Jump to content


Photo

2-strip technicolor look


  • Please log in to reply
19 replies to this topic

#1 Sean Conaty

Sean Conaty
  • Basic Members
  • PipPip
  • 48 posts
  • Cinematographer
  • Los Angeles

Posted 05 February 2008 - 01:49 AM

i'm curious if anyone has had any success achieving this look digitally. through photoshop or magic bullet, what are the steps? i'm assuming desaturation and losing the green channel, but if anyone's done it and can attach some before and after screen caps, that'd be great.

-sean
  • 0

#2 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 05 February 2008 - 01:55 AM

i'm curious if anyone has had any success achieving this look digitally. through photoshop or magic bullet, what are the steps? i'm assuming desaturation and losing the green channel, but if anyone's done it and can attach some before and after screen caps, that'd be great.

-sean


"The Aviator" did it for the early scenes in the movie.

Actually you throw away the blue record, but you have to adjust the red and green records to compensate, sort of making red more of a brick-orange and green more of a cyan. But that's just in theory, I haven't tried it myself.
  • 0

#3 John Holland

John Holland
  • Basic Members
  • PipPipPipPip
  • 2248 posts
  • Cinematographer
  • London England

Posted 05 February 2008 - 02:22 PM

I dont think it worked in "The Aviator" just looked like something had gone wrong with the print .
  • 0

#4 John Sprung

John Sprung
  • Sustaining Members
  • 4635 posts
  • Other

Posted 05 February 2008 - 02:28 PM

To really do it right, you'd have to find out what the actual two color primaries were. I know they used ferric ferrocyanide for the red-orange and potassium ferrocyanide for the blue-green on the prints. The colors they could make on the screen were only what you can get by mixing those two. The shooting primaries weren't necessarily a good match for the print primaries. They used ortho film followed by a red-orange filter layer and pan film in a bi-pack, shooting thru the base of the ortho stock. Figuring out what the shooting primaries were would be even harder to do. If we could, we could shoot B&W stills of a chip chart thru the taking primaries, and superimpose positives of them thru filters matching the print primaries. That would get us a lot of points to interpolate and build a LUT.

To understand what's going on here, look at the CIE 1931 chromaticity chart. With two primaries, you can only accurately reproduce the colors on a straight line between them on the chart. With three, you can get everithing inside a triangle. If we assume that things move to the nearest point on the line between the two primaries, that would give us a way to compute the transformation. But I don't know if it really works that way.

The two color effect in Aviator didn't look as extreme to me as the actual two color stuff I saw back in the early 1970's. I think they decided to make it look better than real two color actually was.



-- J.S.
  • 0

#5 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 05 February 2008 - 02:53 PM

To really do it right, you'd have to find out what the actual two color primaries were. I know they used ferric ferrocyanide for the red-orange and potassium ferrocyanide for the blue-green on the prints. The colors they could make on the screen were only what you can get by mixing those two. The shooting primaries weren't necessarily a good match for the print primaries. They used ortho film followed by a red-orange filter layer and pan film in a bi-pack, shooting thru the base of the ortho stock. Figuring out what the shooting primaries were would be even harder to do. If we could, we could shoot B&W stills of a chip chart thru the taking primaries, and superimpose positives of them thru filters matching the print primaries. That would get us a lot of points to interpolate and build a LUT.


Actually 2-color Technicolor used a single stock (panchromatic probably) with a prism block exposing the red and green frames above or below the other, essentially pulling 8-perfs down at a time.

http://www.widescree...echnicolor1.htm
  • 0

#6 Leo Anthony Vale

Leo Anthony Vale
  • Basic Members
  • PipPipPipPip
  • 2010 posts
  • Other
  • Pittsburgh PA

Posted 05 February 2008 - 03:03 PM

To really do it right, you'd have to find out what the actual two color primaries were. I know they used ferric ferrocyanide for the red-orange and potassium ferrocyanide for the blue-green on the prints. The colors they could make on the screen were only what you can get by mixing those two. The shooting primaries weren't necessarily a good match for the print primaries. They used ortho film followed by a red-orange filter layer and pan film in a bi-pack, shooting thru the base of the ortho stock. Figuring out what the shooting primaries were would be even harder to do. If we could, we could shoot B&W stills of a chip chart thru the taking primaries, and superimpose positives of them thru filters matching the print primaries. That would get us a lot of points to interpolate and build a LUT.

The two color effect in Aviator didn't look as extreme to me as the actual two color stuff I saw back in the early 1970's. I think they decided to make it look better than real two color actually was.


You're describing Multicolor/Cinecolor rather than Two color Technicolor.
Two color Technicolor used a single strip and a prism that produced two adjacent seperations.
Adrian Klein, Cornwell-Clyne in later editions, mentions that the camera filters could be changed to adjust the colors in photography. The example given was to produce neutral grey skies instead of green skies on exteriors. A green and magenta dye was used for the IB prints.
& Technicolor could alter the saturation of the dyes. Prints for Fairbanks' 'The Black Pirate' had low saturation prints.

