The Original Trilogy restored from 35mm prints (a WIP) — Page 7

poita said:

The blue is meant to be there, the sky is quite blue in those shots, which makes the smoke etc. also quite blue.
The original prints do have a lot of blue in the Hoth scenes, not as much as the BD release, but still quite a lot more than the home releases, which appear to have been ‘auto balanced’ when the telecine was done, removing a lot of colour from the snow scenes, probably working on the assumption that snow should always be white.

An interesting update to a topic of much contention over the years here.

poita said:

The blue is meant to be there, the sky is quite blue in those shots, which makes the smoke etc. also quite blue.
The original prints do have a lot of blue in the Hoth scenes, not as much as the BD release, but still quite a lot more than the home releases, which appear to have been ‘auto balanced’ when the telecine was done, removing a lot of colour from the snow scenes, probably working on the assumption that snow should always be white.

Interesting! Could you share the raw frame for this shot? I would like to try another little experiment…

I’m finding in my project (working with the GOUT, SE and BR) that Hoth tends be be bluer than I would guess. The bright, sunlit areas are very white, but the overcast and shadows are more blue and the inside of the base is very blue. This frame (of the gun exploding) tends to match the Despecialised more than the raw 35 mm, but with a bit more blue.

For instance, there is a triangle of blue sky visible through the smoke and the sky to the left should have some touches of blue. Also the background cloud of smoke to the far right should have a very gray, almost bluish tone. The SE (I’m working on a mix of TB and GKAR) may be crap for DNR and details, but they are a fantastic color source.

yotsuya said:

tends to match

should have

should have

I’m never quite clear what you’re discussing with your project. You’re often speaking in these absolute terms of how things should be, but you aren’t quite explaining how you might be using references in an objective way. Is it fair to say that you’re discussing the way you prefer it to look? No judgement for that, since I did a version of SW that was in no way objectively colored. Just curious, since I tend to find myself confused at your posts.

towne32 said:

yotsuya said:

tends to match

should have

should have

I’m never quite clear what you’re discussing with your project. You’re often speaking in these absolute terms of how things should be, but you aren’t quite explaining how you might be using references in an objective way. Is it fair to say that you’re discussing the way you prefer it to look? No judgement for that, since I did a version of SW that was in no way objectively colored. Just curious, since I tend to find myself confused at your posts.

Just the results that I see. What I see over and over on this site are color corrections that all bring it closer to what we think we remember and what we think it should be, but there is a lot of guess work involved. Even DrDre’s color correction tool requires a guess as to what frames to work off of. Most of the guessing isn’t random, but no two people seem to be arriving at the same result so I don’t see there being a single correct answer, more a collective set of answers that keeps getting narrower as everyone gains experience.

In this case, looking at the raw scan, the GOUT, the SE, the BR, the Despecialized Edition, and my color corrections, I was pointing out that the Despecialized Edition of this frame is lacking some blue areas that are in the raw scan and in the GOUT, SE, and BR. To me that is a pretty clear indication that the Depecialised Edition has been over corrected to remove blue tones that should be there. 4 separate scans of the film cannot all be wrong. I posted a screen cap of my color correction of the SE, which came out with very similar colors, but with the blue areas intact.

I suppose there is no right or wrong. There is such a thing as the original color timing, as was present on the original interpositive. That color timing can be found with some variation introduced by the printing process on unfaded 70mm, 35mm, and 16mm prints, although contrast will vary, depending on the generation of the print.

Personally, I trust poita’s judgement in these matters above most others for two reasons:

1. He’s a professional, with much knowledge on the medium and experience working with film
2. He has access to more original prints, than most of us will see in a lifetime. The best estimate of the original color timing will always be an average of a multitude of original prints

yotsuya said:

… Hoth … bright, sunlit areas are very white, but the overcast and shadows are more blue and … a triangle of blue sky visible through the smoke and the sky to the left should have some touches of blue. Also … smoke to the far right should have a very gray, almost bluish tone.

Gentle-curve tweaks of Green (weaker at mid and high) and Blue (stronger at mid but weaker at low) with other points anchored (to not throw off the rest) will get this:

Time to move this thread up the ladder! Here’s a fun little experiment. Let’s correct poita’s red faded TESB print, using a Topps card as a reference.

Raw print scan:

Topps card:

Print matched to Topps card:

Can you throw in the BD frame for comparison?

Here’s a bluray frame from the same shot:

Yikes, the BD looks like crap. Your matched picture looks great as always

Seriously can’t wait to see the restored print of TESB. Once it’s available it will be the only way to watch TESB.

That looks amazing, can’t wait for this release.

DrDre said:

Time to move this thread up the ladder! Here’s a fun little experiment. Let’s correct poita’s red faded TESB print, using a Topps card as a reference.

Raw print scan:

Topps card:

Print matched to Topps card:

You’re just showing off now 😉

DrDre said:

Time to move this thread up the ladder! Here’s a fun little experiment. Let’s correct poita’s red faded TESB print, using a Topps card as a reference.

Raw print scan:

Topps card:

Print matched to Topps card:

I’ve been wondering for a while. Would the topps cards generally be considered good color reference? I’ve got a number of near mint Star Wars and ROTJ cards that all have pretty good looking colors, although I don’t know how accurate they are.