Film-Tech Forum ARCHIVE: How fast are reels printed on high speed printers?

This is topic How fast are reels printed on high speed printers? in forum Film Handlers' Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=1;t=007868

Posted by Mark J. Marshall (Member # 1409) on 03-17-2006, 07:02 PM:

And also, how fast do depot workers rewind reels on their rewind / benches?

Posted by David Graham Rose (Member # 1450) on 03-18-2006, 05:54 AM:

Greetings

Very fast indeed. In some branches of Technicolor, film is printed at a rate approaching the speed of light, which is why the quality drops so dramatically. On an emergency print run of the last Star Wars, the panel printers were pressed into action at 1.5xspeed of light. What actually happened is the the printers actually began to run in reverse, and printed a mirror image of the film, thus causing projectionists here in the UK to have to lace up with the emulsion facing the lens.

Good afternoon

David

Posted by Leo Enticknap (Member # 534) on 03-18-2006, 08:19 AM:

I was once told that the London Technicolor lab has the capacity to produce a million feet of release print per day.

Posted by David Graham Rose (Member # 1450) on 03-18-2006, 08:27 AM:

True, but at what cost to quality? Why are so many SRD tracks failing in the UK on new release prints?

Posted by Randy Stankey (Member # 64) on 03-18-2006, 11:34 AM:

How many machines do they have in that lab?

If they had 100 machines, a million feet of film would only be 10,000 per. Not amazing. But if there were only two or three of them, that would be something to see.

For three machines, that would be 63 miles for each.

Posted by Brad Miller (Member # 2) on 03-18-2006, 01:21 PM:

Technicolor has 4 high speed printers and many other printers in their LA lab. Deluxe has 18 high speed printers (if I recall correctly) in their Toronto lab. I can't remember the specific numbers, but the term "damned fast" should suffice. [Wink]

The depot workers don't wind film at all. That's why they all use clip-together reels.

Posted by Mark J. Marshall (Member # 1409) on 03-18-2006, 02:16 PM:

Faster than one reel per minute do you think? If not, two minutes? Five?

I guess I'm thinking of the distribution center workers not the depot workers. I thought I saw one of them rewinding a reel on one of your photos of the Technicolor warehouse.

Posted by Brad Miller (Member # 2) on 03-18-2006, 02:24 PM:

That's the same place...and they were doing an inspection, NOT rewinding onto reels.

Posted by Mark Gulbrandsen (Member # 72) on 03-18-2006, 03:05 PM:

The last time I heard which was a few years back it was in excess of 1800 fpm but it may have been increased by now.

Mark

Posted by Mark J. Marshall (Member # 1409) on 03-18-2006, 03:52 PM:

Thanks, Mark.

Brad,

Let me rephrase... How fast can that thing that the depot workers are inspecting film on wind film if they crank it up all the way?

Posted by John Pytlak (Member # 331) on 03-18-2006, 05:01 PM:

"Off the shelf" BHP Bi-Directional panel printers normally run at 960 feet per minute for release printing. Customized (proprietary) printers go up from there, and often "loop" the picture and sound negatives.

Remember, as long as the printing negative has short pitch perforations (e.g., BH-1866) and the raw stock has long pitch perforations (e.g., KS-1870), a contact printer can be optimized to have almost no slippage, even at very high printing speeds.

Posted by Dominic Case (Member # 1830) on 03-23-2006, 01:04 AM:

quote: David Graham Rose
True, but at what cost to quality?

No-one here has produced any argument that relates printing speed to quality in any scientific way. As John points out, if the neg and the raw print stock have the correct pitch perforations (they do) and if they are transported around he printing sprocket with sufficiently well-controlled tension, then the film can travel almost as fast as a London bus (OK work that out [Smile]

) without loss of quality. Slippage and steadiness aren't issues if the engineering is done properly, and the technology really is at a very sophisticated level nowadays.

Posted by Carl Martin (Member # 1146) on 03-23-2006, 04:05 AM:

one thing that i understand can't be done in a high-speed print job is color timing changes. so (presumably) the interpos or interneg is timed and the print is made one-light. that is sub-optimal, isn't it? (which is not to say that it would be done optimally even at a slower speed.)

also, what about redeveloper splashing on a non-cyan track print? is that any more prone to happen the higher the speed?

and raw-stock splices: more likely to be poorly aligned? i wouldn't think fogging would be any more frequent, but... is it?

Posted by Stephen Furley (Member # 1277) on 03-23-2006, 05:33 AM:

quote: Carl Martin
one thing that i understand can't be done in a high-speed print job is color timing changes. so (presumably) the interpos or interneg is timed and the print is made one-light. that is sub-optimal, isn't it? (which is not to say that it would be done optimally even at a slower speed.)

