Jump to content

Pathfinder's look?


Adam Paul

Recommended Posts

I recently caught the 2007 movie Pathfinder on DVD and absolutely loved the cinematography. But as I'm pretty sure most of the look was achieved in DI I thought I would post my question here. Does anybody know what type of process or technique they used to achieve that almost "painting like" look? I know it was not bleach bypass as the DP Daniel Pearl said it wasn't what they used in the AC article of the Sept. 2006 issue. It kind of reminds me of 300 but instead of that golden brown aspect it has more of a bluish and metallic feel. But the image itself is beautiful and looks like a painting. The grain and gradation is just so interesting. Any ideas of how they did it? I'm aware of the lighting side of it and the lighting strategy Daniel Pearl used. I'm more interested in the color grading side. Thanks in advance.

Link to comment
Share on other sites

Thanks for the kind words.

 

Being a period picture I thought it was important to keep the texture of the film consistent throughout, and knowing I would be in the low light levels of the British Columbian forests during October, November and December, I chose to shoot the entire picture on only one stock, 5218. Even though I don't consider 5218 a grainy stock, since I would often push the film a stop, it was as close as I wanted to get to the director's request for grain.

 

In the early stages of pre-production, while interviewing potential crew, I had to the luck to meet first assistant cameraman Simon Jori. He impressed me with his confident manner and made a lasting impression when he explained to me that in the B.C. forests at that time of the year, there would not be enough light to properly expose fast film, even at mid-day. He went on to explain the merits of shooting with an LLD filter, which I'd never even heard of before. The Low Level Daylight filter is used instead of the normal orange 85 filter when shooting tungsten balanced stocks in daylight conditions. Appearing almost colorless to the human eye, and having virtually no filter factor, the filter does cut the U/V inherent in the daylight which does tend to effect shots made simply without an 85. It is not intended to deliver a final color temperature balance, but rather a way to preserve some skin tone, while reducing the excess blue.

 

I immediately hired Simon to be the A camera focus puller, and arranged for Panavision to send me a LLD filter for testing. As is frequently my method, rather than ramp up a full test shoot day, I used my Nikon loaded with 36 exp. 5218 to shoot stills in the B.C. forest locations with the LLD filter. Looking at the positive print made by Technicolor Vancouver, the results were fantastic, in this situation the undercorrected negative was a huge step in the direction of desaturating the colors, and set us up very well to achieve the look we wound up with. Obviously the warm tones are suppressed by the infusion of blue, and the predominately verdant greens of the forest are also softened by the blue tone of the ambient light counteracting the yellow in the vibrant green. I was so pleased with the results, I decided to use it for the entire movie, and not only the "Low Light" situations.

 

You may find that my method of testing with a still camera isn't practical since the closure of outfits like RGB that would process short lengths of motion picture film and make positive prints. It's not exactly easy, but if you are persistent and repeatedly point out that you are only looking for test results, and are not trying to generate final photography, you should be able to talk your lab into doing it for you, particularly if you represent the volume of development that comes with a feature. Tell them it's fine if there are staple holes, scratches, etc. as long as you can interpret the results you are after.

 

OK, so how much is in camera and how much is in the grading? After seeing my work print stills, I quit talking about the need to find a look, and we got on with making the movie. Did we take it even further in the DI? If you look at the work of director Marcus Nispel, colorist Rob Sciarratta, and myself, both collectively and individually, over the many years we have collaborated, wouldn't you expect us to?

 

Daniel Pearl, ASC

Edited by Daniel Pearl ASC
Link to comment
Share on other sites

Hey Daniel, it's great to have you posting on this forum, hopefully you can keep stopping by from time to time. I'm a big fan of your work on "Pathfinder" as well, I love the textures you got and the mood you achieved with the lighting. I know you guys shot two camera constantly, can you talk about your strategy for using the second camera and what kind of lighting adjustments you made for it? Thanks for the info so far.

