logo Sign In

Star Wars 1977 Technicolor IB print color references (matched to print)

Author
Time
 (Edited)

This thread is dedicated to presenting and discussing color references based on scanned frames of a 1977 Technicolor IB print. The scanning was performed with a Reflecta RPS 10M scanner:

http://www.filmscanner.info/en/ReflectaRPS10M.html

I also bought the Silverfast software package, which allows you to calibrate the colors using a it-8 color target. An uncalibrated scan generally gives a fairly poor reproduction of the colors of a film frame or photo:

http://www.silverfast.com/highlights/autoit8/en.html

Here I will share 1080p 8 bit versions of the calibrated scans for color reference purposes, and to discuss and compare with other OT versions.

A big thanks to all who contributed!

As a special thank you, everyone who donated will also get access to the raw scans, and the high color depth, high resolution calibrated versions of the scans.

Here’s a set of completed scans:

Author
Time
 (Edited)

Excellent ! I assume these calibrated references will be very useful !

Aniway, it’s very difficult to resist the temptation of “color correcting” these accurate frames ! 😉

Imgur

Author
Time
 (Edited)

Ugh. I think my heart just sank into my stomach.

Edit: Dr Dre figured it all out, heart now restored!

Author
Time

I think we need to figure out how to apply these to the film in our restorations. I am encouraged that C-3PO comes out a bit brassy (on the greener side of gold) and a lot of the skin tones look fantastic. But the darker colors seem far too dark. I think we need to couple these with the Senator and other showings to understand how the dark colors appear when projected. I don’t think they come out quite that dark on the screen. Those Tatooine/Tunesia shots look way too yellow (as in unnaturally yellow) but the rest look great. Can’t wait to see more.

Author
Time
 (Edited)

yotsuya said:

I think we need to figure out how to apply these to the film in our restorations. I am encouraged that C-3PO comes out a bit brassy (on the greener side of gold) and a lot of the skin tones look fantastic. But the darker colors seem far too dark. I think we need to couple these with the Senator and other showings to understand how the dark colors appear when projected. I don’t think they come out quite that dark on the screen. Those Tatooine/Tunesia shots look way too yellow (as in unnaturally yellow) but the rest look great. Can’t wait to see more.

It seems to be a generic property (issue?) with the technicolor prints. When we compare the calibrated scans with Mike Verta’s calibrated photographs, the yellow appears fairly consistent:

I will adjust the contrast and brightess settings somewhat to see how the scans come out, to optimize the scans for view on a computer screen.

Author
Time
 (Edited)

There’s something wrong in the software settings, as the prescan and frames themselves look much less green than the final versions.

Author
Time

I second what other people are saying: it’s a tad too dark. But I’ll wait for you to perfect your scanning process before making any more judgments.

What’s the internal temperature of a TaunTaun? Luke warm.

Author
Time
 (Edited)

I figured out what went wrong. For some reason windows photo viewer altered the colors of image when I opened, rotated, and saved the raw scan. Here’s are the correct colors for the Death Star corridor shot.

Author
Time
 (Edited)

I used the standard levels of the Silverfast scanning software. Here’s the first full set of calibrated scans for the Death Star Conference scene:

Author
Time

If the calibration make the scans match the frame under normal light, they necessarily are darker and have a lower contrast than the projected film in a dark room, no ?

Author
Time

Ah, these new scans excite me. I just checked how close the next to last is to my color correction. Very close. I came at it by manually correcting Mike Verta’s samples for the noted green cast (and other issues I noted with it). But the uniform, skin tone, and walls, are all in the ball park. Very exciting that I got so close. But then they both are IB prints so they can’t be too far off. I think it is sample time, but I won’t clutter your topic with that.

Author
Time
 (Edited)

While the print has the green cast typically seen for Technicolor prints of Star Wars, the colors are in good agreement with the Senator print photo of this scene:

Author
Time
 (Edited)

Here’s what I get when white balancing the wall lights in Premiere Pro. With a slight bump in exposure.

Author
Time

The problem with this approach is, that the lights are supposed to be blue. 😉

Author
Time

Okay, I know I’m not supposed to use automatic features when playing with colors…and I know they aren’t perfect…but I love how these frames look after auto-white balancing them in GIMP:



What, a man builds a giant mound of dirt in his house and you aren’t entertained?

Author
Time

DrDre it’s nice to see accurate scan of star wars
Probably is better to open a new topic where users can try to make their own version of the scan
Just for curiosity silverfast seems a very cool and complex software, can i ask what value do you have in gamma gradation?

Here’s my fast try


Author
Time
 (Edited)

That looks great! I think the settings are now such that no information is lost in the dark and light areas. At this point I’m mostly concerned with getting color accuracy, but your gamma correction is looking pretty sweet, so I will be using yours for future reference.

Author
Time
 (Edited)

Thanks DrDre!!


edit:i’ve reupload this one

This comes out a bit too much saturated when i increase the contrast

Author
Time
 (Edited)

age said:

Thanks DrDre!!


This come out a bit too much saturated when i increase the contrast

The colors for the latter frames are not correct, as there was some color conversion issue. I will scan those again with the correct settings. Only the Death Star conference room shots are correct.

Author
Time

The confrerence room looks up better in the calibrated scans, DrDre.

Author
Time
 (Edited)

I decided to go for an alternate gamma correction, as the previous one seems a little too contrasty and saturated, compared to the actual frames, and less representative of the actual photography:

Raw scans:

Gamma corrected scans: