Everything you always wanted to know about printing but were afraid to ask

[Simon Wyss]

How do you do ?

 

There is that ugly duckling of motion-picture film printing or copying, something always a bit disregarded by cinematographers, perceived as a technical necessity but only rarely taken into consideration for creative reasons. Just look at Stan Brakhage who did Mothlight in 1963 in Denver on an old step printer.

 

I'd like to share some knowledge and secrets of mine with those who are not afraid of such things as travelling mattes or flashing or high speed bulk printing and so on.

 

Care for a post ?

 

 

http://en.wikipedia.org/wiki/Stan_Brakhage

[Guest Glen Alexander]

How do you do ?

 

There is that ugly duckling of motion-picture film printing or copying, something always a bit disregarded by cinematographers, perceived as a technical necessity but only rarely taken into consideration for creative reasons. Just look at Stan Brakhage who did Mothlight in 1963 in Denver on an old step printer.

 

I'd like to share some knowledge and secrets of mine with those who are not afraid of such things as travelling mattes or flashing or high speed bulk printing and so on.

 

Care for a post ?

 

 

http://en.wikipedia.org/wiki/Stan_Brakhage

 

I'd be happy to find a lab to make some dupe negatives from SLR stills with a step printer.

 

If you're going to be in Los Angeles December/January timeframe, I'll invite to a showing of my work. I guarantee you will be blown away.

Edited November 9, 2008 by Glen Alexander

[Chris Millar]

yeh, ok I'll bite - tell me everything B)

 

I'd love an optical printer but in the meantime have bought this: http://www.cinematography.com/forum2004/in...showtopic=34544

 

Aside from it intended use are there any modifications you know of that could be interesting ?

[armin dierolf]

How do you do ?

 

There is that ugly duckling of motion-picture film printing or copying, something always a bit disregarded by cinematographers, perceived as a technical necessity but only rarely taken into consideration for creative reasons. Just look at Stan Brakhage who did Mothlight in 1963 in Denver on an old step printer.

 

I'd like to share some knowledge and secrets of mine with those who are not afraid of such things as travelling mattes or flashing or high speed bulk printing and so on.

 

Care for a post ?

 

 

http://en.wikipedia.org/wiki/Stan_Brakhage

 

Hej Simon,

 

i´m actually working on my graduating movie at dffb (deutsche film- und fernsehakademie berlin). It´s a feature film to be shot on 16 and 35 mm. As the director and me created in our short movies a form of telling the story using analog sfx work like double-exposure and matte-photography we now want to put this on a higher and more sophisticated level. Maybe you can help us in thinking and playing? We´re trying to keep our material as pure as possible, as far as digital stuff is concerned.

 

gruesse aus berlin

 

armin

[Simon Wyss]

I'd be happy to find a lab to make some dupe negatives from SLR stills with a step printer.

With 18 X 24 mm (half frame) stills no problem, with 24 X 36 mm you get a thin double exposed line in the middle of your stills. What raw stock did you have in mind for the duplicates ?

[Simon Wyss]

Q. Is there any new technology with printers at all ?

 

A. Well, yes. We have replaced the incandescent bulb by an r-g-b gun that reproduces light amounts exactly from day to day and over years. When you order 25-25-25 you get 25-25-25. This printer light control actually replaces a sensitometer. The accuracy lies within fractures of the thousandth of unit. You have 50 steps. With a 0 no light is emitted. And then you have complete freedom with programming. You may want 34-28-11 for frame 1, 49-2-17 for frame 2, whatever. You can introduce a flash effect by entering 0-0-0 for frame 15,367. Frame 15,367 will receive no light and thus be blank after processing.

 

There are no more limitations to editors. Cuts can be literally only one frame long. A programme holds up to 32,000 values for each printing light colour. We deal with Excel files.

 

Q. What is the working speed of the printer ?

 

A. Step printers can be run at changing speeds between 1 and 25 frames per second.

 

Q. Can you repeat sequences ?

 

A. Oh, yes. Press a button on the control unit, zero the counter once the printer is laced up, start the machine.

[Guest Glen Alexander]

With 18 X 24 mm (half frame) stills no problem, with 24 X 36 mm you get a thin double exposed line in the middle of your stills. What raw stock did you have in mind for the duplicates ?

 

Half frames wouldn't work, I did some macro photography to make dup negs. I used Rollei Techpan.

 

I'm nearly done editing now, just some titles and credits.

[Chris Keth]

We have replaced the incandescent bulb by an r-g-b gun that reproduces light amounts exactly from day to day and over years. When you order 25-25-25 you get 25-25-25. This printer light control actually replaces a sensitometer.

 

You're not meaning that you don't do any sensitometric control work, right? Not every batch of print stock (or negative stock, reversal stock, light sensitive printing paper, etc) is exactly the same. I have to think that you must still do density and linearization checks every so often to keep all of your systems working together properly.

[Guest Glen Alexander]

How about wet gate from VV to 5-perf 65mm print?

