellent. Sometimes a few shakey days in the beginning but then smooth sailing. The secret has not been gray cards or any other on set gyroscopics but communication. Dialogue before the start. Faxed instructions directly to telecine each night and faxed feedback after each days work. I have also set myself up with a very good 21" monitor and ¾" playback deck so I have a constant reference. It would be really nice however to have an uncorrected feed as there are always those scenes that are difficult to pull back....just like in a film lab. It would also be great to be working from what would essentially be "camera original".
>Rob Draper ACS
>I remember attending a SMPTE meeting a while ago where a SONY representative gave a long speech about a video (or computer) interpositive. The exiting thing about this proposal was that the bandwidth of the proposed system was wide enough to capture a lot more of the film than can be captured on normal video (RS 170A). That way you could always "downstep" to normal video later, and would not be locked into a specific setting during transfer. I do not remember the name of the system, nor how far they are with it. This speech was given as part of their big push for "Digital Cinematography". If you guys really want to know, I could delve into my pile of "Interesting stuff that needs to be filed some day" and pull my notes out.
>Rob Draper Comments :
>>"I have also set myself up with a very good 21" monitor and ¾" playback deck so I >have a constant reference."
>Do yourself the favor and don't get a ¾" Umatic playback deck. Get a BetaCamSP playback deck and a component monitor, and you will never go back. I did several years ago and have been spoiling Directors ever since.
>Since you are not looking at a Composite NTSC (or Pal) video signal, but rather a Component video signal, the quality is MUCH better. Also since it is not NTSC the blacks can be at 00% IRE not the 7.5 RIE specified by the NTSC specs, so your blacks are blacker to start with! (From what I understand, PAL has 0% IRE black) BetaCam SP decks used to be quite expensive. A while back I learned that Sony has come out with a new line of less expensive Betacam SP decks called the UVW series and that is what I bought.
>I got a UVW 1200 player and a Sony PVM1354Q 13 inch component monitor. The player was $4,245 and the monitor was $1,085. The monitor is what they call the High Res (HR) Trinitron series and is capable of Component video input. The monitor is capable of 600 lines of horizontal resolution, and will show Component video, RGB video, NTSC, PAL, or SECAM. There is a 19" diagonal HR Component monitor in the series that is called the Model PVM1954Q, if you need a bigger screen. The Component HR monitors are expensive compared to composite only monitors, but you get what you pay for! The player will accept both the small (up to 30 minute) tapes as well as the large 90 minute BetaCam SP tapes. The player will only play NTSC video: in either NTSC Composite (OK, but not quite so good if you only have a composite monitor), or Component (the best way if you have the better monitor that will accept the Component signal) There is a UVW 1400 player/recorder available if you need record capability for about $1,000 additional.
>Be certain that the transfer facility keeps the video signal in the component realm all the way from the transfer machine, through the color corrector and on to the BetaCam SP tape. If they, for instance, run it through a composite switcher, you will loose some of the startling image quality of a pure component signal. Even though the signal in this case is recorded on the tape as component, it was smushed into NTSC composite and then "torn" apart into component and much is lost in the process. Sometimes the loudest whining I hear about doing BetaCam SP dailies is coming from the editors. (They don't want to spring the $ for the playback deck!) What they don't realize is that if they feed a pure component signal into their AVID or whatever digital editing system, the pictures look better, as they are stored and manipulated as digital component signals inside the computer. Better signal in, better signal out.
>Though I think Encore's system of taking an uncorrected feed from the telecine and feeding one tape machine with that as a source for later tape to tape correction, and then taking a density and color corrected feed and recording it on a separate tape for viewing and editing it is a step in the right direction. I would feel a LOT better about it if the telecine and the recording medium had greater bandwidth as Marc mentions in the above quoted statement. If there was more data about what the negative really contained and was available when you do the final tape to tape, the color corrector will have much more capability than s/he does now. Sort of a Super Format, not really intended for broadcast, but intended as a high resolution source for "bump downs" into NTSC, Pal, Secam, etc. Kodak has also proposed this. Going from film to NTSC video is a form of "compression." The less compression that early in the process, the more variability later and the better the final result will be.
>I believe in the Encore VIP system. It makes good sense to have a relatively untouched low contrast master for final color correction and a boldly 'interpreted' dailies tape to show all what you have in mind for final printing. No other system allows you to have both these necessary elements.
>I also believe that if you are kind to your colorist, your colorist will be kind to you. I therefore send him/her an entertaining microcassette with the crew joking around, etc. to relieve his/her late night tedium that also includes my description of the mood of each scene and where I want barely visible detail, etc. This coupled with a pure white light grey card (I prefer the new Kodak card) generally allows the colorist to line up the grey card to see basically what I had in mind, and then tweak as necessary to ensure what I'm going for actually gets to tape.
>>"Shows such as NYPD Blue and Murder One, which do not finish electronically (they >assemble the negative and we transfer directly from that)"
>That fact that these shows mentioned above have an arguably better "look" than shows that are color corrected "tape to tape", fly in the face of your comment that the "tape to tape" finish system, be it your "VIP" system or others, is "just as good" as going back to the original negative for the finish color correction? I think not. "Tape-to-Tape" is a case of the bean counters ruling out the best possible method to have a top quality image.