Film-Tech Forum ARCHIVE: Why are Interpositive elements pink?

This is topic Why are Interpositive elements pink? 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=003614

Posted by Michael Barry (Member # 237) on 06-20-2002, 01:13 PM:

I spent some time gaining work experience in DVD quality control, and one of the things I learned is that one of the signs that a transfer has been derived from an IP is that everything looks a bit reddish - the whites are a bit on the pink side and the reds are a bit too hot (although this depends, as some telecines can handle IPs better than others).

Recently, I visited a lab where they were screening an IP, and indeed the film is pink - very pink!

Why is this? Is this done for the same reasons that colour negative film is orange in colour? (Called a mask, if memory serves correctly.)

The second question is: why would someone want to project an IP? What would this tell you that screening a one-light workprint or timed print would not?

Thanks!

Posted by Bill Carter (Member # 156) on 06-20-2002, 03:03 PM:

Except possibly to evaluate quality in a lab setting, I don't know why anyone would project an IP. IP's typically aren't "married" composites, meaning they don't have any soundtrack. The IP does typically have all the opticals printed in.

The odd looking color saturation and gamma are to compensate for the amount that color and contrast will change when you dupe it.

Posted by John Pytlak (Member # 331) on 06-20-2002, 04:57 PM:

Some labs project IPs to inspect them before spending any money to make duplicate negatives. Usually a special projector with velvet gates is used to minimize the risk of any damage.

For review, here is a link to Kodak information about film duplicating systems:
http://www.kodak.com/US/en/motion/support/h1/dealing.shtml

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Research Labs, Building 69, Room 7525A
Rochester, New York, 14650-1922 USA
Tel: +1 585 477 5325 Cell: +1 585 781 4036 Fax: +1 585 722 7243
e-mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion

Posted by Steve Kraus (Member # 476) on 06-20-2002, 10:01 PM:

Orange Masking

As you know the image is formed of magenta (minus green), cyan (minus red) and yellow (minus blue) dyes. No dye is perfect in terms of what colors it passes and what are absorbed. Furthermore, the choice of what compounds to use as dyes is limited to what can be formed by the complex photographic chemistry. Thus, in the real world the dyes forming the image are absorbing some colors of light that they ought not to. This leads to errors as the image is printed. This isn't something you can correct for with the printing light because it varies across the image with the image itself.

The solution to this problem is the colored color coupler. The color coupler is the molecule which will form the color dye during development IF it's been activated by the appropriate color light. (OK, technically it's usually a reaction with the oxidation product of the developer after reaction with the silver compound but let's not go there.)

A colored color coupler is exactly like it sounds: It has a color to it before processing. The color, in this case, is designed to be very close to the UNWANTED light absorption of the dye molecule, if one is formed.

Consider the old adage, "if you can't beat'em, join'em". With colored color couplers this unwanted absorption is the same everywhere. At one point it's because dye has been formed and the dye has the erroneous absorption as well as the desired one. At another point no dye has been formed and the colored color coupler left behind creates the same unwanted absorption.

Now the unwanted absorption is the same everywhere and it can be compensated for with the color of the printing light but also more especially by the sensitivity designed into the material being printed on to. The combination of these unwanted absorptions appears as orange. On the unexposed parts of the film like the edges what you are seeing is entirely the colored color couplers. Within the image you are seeing a combination of them and spectrally deficient dyes. The same techniques is used on IP and IN stocks for the same reason it's used on camera negative.

That's why these film stocks are orange. But not why you are observing different looks on DVD transfers.

Posted by John Pytlak (Member # 331) on 06-21-2002, 05:55 AM:

Gee Steve, good explanation of colored couplers. Did you ever work for Kodak?

Kodak invented colored couplers during the 1940's. The technology greatly improved the color reproduction of the negative-positive color system. Over 50 years later, every manufacturer of color film still uses this Kodak innovation in their color negative films.

A complete discussion of colored couplers and how they work is in the book "Principles of Color Photography" by Ralph M. Evans, Wesley T. Hanson and W. Lyle Brewer (the Kodak inventors), published by John Wylie and Sons, Inc., 1953. LOC Catalog 53-6722.

And some links:
http://www.kodak.com/country/US/en/motion/support/processing/sequence1.shtml
http://www.spinics.net/lists/scan/msg02606.html
http://frontiernet.net/~rlmsmw/photo/c41.htm
http://www.kodak.com/US/en/motion/support/h1/

Colored couplers are only one of the many sophisticated technologies that Kodak has developed to improve the color and sharpness, and reduce the graininess of film. More improvements are on the way!

Posted by Michael Barry (Member # 237) on 06-21-2002, 10:16 AM:

Thanks, Steve!