Link to comment
Share on other sites

  • Premium Member

When I tested the LLD vs. using no filter at all in daylight on tungsten stock, I basically noticed that the main difference was that the blue layer wasn't printing so high, it was closer to normal. Without the 85B, blue prints near the top of the scale because it is so dense, but with the LLD, it comes down about seven points as I recall.

 

Copied from my "Jennifer's Body" posts:

 

The printer lights were interesting to compare (and at this lab, generally a normal negative should print in the 25's):

 

5212 100T (rated at 80 ASA)

3200K lighting R=32 G=37 B=32

5500K lighting with an LLD correction R=27 G=38 B=42

5500K lighting with no filter R=29 G=41 B=49

 

5217 200T (rated at 160 ASA)

3200K lighting R=28 G=34 B=29

5500K lighting with an LLD correction R=25 G=34 B=37

5500K lighting with no filter R=26 G=37 B=45

 

5219 500T (rated at 320 ASA)

3200K lighting R=27 G=35 B=24

5500K lighting with an LLD correction R=23 G=37 B=34

5500K lighting with no filter R=25 G=40 B=41

 

The slight drop in red density is basically just due to the density of the filter since I didn't do any exposure compensation when I put on the LLD, when in fact, no colored filter has zero density. But it basically has less than one-third of a stop density so you can ignore it, especially since one point in using the LLD is to reduce the excess blue exposure.

Link to comment
Share on other sites

As is frequently my method, rather than ramp up a full test shoot day, I used my Nikon loaded with 36 exp. 5218 to shoot stills in the B.C. forest locations with the LLD filter. Looking at the positive print made by Technicolor Vancouver, the results were fantastic, in this situation the undercorrected negative was a huge step in the direction of desaturating the colors, and set us up very well to achieve the look we wound up with. Obviously the warm tones are suppressed by the infusion of blue, and the predominately verdant greens of the forest are also softened by the blue tone of the ambient light counteracting the yellow in the vibrant green. I was so pleased with the results, I decided to use it for the entire movie, and not only the "Low Light" situations.

 

You may find that my method of testing with a still camera isn't practical since the closure of outfits like RGB that would process short lengths of motion picture film and make positive prints. It's not exactly easy, but if you are persistent and repeatedly point out that you are only looking for test results, and are not trying to generate final photography, you should be able to talk your lab into doing it for you, particularly if you represent the volume of development that comes with a feature. Tell them it's fine if there are staple holes, scratches, etc. as long as you can interpret the results you are after.

 

Daniel Pearl, ASC

 

Hello Mr. Pearl. Really honored to see you here in our midst. Thanks for posting!

 

I've been running behind for several weeks now of my original intended startup date, but I am trying to put together a replacement to RGB's service using Wing Lynch equipment normally used with E6 with ECN-2 and ECP-2 chemistry.

 

It is exciting to know that professional cinematographers, not just students use this service still, so I would, if you don't mind, like to ask a few questions to you about the process that would be very helpful in determining what sort of services professionals such as yourself would need that currently aren't being offered.

 

First off, do you feel that in evaluating tests, they need to be run through the same processor at the same lab, or do you feel that, assuming the chemistry is in control, a dedicated short-length processing laboratory would provide valuable results? My biggest concern is, not having the volume to justify using a roller transport processor like you would find in movie labs, that there might be variations between an RT and something like a wing lynch, that uses nikkor reels. They wouldn't be better or worse (though no staples or scratches), just different.

 

I had the idea to have a same-day turnaround on ECN-2 we receive by scanning the negatives with a high-res Nikon and emailing them back the day they are received, so the only delay in seeing the results would be the shipping to the lab. Do you feel that digitized stills would give you a good idea of what you are seeing, or is a contact print a must for this sort of thing? Keep in mind, we can do both, but because of my unfortunate location, it'd be impossible to have physical results back same-day without paying a small fortune in shipping.