The colors on Cinecolor prints did not produce a neutral black, but a greenish black. So an actual print would have some greens on it. I have some frames from a Cinecolor print of a fashion show from a newsreel where the dresses are green and the fleshtones are like ivory.

An article in a 40s AC Handbook suggests using a yellow filter outdoors for darkening the skies.
That further complicates figuring out the shooting primaries.

'The Aviator' was mostly triing to copy the Cinecolor look instead of the two-color Technicolor.
  • 0

#7 Leo Anthony Vale

Leo Anthony Vale
  • Basic Members
  • PipPipPipPip
  • 2010 posts
  • Other
  • Pittsburgh PA

Posted 05 February 2008 - 03:04 PM

I dont think it worked in "The Aviator" just looked like something had gone wrong with the print .


The three-strip Technicolor look didn't work either.
  • 0

#8 John Sprung

John Sprung
  • Sustaining Members
  • 4635 posts
  • Other

Posted 05 February 2008 - 03:22 PM

Actually 2-color Technicolor used a single stock (panchromatic probably) with a prism block exposing the red and green frames above or below the other, essentially pulling 8-perfs down at a time.

Then what Trimble showed us back in the 1970's must have been one of the other two color systems. It definitely used a bi-pack in the camera, and the prints were coated with emulsion on both sides, one for red-orange, the other for blue-green. It looked like the 1926 Ben Hur on the web site, only the subject matter was some people on an ocean liner at sea.

Edit to add: Yes, Leo even knew the name: Cinecolor. Sometimes my brain doesn't work any better than a two color system. ;-)




-- J.S.
  • 0

#9 Bernhard Zitz

Bernhard Zitz
  • Basic Members
  • PipPipPipPip
  • 342 posts
  • Other
  • Zürich, Switzerland

Posted 05 February 2008 - 05:59 PM

i'm curious if anyone has had any success achieving this look digitally.

People rarely cite "tears of the black tiger" when talking about imitating technicolor.
http://www.imdb.com/title/tt0269217/
http://en.wikipedia....the_Black_Tiger

They exaggerate it in this movie. More funny colors, pastel and candy than the real stuff back in the old days. Some scenes look more like 2-strip some like 3-strip and others completely extraterrestrial.

Some of the look is the production design. They did a digital color grade in SD, I guess on digibeta. It's a pitty that 2k DI wasn't popular or affordable back in the year 2000.

This is one of my favorite movies in the last decade. I'd recommend it also to people who don't care about technicolor.
  • 0

#10 Bernhard Zitz

Bernhard Zitz
  • Basic Members
  • PipPipPipPip
  • 342 posts
  • Other
  • Zürich, Switzerland

Posted 05 February 2008 - 06:07 PM

http://en.wikipedia....the_Black_Tiger


Sorry that's the good link: http://en.wikipedia....the_black_tiger
  • 0

#11 Mike Simpson

Mike Simpson
  • Basic Members
  • PipPipPip
  • 113 posts
  • Cinematographer
  • Austin

Posted 05 February 2008 - 10:46 PM

Sorry that's the good link: http://en.wikipedia....the_black_tiger


Oh god that movie is amazing.
  • 0

#12 Steve Phipps

Steve Phipps
  • Basic Members
  • PipPip
  • 59 posts
  • Other

Posted 06 February 2008 - 06:37 AM

I hesitate to post this since (1) it?s wandering a bit off-topic, and (2) I realize that not all members will be able to access the resource. Apologies for that. But I thought of this as similar to a ?in limited release and screening in a theatre near me? recommendation: If you?re interested and able to see it, possibly you might enjoy this.

I thought this was a really nice article: alternative processes to Technicolor, in the context of post-war France.

Andrew, Dudley. The Postwar Struggle for Color.
Cinema Journal, Vol. 18, No. 2. Spring, 1979. pp. 41-52.

Alternatives processes discussed in the article are: Rouxcolor, four (R-G-B-Y) images composited on black-and-white thirty-five millimeter; Thomsoncolor, a lenticular process whose mechanics, I have to admit, I still don?t actually understand (or believe!); and the two descendants of Agfacolor, the German wartime three-color, one-strip stock.

They?re presented in the context of an analysis of post-war French film: Dudley, if I have understood his thesis, argues that the relatively late adoption of color by the French film industry was a response to the restrictiveness and cost of using Technicolor, and the technical deficiencies of these alternatives (and also attributes it to issues of nationalism, and national cinematic ideology).

I thought the article was a nice piece of film scholarship. I was able to read it (for free) through JSTOR ( http://www.jstor.org/ ), courtesy of my public library (USA). The process for me is: I log into my account on my local library?s website, and choose ?Articles and Databases?. JSTOR shows up as a search choice.

Anyway, for what it?s worth. I did think it was a nice piece of film scholarship.