And thanks for those links as well, John!

I had remembered the basic principle behind orange masking from my darkroom theory, but a) I have never understood it in such excellent detail and b) having never seen a motion picture IP or IN I didn't realise that they also used the same technology.

So once again, thanks for the excellent explanation.

As you say, the fact that IP transfers to video are a little on the pink side is not therefore related to this masking system. Can anyone tell me why IP and low-contrast IN elements produce such different results (again, depending on the telecine chain's ability to handle IP) particularly with respect to the pink bias that results from using IP elements?

Posted by Steve Kraus (Member # 476) on 06-21-2002, 10:32 AM:

Thanks John. High praise coming from you. I was going to joke that I'm after your job but I think you get too much snow out there. Then again, you do have your own railroad and then there are the LAD girls...

Posted by John Pytlak (Member # 331) on 06-21-2002, 10:38 AM:

For optimum quality, telecine transfers today are almost always made from a color-masked pre-print element, rather than a print. If a print must be used, Kodak makes a special low contrast print film for better telecine transfers:
http://www.kodak.com/US/en/motion/products/lab/h12395t.shtml

There should be no contrast mismatch or color "bias" in a telecine transfer, regardless of the film type used. With modern telecines, even films faded by improper storage can usually be transferred with high quality results.

Posted by Michael Barry (Member # 237) on 06-21-2002, 11:04 AM:

Oops...in my last post I referred to low-contrast internegs. That doesn't make any sense! I meant to say low-contrast prints...

OK, I think I've got it...with a low-contrast print, the telecine system doesn't have to deal with a mask at all, which is why the results are so (comparitively) neutral. An IP, on the other hand, is masked so the results depend on the telecine's ability to compensate for said mask. Any kind of film elements that are masked (which would be other than a regular pos print or low-con print) would be similar in this respect.

Thanks again for the ongoing help!

Posted by John Pytlak (Member # 331) on 06-21-2002, 11:51 AM:

Almost any modern telecine handles color masked pre-print elements very well. A transfer from an IP or DN is the optimal way to go, followed by the Kodak Teleprint film. A projection contrast print is usually the least optimal.

Posted by Michael Barry (Member # 237) on 06-21-2002, 03:47 PM:

I suspect I may have been looking at slightly older transfers. To verify, I remembered that the new DVD of 'Seven' was transferred directly from the camera negative (which is very unusual according to the commentary track). They compare the new transfer with the previous one, and indeed the new transfer, done on a 2K spirit datacine with a DaVinci 2K colour corrector, has a very clean, neutral colour balance and a full contrast range, despite coming from a pre-print element.

Thanks for helping me sort out the confusion...

Posted by Stephen Furley (Member # 1277) on 06-21-2002, 04:57 PM:

Question to john,

As I understand it, there is just one type of intermediate stock, but it may be necessary to make an interneg from many various types of positive material, e.g. an interpos, or a 16mm Kodachrome original. These will obviously have vastly different contrast, how can they both be printed onto the same stock, to produce internegs of suitable contrast for standard prints to be made? It may even be necessary to intercut material from different sources, so they would have to be quite closly matched. I wouldn't have thought it would be possible to vary the standard ECN-2 process enough to produce the required contrast without creating other problems.

Posted by John Pytlak (Member # 331) on 06-21-2002, 10:21 PM:

Kodak makes a wide variety of duplicating films specifically for motion-picture labs:
http://www.kodak.com/US/en/motion/products/postProducts.shtml

To make an internegative from a reversal original, you would normally use EASTMAN Color Internegative II Film 5272 / 7272:
http://www.kodak.com/US/en/motion/products/lab/h15272.shtml

The ECN-2 process can be "pushed" or "pulled" slightly by varying developer time, but usually the process is run per Kodak specifications:
http://www.kodak.com/US/en/motion/support/processing/h24m7.shtml
http://www.kodak.com/US/en/motion/support/processing/push.shtml

Posted by Steve Kraus (Member # 476) on 06-23-2002, 08:22 AM:

Is it somewhat surprising that there is still just one intermediate stock for both IP's and dupe negatives? Granted that both situations require a low contrast / negative-going / masked stock but one might have thought there would be enough difference between these two situations to have them branch off into specialized products by now. Apparently not.

Posted by John Pytlak (Member # 331) on 06-23-2002, 10:19 AM:

Whether the image is a positive (IP) or negative (DN), the characteristics required for the film are the same, since the scene information is placed on the "straight line" portion of the duplicating film. My Laboratory Aim Density (LAD) Control method helps labs achieve the proper "curve placement" for optimum duplication:
http://www.kodak.com/US/en/motion/support/h61/