[Simon Wyss]

You're not meaning that you don't do any sensitometric control work, right? Not every batch of print stock (or negative stock, reversal stock, light sensitive printing paper, etc) is exactly the same. I have to think that you must still do density and linearization checks every so often to keep all of your systems working together properly.

Of course we countercheck raw stock. The trimming goes optically (condensor). Let me add that also r-g-b are of very constant wavelengths. Deviations have been a dispute subject with other people of the branch here in Europe, and when I wanted to explain that I regard them (deviations) as variables of the processing, not of light or manufacture origin, imagine what storm broke out. Our WRR printer light control system seems to be too precise for some colleagues.

[Simon Wyss]

How about wet gate from VV to 5-perf 65mm print?

How many miles are we talking of ?

[Guest Glen Alexander]

How many miles are we talking of ?

 

about .8 mile, ~4000ft, oh yeah and the B roll will be a DTS timecode track.

Edited November 15, 2008 by Glen Alexander

[Simon Wyss]

about .8 mile, ~4000ft, oh yeah and the B roll will be a DTS timecode track.

Now, let's see - $ 200,000 to build the machine everything included divided by 4000 equals $ 50 the foot.

 

Serious, you can't find anybody who did that job, can you ?

[Guest Glen Alexander]

Now, let's see - $ 200,000 to build the machine everything included divided by 4000 equals $ 50 the foot.

 

Serious, you can't find anybody who did that job, can you ?

 

 

ha,ha, yeah right, VV runs horizontally, 5-perf 65mm, runs vertically.

 

optical printing is about $2/ft, your wet gate has to be less than that.

Edited November 15, 2008 by Glen Alexander

[K Borowski]

How many miles are we talking of ?

 

Lol. Can't you silly EU guys get thrown in jail for talking about miles?

 

Just a customary unit protocol, you generally don't talk about miles unless it is approximate with film. The unit stays in thousands or millions of feet, which should be easy for a metric guy to understand :P

 

Since I'm nitpicking already, shame on you for using DECIMAL miles Glen! If you guys do want to use miles, please put them in pretty fractions, eighths or sixteenths. 0.8 miles is actually 4224 feet, so that'd be more like 4250 feet than 4,000. It'd be most-closely 3/4 mi., 3,960 feet.

 

1/8 mi. is 660 feet, a furlong

 

1/16 mi. is 330 feet.

 

Sot 90 fpm, 3:40 of filming 4-perf. 35mm is 1/16 mi. 7:20 min is 1/8 mi.

 

3/4 mi. would be, therefore, 44 min. or 44:27 of shooting at an even 4,000 feet.

 

These are also, relatively, even numbers for doing calculations in your head and more easily enable you to convert from feet to miles or vice versa, not that anyone cares.

 

You've both received warning from the customary-unit police on this board. Don't let it happen again! ;)

[K Borowski]

On a more serious note, Simon, I am still a new-be when it comes to some of the more theoretical relationships in sensitometry.

 

I understand that additive color is better for consistancy's sake, as you aren't using fade- and light-loss-prone dichroic subtractive filters that make maintaining consistency in exposure across filter-packs and print stock, and chemistry changes a nightmare.

 

Assuming we're working with additive here, using RGB timing must be such a relief in a very complicated system.

 

What I find perplexing is how you have to make corrections, basically where reciprocity laws fail, i.e. the time has to be changed and the system doesn't behave as expected.

 

So also, I see correction factors all the time and there are nice even numbers for a shift from one emulsion to another. Are these just approximate?

 

What about with chemistry changes?

 

 

 

OK, say you have a piece of film that is printing at 25, 25, 25. You need to do an optical print to crop in on a portion of the frame to hids a protruding microphone, so that involves making an optical print zoomed out. Then you run out of print stock, and your developer batch is different than the one you were using when you printed the original dailly.

 

How the hell do you sort through all of these variables systematically?

 

What do you do first to try to match color, and, despite the perfection of additive printers that have been developed recently, CAN you really get to the point where colors truly do match to the same extent as they would with a digital print? I.e., let's say they match to within less than 1/4 of one point each.

 

And then, because of my love of numbers, I want to understand the relationship mathematically. Are changes in stock speed able to be expressed logarithmically, or is it better to think of changes in terms of arithmetic terms?

 

Are there software or formulae that one can use to help model the changes and actually make mess of these variables?

 

For instance, say I want to map all of my printers so that I can relate a print I get out of printer A, using a different lightbulb, to printer B.

 

 

 

I guess what I am saying is I would like to be able to, using math, to come into the lab, plot the chemistry, run a test strip on each one of my printers, determine the film and print emulsion batches I am using, plug all of these into a computer, and be able to use standard, "ideal" printing numbers I've used before as a starting point with each new piece of film I bring in.

 

I also want to be able to understand, through pure math, good starting exposures for everything I print.

[Simon Wyss]

Don't you, Karl, think I'm a metric guy. I've been raised on foot and inch and line.

