Home of Professional Cinematography since 1996
Proofing A Film
>If the HD Wars (Episode 1) are over, or at least there is a temporary truce, I have film related item that I found very exciting and I thought I should report on. I'm sure many on CML will report that they have been doing this for years, but for me it was a first and a tremendous improvement over the old way.
>It's at CFI in Hollywood and it's called "proofing" a film. Instead of viewing a first answer print with the timer and giving instructions which he then (hopefully) makes notes on, proofing is basically a slide show. For every shot in the film, you see the first frame, the last frame and a middle frame, printed on the actual stock and same conditions as your print will be.
>It is linked to a computer by a frame counter and displayed on a monitor are the printer lights for the shot and a history of the printer lights. There is no time pressure, you can stop and discuss each shot with the timer and even ask him to back up to previous shots. I found this extraordinarily helpful as I sometimes ask for a correction and then three or four shots later find that although it looks good, it's not a good idea for continuity or overall color/density balance of the entire scene or sequence.
>If you ask for a certain correction, the timer can also tell you what he intends to do and you can discuss with him how it might affect other items in the frame or continuity. I always found the old method very frustrating: you had to make all your choices and scramble to communicate them to the timer so he could scribble them down before the next scene starts while at the same time trying to explain to the director why you are doing it - things always got missed. You could ask for another screening but it was seldom practical.
>They are absolutely upfront and open about the fact that this is NOT a replacement for seeing it in motion (which will happen next week). Clearly some things that don't look OK in a still frame will be just fine in motion and vice versa. Also, on some shots those three frames just don't represent what the important part of the shot is like. Cost wise, it will certainly reduce the number of answer prints necessary.
>With shots that were problematic on the set (losing the sun, changes in weather, etc) I was able to stop, take a moment to recall how I had exposed it and compare directly with the actual printer lights for that exact shot. It was a real reminder of how much metering is an interpretive art as much as a science. Also extremely helpful as (for budget reasons) I was using a stock I don't normally use. I could even see the results of filtration and color balance decisions in a direct, quantitative way. (Yes printer lights have always been available but not in such a direct, accessible manner.)
>The timer was very excited about the ability to examine and discuss the history of printer lights for each shot, for those inevitable times when you see a next go-through and the correction is a little too much or not enough - now you can discuss it in quantitative terms, not vague generalities.
>It also helps that the timer there is extremely knowledgeable, not only about printing and color theory but about production constraints and storytelling needs as well. The sales person told me that it was invented at CFI and they won a technical Academy Award for it but the timer seemed to think it was based on a previous machine. I've forgotten the name he told me, maybe Quantex or something like that? Perhaps someone here on CML can fill in the history.
>I don't know how long they have been doing it, this just happens to be the first time I've taken a movie to a film finish at that particular lab.
>Anyway I found it a huge improvement. I think anybody who tries it will want to push for whatever lab they use to implement some similar system. Obviously it saves money for the lab. Their sales point is that is saves wear and tear on the negative, but I have no doubt that it makes for a better movie. I'll know for sure in a week or two when we finish.
>End of report.
>Sounds like a great system... though quite similar to the Cinexes of old. Hope it catches on.
>Jeff "channeling the ghost of Art Reeves*" Kreines
>*10 free quoted middle names to the first CMLer to identify Mr. Reeves' invention (in this field) by its name
>Actually, long ago, a filmmaker named James Benning had a similar system. (Usually the beginning and end of any take in one of his films is the same as the beginning, or close to it. In fact, one recent film was edited using his own unique non-linear editing system -- a slide projector. He just put a frame from each shot in a slide mount, and could screen a cut very easily.)
>Anyway, he'd just make up a little roll of trims, a couple feet per shot, and have the lab print and print and print it until James liked the printing lites. Then these numbers were used on his A/B rolls.
>Worked very nicely.
>(For you Milwaukeans, this was at Kluge, a little lab long gone.)
>I've timed 3 films at CFI with this method and one at FotoKem with a similar method. At CFI I worked with Chris Regan and at FotoKem Dave Jankin (sp). Both are excellent timers. This is a great system and has been around for quite some time. If you finish in film, it's a great way to reflect on your work not in "reel" time.
Mark Woods, Director of Photography
Stills That Move, Pasadena, CA
>This story reminds me of the old days of substractive printing. The Debrie printers (made by Andre Debrie Co. in France) had a mechanical device that changed the filter band (a 35mm black cardboard strip, with diafragm perforations in 20 different diameters, each 1/6 stop (0,05 log exposure) from the next. To this band coloured filter foils (Yellow, Magenta and Cyan) were attached by a special sheet metal frame.
>As you did your negative cutting, the cutter set aside two frames from each scene (you could also take two frimes from start and two from end) and prepared a 2X2, as we used to call it, a small roll where each scene was represented by 2 frames.
>This was printed with the 2 frame-filter-band-advance device, as many times as necessary, and could be viewed in a modified slide projector, in fact the same leisure paced session with director, cinematographer, and grader, where they would discuss at length what changes to use for the next run. Only after as many as 10 of these 2X2s was a full answer print struck.