:)

Edited by Steve Phipps, 06 February 2008 - 06:42 AM.

  • 0

#13 Steve Wallace

Steve Wallace
  • Basic Members
  • PipPipPip
  • 163 posts
  • Other
  • Burbank, CA

Posted 06 February 2008 - 11:13 PM

If you are interested in the techinique used in the Aviator see:

http://www.aviatorvf...r...ew&id=color

http://www.aviatorvfx.com/
  • 0

#14 Sean Conaty

Sean Conaty
  • Basic Members
  • PipPip
  • 48 posts
  • Cinematographer
  • Los Angeles

Posted 07 February 2008 - 02:25 AM

If you are interested in the techinique used in the Aviator see:

http://www.aviatorvf...r...ew&id=color

http://www.aviatorvfx.com/


steve, thank you so much for your post. this was exactly what i was looking for.

if you take a look at the link, it goes over converting the RGB to CMY. how would one actually do this in photoshop or after effects?
  • 0

#15 Nick Deacon

Nick Deacon

    New

  • Basic Members
  • Pip
  • 2 posts
  • Cinematographer

Posted 07 February 2008 - 05:38 AM

There's some free FCP plugins that simulate the 2 and 3 strip Technicolor process here
and a list of films shot in Technicolor here
  • 0

#16 Will Earl

Will Earl
  • Basic Members
  • PipPipPipPip
  • 385 posts
  • Other
  • Wellington, NZ

Posted 07 February 2008 - 06:40 AM

if you take a look at the link, it goes over converting the RGB to CMY. how would one actually do this in photoshop or after effects?


This is much easier in Shake, but here's how to do it in After Effects...

http://www.earlyworm...er-effects.html
  • 0

#17 Leo Anthony Vale

Leo Anthony Vale
  • Basic Members
  • PipPipPipPip
  • 2010 posts
  • Other
  • Pittsburgh PA

Posted 07 February 2008 - 04:55 PM

There's some free FCP plugins that simulate the 2 and 3 strip Technicolor process here
and a list of films shot in Technicolor here


Do those fleshtones really match these?

Posted Image

Posted Image

Posted Image

Posted Image
  • 0

#18 Will Earl

Will Earl
  • Basic Members
  • PipPipPipPip
  • 385 posts
  • Other
  • Wellington, NZ

Posted 07 February 2008 - 08:01 PM

Do those fleshtones really match these?

Posted Image


Skull tones!? :blink:

I think the problem with all these simulated looks (including mine) is that it's all based off the process, and that no one has actually compared that to the real process. I'm sure you'd get much better results if you went out and shot side-by-side 2-strip against a modern stock and then went about trying to match the look. I'm sure you'd learn about more about the process than just reading about it at the Widescreen Museum.

To be fair though the woman in the FCP example is rather pale to begin with.
  • 0

#19 John Sprung

John Sprung
  • Sustaining Members
  • 4635 posts
  • Other

Posted 07 February 2008 - 08:08 PM

Do those fleshtones really match these?

No. To do that, you'd have to plot your target fleshtones in CIE 1931 (x,y) space, draw a straight line that sorta goes thru them, and hopefully also hits or passes near the color temperature locus, then pick primaries out near the ends of that line. Because a two color system can only do a line, not an area, the choice of the primaries is absolutely critical. To get the exact same fleshtones, you need the exact same primaries, or primaries on the line thru the ones you're trying to match. With a three color system, most any reasonable choice of primaries will completely cover the subtle stuff that we really care about the most.

I've found a wonderful web site that actually has the numerical data for plotting the spectral locus in (x,y) space, the color matching functions, etc.:

http://cvrl.ioo.ucl.ac.uk/



-- J.S.
  • 0

#20 Nick Deacon

Nick Deacon

    New

  • Basic Members
  • Pip
  • 2 posts
  • Cinematographer

Posted 07 February 2008 - 08:14 PM

Do those fleshtones really match these?

nope - nothing like them : )
the AfterEffects link in Wills post above looks about right though
  • 0


Media Blackout - Custom Cables and AKS

Broadcast Solutions Inc

Willys Widgets

Visual Products

Gamma Ray Digital Inc

Metropolis Post

Opal

rebotnix Technologies

The Slider

Ritter Battery

Technodolly

Glidecam

FJS International, LLC

Tai Audio

CineTape

Aerial Filmworks

Wooden Camera

CineLab

Abel Cine

Rig Wheels Passport

Paralinx LLC

Rig Wheels Passport

CineTape

Tai Audio

Ritter Battery

Media Blackout - Custom Cables and AKS

Wooden Camera

Glidecam

rebotnix Technologies

Metropolis Post

Opal

The Slider

Aerial Filmworks

Willys Widgets

Gamma Ray Digital Inc

Abel Cine

FJS International, LLC

Paralinx LLC

Visual Products

Technodolly

CineLab

Broadcast Solutions Inc