If the interneg is correctly timed (graded), either when it was made, or by doing it at a previous film element or digital stage, then there is no need for any light changes when making the release prints. There's nothing wrong with doing it that way, it makes far more sense to make the changes once, or a rew times, rather than hundreds, or thousands of times. Do high-speed continuous contact release printers even have the means to make scene by scene light changes? I've never seen one, but I would doubt it, because it wouldn't be necessaty, and because it would be difficult to do it at the high speed.

Posted by Carl Martin (Member # 1146) on 03-23-2006, 06:13 AM:

it works, in the sense that the colors are correct overall, but my understanding that this comes at a certain expense, but i don't know exactly what to call it. maybe to do with detail, contrast, and grain?

let's say a scene was shot with normal exposure, but intended to be printed dark. if the interpositive is timed, then the range of densities in that shot will be compressed into the densest portion of the density "spectrum". then this is printed into the least dense portion on the internegative, then dense again onto the final print. whereas if the final print is timed then this tonal compression won't happen until the final step and will retain its detail, etc, better.

now, i grant you that i don't really know what i'm talking about and certainly don't have the correct terminology for what i'm trying to say. also, my source for this explanation is a guy who doesn't always have his facts straight, but has spoken with great cinematographers about these things (many of you probably know who i'm talking about). it seems plausible to me.

anyways, if i've got this all wrong, someone correct me. i'd also like to know the right terms for things. actually, my uncle used to be a color timer so i'd like to be able to pick his brain about this stuff.

Posted by John Pytlak (Member # 331) on 03-23-2006, 08:46 AM:

Timing/grading of the cut negative is usually done using an electronic color analyzer, that displays a positive image of the negative, and allows the color timer to see the effect of any changes in printer exposure (printer "points" or tape values). The cut negative is then actually printed to produce an "answer print" for approval of the director and cinematographer, and additional small changes are made to finalize the scene-to-scene color and density.

After the final scene-to-scene timing has been approved by the filmmakers, the timing information is used to make a master positive on a film like Kodak VISION Color Intermediate Film, most often with a wet gate additive continuous contact printer (e.g., a Bell and Howell Model C printer with a Schmitzer gate at about 180 feet per minute). Short-pitch duplicate negatives for release printing are made from the master positive, ideally on a pin-registered step printer.

Before final release printing, "check prints" are usually made for final approval of the color timing and continuity of the release prints. Many times, this is directly compared to the approved timed answer print to verify that everything was done correctly. If necessary, additional small scene-to-scene changes could be made during high speed release printing (done using additive printers), but this is rarely necessary.

Most laboratories use the Kodak Laboratory Aim Density (LAD) system (which I developed), or a variation of it, to control the color timing and duplication process. Here is a link to information about LAD:

http://www.kodak.com/US/en/motion/support/lad.jhtml?id=0.1.4.15.8.6&lc=en

quote:
Laboratory Aim Density (LAD)
In the Laboratory Aim Density (LAD) printing control method, a standard control patch specifies densities midway between the minimum and maximum of those typically obtained for a normal camera exposure. These specific densities have suggested tolerances for each film type in the duplicating and print system, and are listed in this publication, along with helpful hints for their use and evaluation. All color films in a production, regardless of film type or origin, are timed with respect to the LAD control film. Each frame of LAD control film, having the standard patch, with proven accurate exposures plus other information, is available from Eastman Kodak Company.

Here is an interactive tutorial about the "Imaging Chain" for film duplication:

http://www.kodak.com/US/en/motion/support/imageChainDetect.shtml

And tutorials about printing and other laboratory operations:

http://www.kodak.com/US/en/motion/support/h1/dealing.shtml

quote:
Dealing With A Laboratory
Tips on Selecting a Laboratory
Laboratory Services: a Walk-Through
Laboratory Operations
Technical Assistance From Kodak
Motion Picture Film Cleaning
Motion Picture Printing

And a Handbook from the Association of Cinema and Video Laboratories (ACVL), that Dominic Case helped write:

http://www.acvl.org/acvl_manual/index.html

quote:
Published by The Association of Cinema & Video Laboratories, Inc., the ACVL Handbook exists to acquaint users of motion picture laboratory services with terminology, pre-print preparation and procedures which are standard in the industry.

The film laboratory plays an essential role in the transition from conception of story, plot or theme to the finished product -- the motion picture film. In carrying out this function, there must be teamwork and understanding between the producer, the director, the cinematographer and the laboratory so that the finished product will be of maximum quality, interest and effectiveness.