>This was also used to determine the best printing filter combination in the days before a video analyzer. The filter band would be prepared with an assorted combination of filters, and a 10 feet scene could get as many as 80 different filter combinations in one pass. This type of filter band was called "The 1000 light Band", even though it was only 120 different filter combinations, which would print in 240 frames (15 feet).
>Those were the days...
>I spent many an evening changing filter foils in grading bands...
José E. Llufrío
Technical Advisor Former Laboratory Supervisor
>Thanks, Mark. I was hoping we would hear from you on this topic as clearly you are an authority on matters of exposure and densitometry. Chris Regan is the fellow I'm dealing with at CFI and I'm very impressed with his knowledge and taste.
>I'm glad to hear that Fotokem has something like this. Do you know if any other labs have it? Is there a copyright or patent issue? I'm sure the basic process can't be patented, but I could see how the software that coordinates the database of the printer lights to the projector is certainly copyrighted.
>Chris did mention that it was based on existing technology that had been around for some time. As it clearly saves both the lab and the production company money (I'm not sure how that divides up) and saves wear on the negative I would hope to see it more widely available. Given the savings for the lab (not just prints but projection time, etc) it seems crazy for them not to offer it.
>>I'm glad to hear that Fotokem has something like this. Do you know if any other labs >have it?
>I would think Delux and the other labs have something like this. Anyone? I haven't shot at Delux for ages, and then it was commercials, no film finish.
>Is there a copyright or patent issue?
>I wouldn't think so, because the little projector is one of the old film strip projectors I remember as a kid. Boy, does that date me, or what?
>>but I could see how the software that coordinates the database of the printer lights to >the projector is certainly copyrighted.
>It's also an interface with the system the lab uses to instruct their printers regarding the color timing.
>If I remember right, the printer both labs used was called a proof printer. BTW, if you ask the timer nicely, he will give you the little rolls of timed film ;-) I have a couple that I intend to cut and mount on slides for examples when I speak to students. Things like matching image size, camera angles, lighting. Whatever seemes appropriate.
>Enough of a ramble. I'm happy you had a good experience with Chris Regan. He's a good guy.
Mark Woods, Director of Photography
Stills That Move, Pasadena, CA
>Only just caught up on this thread.
>The "earlier machine" was a Cinex machine, used in B/W printing. It was similar to the old exposure wedge used when enlarging stills, and allowed the lab to print a short strip of each shot with each frame printed one light darker than the previous frame. Then it was just a matter of picking the best one from each strip.
>Jeff's mention of the Benning system is what I know as the old "two-frame pilot", which was made up by the negative cutter. In the early days of colour printing, when all corrections were made with filters in the printer, the pilot would be printed several times with a range of filter packs, and the best pack for each scene could be picked from the range of prints. It worked as long as the pilot frames were representative, and not part of a camera run-up, or fogged, or whatever.
>The key to CFI's sytem is not just the little filmstrip projector, but the proofing printer itself, which is cued to print (typically) the first, midle, and last frame of every shot at the timer's selected light. It's made by Hollywood Film Company, but what CFI have done is to interface the filmstrip projector with their Colormaster grading machines (the more modern version of Hazeltines for those who are more familiar with that term), so that the grading data created at the Colormaster session is linked to the projector, and subsequently to the printer. The system was described in the AC report on "Titanic" a couple of years back.
>Most labs use variations on this sort of computer assistance. It's not as sexy or "cutting edge" as digital imaging, but (contrary to some views) labs have advanced in their use of technology since electronics was invented. Anything that helps communication between DoP and grader/timer is a good thing.
Group Technology & Services
Manager Atlab Australia
>>Most labs use variations on this sort of computer assistance. It's not as sexy or >"cutting edge" as digital imaging, but (contrary to some views) labs have advanced >in their use of technology since electronics was invented.
>Out of interest Dominic, does Atlab utilize a system along these lines?
Film & Digital Cinematography
>>Out of interest Dominic, does Atlab utilize a system along these lines?
>No, we don't use a proof printer, but that's because we have a slightly different system of grading. Our graders put heart and soul into getting as close as possible with the first print, and we expect that to be a useable print. Of course there are often small corrections to be made, but we get closer with that first full length print, made on the actual printer, than we would with a filmstrip - made on a different printer. (There are usually slight differences in response between printers).
>I put this success down to the skill of our graders, their relationship with the DoP, and also, I'm certain, to the Australian tradition of one-light rushes. It breeds a remarkable consistency in exposure, makes the job of grading just a little easier, and, in part, is what makes the letters ACS stand for quite a lot. - e.g. Accurate Camera Settings.
>We do of course use the Colormaster and INPS computer system (the same as CFI) to manage the grading data.
>By the way, a previous message referred to saving wear & tear on the negative by using a proof printer. That isn't the case at all, the entire negative still has to run through the printer, intermittently rather than continuously. But of course it's true that there is a stock and time saving in reviewing a proof print compared with a full length print, and this adds up to many hours if repeated prints have to be made. And I do like the end-use of the filmstrip as a convenient visual reference to every shot in the film.
Group Technology & Services Manager