 

Does having the ability to push, pull, cross process in E-6 or C-41, full or partial bleach bypass, and flash film, in your opinion, seem helpful, or do you tend to test all stocks at normal rating and process? I've noticed that that is something that RGB didn't even offer.

 

One thing you should be aware of, Mr. Pearl, is that there is a lab in Florida, Dale Labs, that, as of the last time I called them, still does do ECN-2 processing once a week, and can make "prints" onto ECP. They do very good work, from what I hear. So perhaps something you should look into as well.

 

Finally, the biggest obstacle I am facing thus far is the lack of an actual MP contact printer to work with. I have rigged a slide copier to the task, but obviously it won't produce as crisp of results as a contact print, somewhere in the middle between a show print and a S35 optical blowup release print. However, I do have the ability to video analyze each frame instead of one-lighting it, as was the practice of most of the "Seatle Film Works" laboratories that were out there. This can't really be done with a contact printer. Do you feel that this is a further setback and that a contact printer is a real necessity for this sort of work? Maybe I just need to track down an old cinex strip printer. . .

 

Once again, I appreciate the feedback.

 

When I am up and running, perhaps you would consider doing a side-by-side comparison between the lab you are currently using and myself, shooting a pair of test rolls instead of just one. Your evaluation would be invaluable in determining the worth of this project and whether or not it is useful as an evaluative tool.

 

 

Regards,

~KB

Edited by Karl Borowski
Link to comment
Share on other sites

Hey Daniel. I was wondering how you pushed the look further with your lighting in the forests? I'm working with your gaffer Owen Taylor right now on a film but I haven't had a chance to discuss anything... figured I'd ask you instead. :rolleyes:

 

As I'm sure you know, light is about contrast, direction, and quality. Watch the film with that in mind and you will answer the question for yourself. Owen can fill you in on some broad strokes. Please tell him I said hello. dp

Edited by Daniel Pearl ASC
Link to comment
Share on other sites

"I am trying to put together a replacement to RGB's service using Wing Lynch equipment normally used with E6 with ECN-2 and ECP-2 chemistry."

 

 

I'm glad to see that you are trying to do the honorable thing and help to keep alive an endangered process. Where are you located? I could see how your services could be useful for some tests, but perhaps a bit too renegade for some of the more controlled work.

 

Consistency of development is fairly essential to the process of testing. I guess you'd need to define the range of variation to know if it would be a problem.

Mainly I test low light level items like night vision, Predator blood, etc. and pushing the film is almost always a part of any testing I do.

Digital output probably is more convenient, but once we're in that world, I think we are becoming more and more willing to simply shoot our stills with a pro model digital still camera.

Convenience is a huge part of having my production film development house handle the still work, we will have, or need to set up a system of communication, and there is or will be an established flow of materials.

Printing sounds like a nightmare for you. Is your slide copier as manual as it sounds?

 

I'd be happy to provide you with two test rolls when you are up and running. Good luck, Daniel Pearl, ASC

Link to comment
Share on other sites

I know you guys shot two camera constantly, can you talk about your strategy for using the second camera and what kind of lighting adjustments you made for it? Thanks for the info so far.

 

Pondering your question has led me to realize a pattern to my behavior that I was previously unaware of. The simple answer of course, is to simply say to keep the two cameras as close together as possible, hopefully so close that the lighting works for both. I'd guess the single biggest change would be the slowing down on the back lights which might play too frontal or unflatteringly to a second camera position. I tend to get the B camera operator very involved with the light of the movie. I'l explain my philosophy of light and the relationship with the camera, and define what I mean when I tell them to "shoot the good light". My realization is that on my last four pictures my B camera operator has either become my 2nd unit DP when the time came, or at the very least been responsible for lighting the picks ups and inserts. I hadn't really thought about it before, but I guess B camera is a great place to learn about light.

 

Daniel Pearl, ASC

Edited by Daniel Pearl ASC
Link to comment
Share on other sites

jesus!