Now in it's Fifth Edition, the ACVL Handbook offers recommended procedures for motion picture and video laboratory services.

With modern duplicating stocks and techniques, almost all of the image qualities contained in the answer print are in the release prints. This is checked on a regular basis when the answer print is directly compared to the check prints.

Posted by Dominic Case (Member # 1830) on 03-28-2006, 06:00 PM:

The benefit of a timed (colour corrected) interpositive and dupe negative is that you are using the best part of the characteristic curve to record the tonal range of the image.

The IP and DN stock has a very long straight line portion - so if both stages are exposed correctly by the lab (using the LAD system that John describes), then the full range of tones from the original negative can be reproduced with no alterations. However, by putting the colour correction in at the first stage, you are clearly positioning the tonal range optimally on the IP and then the DN stock, as well as removing the need to do it later. Scene to scene light changes can be done on any printer, but not at high speed.

Also of course, once the grading is locked in, the scene-to-scene balance, and the "look" of the film, is locked in too. Negatives sent to other labs for international release prints will produce the desired results without the cinematographer having to approve a new grade.

Carl's theory about normally exposed negative intended to be printed down is one of those things that was probably true of earlier generations of emulsion, and may still be true if the IP and DN aren't made to exactly the right standards: but otherwise it's no longer a problem. Still, old wisdoms are hard to kick off.

Applicator splashes aren't a risk in high speed printing but in high-speed processing. Something else that is now (at last) becoming a thing of the past.

Posted by Carl Martin (Member # 1146) on 03-29-2006, 05:45 AM:

dominic,

thanks for addressing my question more directly. and for the term "characteristic curve".

i was thinking of timing mainly as the process of giving a certain look to footage properly exposed, like in godfather (for which timing at the last stage seems more ideal), while you focus on bringing footage that perhaps wasn't perfectly shot back into the middle of the characteristic curve (which should ideally be done earlier rather than later--sounds reasonable). i guess it's unrealistic to hope for two timing stages. too bad more prints can't be made off the o-neg.

i do all too often see prints that just look like mush and have no shadow detail whatsoever. from what you're saying that's not due to high-speed printing per se. so that leaves...

1) "the IP and DN aren't made to exactly the right standards"
2) digital intermediate: another reason to hate those.
3) artistic intent
4) indifference

is it your opinion that (1) is a fairly rare occurence? maybe it varies a lot by lab?

Posted by John Pytlak (Member # 331) on 03-29-2006, 09:39 AM:

"Lost shadow detail" is most often due to underexposing the original negative, and is likely to vary on a scene-to-scene basis. If the "curve placement" onto the master positive or duplicate negative was grossly wrong, the entire reel would have an issue. In my experience, since the development of LAD to help labs in the early 1970's, this has rarely been a problem. As Dominic notes, best practice is to use the scene-to-scene timing determined in making an answer print to make a timed master positive --- Kodak VISION Color Intermediate Film has more than enough latitude to fit all the scene information on the "straight line" portion of the sensitometric characteristic:

http://www.kodak.com/US/en/motion/products/intermediate/5242.jhtml?id=0.1.4.6.4.4&lc=en

quote:
For the filmmaker, these intermediate films help to preserve all the sharpness, clarity, and color of the original images and carry them more faithfully to the final display screen.

For the laboratory, VISION Color Intermediate Films enable the production of printing elements that are closer in quality to the look of a print made directly from the original negative.

Postproduction facilities will appreciate the features of the VISION Color Intermediate Films, especially improved sharpness for digital output. They're the choice for production of masters for theatrical re-release, video and DVD release, satellite, cable or broadcast TV presentation - or use in any media of the future.

Posted by Dominic Case (Member # 1830) on 04-03-2006, 08:44 AM:

quote: Carl Martin
1) "the IP and DN aren't made to exactly the right standards"
2) digital intermediate: another reason to hate those.
3) artistic intent
4) indifference

Well that pretty much covers all possibilities! Though as John points out, underexposed original negative is another cause of muddy blacks in prints - though that would be rare to happen throughout a production.

Most labs that handle studio feature work follow the same recommendations for duplication - and the stock works very well when those standards are followed. Obviously I can't speak for them all though. Similarly with Digital Intermediates: when set up correctly, the results are superb. If you've seen DIs that give you reason to hate them, then they aren't typical of good DIs. It's still newish technology though, and the results aren't as consistently good as they could be: sometimes though, this is as much to do with overambitious artistic intent as with technological imperfections. Again, the DI process brings wonderful benefits to a film when used correctly: it shouldn't degrade the results.