Ghost in the Shell

[CSchmidlapp]

Could you try the correction you have already done, on the scenes from the book pics?
Just to see how close it looks

[BronzeTitan]

Sure! Only I'll use what looks like a scan, as opposed to a camera photo. Just check out the photos' "white" areas - anything but white, and not even uniformly not-white. For comparison, I'll continue to correct the Judgment version, to compare the previous settings to the new, independent settings to the particular shot (for which I must look around).

[BronzeTitan]

I found a few good background scans to try (thumbnails & links below). Fortunately, there was one scene that had some variety of color -- hard to do in the bleak future of GITS. I also came across a blown-out scan of one of these scenes. It served to demonstrate that anything can be messed up and that all this "color correction" is an educated guess.  

1. https://i.pinimg.com/736x/af/fe/80/affe8...703bfd.jpg
   
2. https://i.pinimg.com/736x/19/15/ab/1915a...aa8672.jpg
   
3. https://creators-images.vice.com/content...d51a4b.jpg
   
4. https://i.pinimg.com/736x/5e/f4/7e/5ef47...ecture.jpg
   

[BronzeTitan]

As one may guess from this comparison between the 2 HDTV(J) snapshots and their original artwork backgrounds . .

. . the 1st color-correction settings won't work on the 2nd snapshot. And, of course, it didn't. To be fair, the two backgrounds are obviously dissimilar to begin with. Even individually normalizing each book-reproduced background to compensate for the scan/photography/printing would not smooth out the difference between them. One would have to further homogenize them on one's own authority, which would invalidate the legitimacy of this whole approach ... unless they had done so in production. Hmmm ...

[CSchmidlapp]

Bar the Video master, the scan, digitizing, compositing, film out, all could add various amounts of inconstancy.
It was probably colour graded to a degree scene by scene during the process also.

Great work good sir.

[Doctor M]

I don't recall it ever being mentioned, but is there anything useful in GitS 2.0 that could be used as a reference?
They may have monkeyed with a lot of scenes, but for anything they didn't should be well restored from the source.

[zoidberg]

It would have definitely been colour timed which probably explains the 35mm's teal. Even Toy Story was colour timed despite being 100% filmout, by all accounts John Lasseter was somewhat clueless to the process. I'll find the video later.

The 2.0 is heavily cropped and regraded. I'm not sure if the composites (focus, movement of layers etc.) have been altered in any way from the original. One watch was enough for me

[CSchmidlapp]

Yes 2.0 was a horrible experience.
Why they didn't remaster the original in the process is beyond me!
I looked at it in detail a few years ago for a project I was working on for exactly the reasons you stated Doctor M, and like Zoidberg has stated it's not very useful at all
Would love to see that Toy Story clip!

[zoidberg]

Here's the video I mentioned earlier:

[BronzeTitan]

Great work good sir.

Thanks! I love this digging-for-gold research (my pedestal is full of the treasures of the gods    ).

I also came across this Siskel & Ebert review of The Ghost In The Shell YouTube video from 1996. That's probably the closest we'll ever get to the original release.