 

Glen, it was crystal clear to me from the beginning that the films run in perpendicular fashion. What stirred me up to some slight extent is your request for 5-perf 65mm print. Projectionists will refuse 65mm prints. VistaVision furthermore has the picture image aspect ratio of 1:1.85 whereas Todd-A. O. 70 mm is standardised to 1:2.2(5).

 

But it is a pleasure to dream of a brand new printer that is going to occupy another room for a single order of 4000'.

[Guest Glen Alexander]

Don't you, Karl, think I'm a metric guy. I've been raised on foot and inch and line.

 

Glen, it was crystal clear to me from the beginning that the films run in perpendicular fashion. What stirred me up to some slight extent is your request for 5-perf 65mm print. Projectionists will refuse 65mm prints. VistaVision furthermore has the picture image aspect ratio of 1:1.85 whereas Todd-A. O. 70 mm is standardised to 1:2.2(5).

 

But it is a pleasure to dream of a brand new printer that is going to occupy another room for a single order of 4000'.

 

 

Simon,

 

I can easily go VV to 4-perf for $2/ft via optical printer, but why go to 4-perf land when a little more cost I can go to 5-perf 70mm?? large format projectors can project 5-perf, only IMAX uses the full frame. Berlinale has a 70mm projector. I'd rather go large format than distortions with anamorphic glass, (yuk). I like that big 5-perf frame.

 

from what i've been reading, 5-perf widescreen can be projected in many theaters.

 

for my print, i'm bypassing the standard fine grain ip and going straight to timed answer print.

 

 

lots of goodies for fans of large format

 

http://www.in70mm.com/widescreen_weekend/index.htm

 

 

Karl,

ha ha, i think in terms of frames since VV is 180 ft/min.

Edited November 16, 2008 by Glen Alexander

[Simon Wyss]

Look, additive colo(u)r printing has to be understood in connection with the films, i. e. Red-Green-Blue are distinct fields within the spectrum. With subtractive printing you employ Cyan-Magenta-Yellow filters for which must be said: Cyan lies between blue and green. Yellow is quite a narrow band between green and red. But Magenta does not exist as a light colour. You won't find it in sunlight. It is half red, half blue - two colours that are way apart from each other. A transparent magenta filter lets simply pass red and blue. Our eye-brain complex understands this as a colour. Balancing is therefore much more intractable and has always been.

 

Manufacturers would have designed their -color materials for red, green and blue since 1950.

 

You can't develop quarter points out of -color stocks. The corridor given by Eastman Kodak which one can keep up embraces about two points like between 24 and 26. So the customer sees his precision sunk in the suppleness of the combination film-chemistry. We more or less beat around the bush while still maintaining varying batches within one printer point.

 

Chemistry changes? The most important factor is time, then comes agitation, next temperature and last composition. The film does more to the process than the baths do. You understand that it is a simple task to control time. Agitation remains constant in a given machine. Temperatures tend to change at slow paces.

 

We now know that the printer light output is pretty constant. There is nothing else we can do than have test strips processed and measured. Densitometer readings are "calculated" back: It's collected data to show the degree of change necessary to parameters. One method to clear out chemistry faults is to replace the developer as a whole, I mean dump one after the other like with one-shot preparations.

 

We must not confuse different things. I find it hard to accept the approach of perfection when we're at dailies and first answer and soon after negligence when a picture goes into series. What the paying public is served can be sort of different from what the director saw.

 

No trick specialist runs out of a batch of duplicating stock during a given production. Such a man is a beginner, although it's been reported that funny things have been sold to producers. He who manages to persuade a cinematographer into his mistake is a genius and I am looking for that man.

 

Hoping everything is totally unclear now, I'm going to bed.

[Guest Glen Alexander]

Color? I filmed in B&W and filtered out everything below 600nm.....

[Simon Wyss]

Color? I filmed in B&W and filtered out everything below 600nm.....

That relieves us of trouble. <_< No, honest, I don't hope you had blank negatives.

[Guest Glen Alexander]

That relieves us of trouble. <_< No, honest, I don't hope you had blank negatives.

 

Quiet opposite, spectacular. I didn't use standard Kodak or Fuji stock but a special stock.

[Guest Glen Alexander]

With wet gate, the lab can do good answer prints to time out or in many things but with optical printing how much leeway is there?

[John Sprung]

We have replaced the incandescent bulb by an r-g-b gun that reproduces light amounts exactly from day to day and over years.

What is the technology that produces this RGB light? Is it lasers or LED's? In the old days when I knew something about printers, it was an incandescent source split out by dichroics into RGB, controlled, then re-combined.

 

 

 

 

-- J.S.

[Simon Wyss]

What is the technology that produces this RGB light? Is it lasers or LED's? In the old days when I knew something about printers, it was an incandescent source split out by dichroics into RGB, controlled, then re-combined.

-- J.S.

We have a high power L. E. D. that emits Red, Green, and Blue nearly from one point. This is helpful inasmuch it can also be placed rather close to the aperture. A long throw is optically more convenient but not feasible with all printer makes.