 

You've got to love this place. You are curious about how some look or image was achieved and the CREATOR of those images can tell you...

 

Glad to be of service. My friend and associate Wendell Greene let me know there was a question that needed clearing up. DP

Link to comment
Share on other sites

Mr. Pearl, I've sent you a private message, as I think RGB services are off the main track of this thread. . .

 

When I tested the LLD vs. using no filter at all in daylight on tungsten stock, I basically noticed that the main difference was that the blue layer wasn't printing so high, it was closer to normal. Without the 85B, bl ;) :huh:ue prints near the top of the scale because it is so dense, but with the LLD, it comes down about seven points as I recall.

 

Copied from my "Jennifer's Body" posts:

 

The printer lights were interesting to compare (and at this lab, generally a normal negative should print in the 25's):

 

5212 100T (rated at 80 ASA)

3200K lighting R=32 G=37 B=32

5500K lighting with an LLD correction R=27 G=38 B=42

5500K lighting with no filter R=29 G=41 B=49

 

5217 200T (rated at 160 ASA)

3200K lighting R=28 G=34 B=29

5500K lighting with an LLD correction R=25 G=34 B=37

5500K lighting with no filter R=26 G=37 B=45

 

5219 500T (rated at 320 ASA)

3200K lighting R=27 G=35 B=24

5500K lighting with an LLD correction R=23 G=37 B=34

5500K lighting with no filter R=25 G=40 B=41

 

The slight drop in red density is basically just due to the density of the filter since I didn't do any exposure compensation when I put on the LLD, when in fact, no colored filter has zero density. But it basically has less than one-third of a stop density so you can ignore it, especially since one point in using the LLD is to reduce the excess blue exposure.

 

David, it's been a while since I used the additive system, but 3 "points" is a stop with additive, correct? If I recall correctly, subtractive (CMY) color printing has inflated points, with twenty or thirty comprising a stop.

 

So it looks as if the filter cuts out about two stops of blue, a stop of green and 1- to 2/3 stop of red, correct? So, granted you've dealt with the problem of color crossover in the shadows, but wouldn't an extra 2/3 stop of exposure be a definite improvement? It definitely looks to be more than a third of a stop of light loss, unless I have my additive stops mixed up and it isn't 3 points to a stop. I'm sorry if I've misread them. I haven't even seen additive stop references in about four years. Isn't red, like in subtractive with cyan, the most critical color in timing, as they try to leave that one constant to maintain consistent density and to try and avoid as much as possible the breakdown of the reciprocity law with the print stock?

Link to comment
Share on other sites

  • Premium Member

Dominic Case can answer better than me, but I think 6 points equals a stop, certainly not only three -- that would mean each printer light point would create quite a visible shift in color and density.

 

The printer light scale goes from 1 to 50, with 25 being in the middle. Ideally you wouldn't have a layer printing in the high 40's because you quickly hit the ceiling at 50 and have no flexibility to adjust that color further without retrimming the printers. So if your blue prints at 47 just to be "normal" or neutral, what if you wanted to then warm up the image? You'd hit 50 right away.

 

Plus a super dense layer is more than likely going to be a bit noisy in a telecine transfer.

Link to comment
Share on other sites

Dominic Case can answer better than me, but I think 6 points equals a stop, certainly not only three -- that would mean each printer light point would create quite a visible shift in color and density.

 

The printer light scale goes from 1 to 50, with 25 being in the middle. Ideally you wouldn't have a layer printing in the high 40's because you quickly hit the ceiling at 50 and have no flexibility to adjust that color further without retrimming the printers. So if your blue prints at 47 just to be "normal" or neutral, what if you wanted to then warm up the image? You'd hit 50 right away.

 

Plus a super dense layer is more than likely going to be a bit noisy in a telecine transfer.

 

OK, that makes more sense. By those calculations, it's about a third of a stop reduction in red, a half of green and a little over a stop of blue. Thanks for clearing that up.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...