Fake colors on the restoration process?

Since I am not an expert in this area, could someone please explain to me how these digital restorations are is done?

I've been looking into DVD Beaver comparisons and noticed that they have removed artifacts like EE and others from the negatives, but sometimes the colors are too much vivid or false to me.

I mean, how is it possible that the previous picture is white and the new one is blue (Poltergeist shot)?

I will pick you three examples.

Poltergeist, 25th edition and the old copy:

dvdbeaver.com/film2/DVDReviews33/poltergeist.htm

Clockwork Orange - SE and Non-Anamorphic Edition:

dvdbeaver.com/film/DVDReviews16/clockwork_orange_dvd_review.htm

Matrix (1999)

dvdbeaver.com/film/DVDCompare10/the_matrix_.htm

If I am not mistaken, the correct restoration should make the skin of people not look false like the old editions, when it looks like they were in the beach too much time.

The palette of colors should look like original, not enhanced on the brightness or contrast if this was not the intention of the producers, but sometimes this is not adjusted properly on the restoration process.

What I can't understand is why that Clockwork Orange SE picture is smoother than the first and still is better from the perspective of the reviewer. Did I miss something here?

See Matrix for example. The third shot from DVD Beaver. We have a green picture on the new Ultimate Edition. Those green colors were seen on the theatrical release? Or they were just a modification made by Warner to make the movie look more green than it was before, on the previous DVD edition, VHS, TV broadcasts, etc.

See the Poltergeist picture where Craig T. Nelson is shot. The wall colors are modified to match all the others. Are these colors created by Warner, or they were original, seen on the cinemas? Every other aspects seem correct on the new SE edition. The skin color seemed correct this time. But it looks like the picture was 90% painted using the same colors.

I mean, it looks like the wall have almost the same color used for the skin. Even if the skin color looks more natural now. It looks like, from a naked eye, that everything was painted using the same colors.

Re: Fake colors on the restoration process?

The first Poltergeist DVD is almost ten years old, so it's a matter of the new transfer using newer technology.

From what I can tell, there's minimal difference in A Clockwork Orange, except that the '07 DVD is a bit sharper because it's 16x9. The '01 DVD was virtually perfect except for being non-anamorphic. The '99 DVD should be compared since it looked awful.

The Matrix was originally released with conventional 35mm prints that were color timed photochemically. The weird tinting had to be achieved with filters and other processes. The remaster used the same digital color timing tools as the sequels, so they could use a more pure green look.

As for what colors look "natural"... it's not really a good indicator. A lot of films end up looking exactly as they should on DVD, only for them to get complaints about not looking right since white levels aren't pure or lipstick isn't the right shade of red.

(Also, they don't remove edge enhancement from film negatives)

Re: Fake colors on the restoration process?

One of the major problems with remasters of older fims is that the source material ,ay have faded since the previous transfer was made. While digital color correction has come a long way, it can't completely compensate for faded source elements.

Often times, it's a question of making do with what you have from the source elements in terms of color correction which often times results in subtle differences in color saturation and tints between transfers.

(And negatives don't have edge enhancement) Interpositives might if there was a digital intermediate somewhere in the chain, however.

Vern

Re: Fake colors on the restoration process?

Absolutely right. We can accept a huge range of permutations of lighting as 'natural' in real life and often don't notice these until they are shown to us in photos of in films. Take one small example - when you walk from outside into a lighted room, do you notice the ambient lighting change from blue to yellow? Thought not. But have a look at unfiltered film of someone moving from outdoors to indoors and you will notice a massive change, even if the brilliance of the light is controlled for.

Re: Fake colors on the restoration process?

Surely if a film was produced using a digital intermediate, then the D.I. would be considered THE master of the film, and would be used to make all lower resolution video transfers?

Re: Fake colors on the restoration process?

Film coloring changes because celluloid colors fade and the telecine supervisor has to make his own judgement about how to deliberately manipulate the color so as to be accurate; sometimes it is, sometimes its not.

Films taken from the O-negs have to basically be re-timed from scratch--the negative does not have any color timing on it, thus coloring often changes drastically (ie a nighttime scene might have been shot "neutral" on the negative but really the DP intended to have it tinted blue in the timing, but the telecine colorist has no idea about this; or maybe it was intended to be neutral looking but the colorist assumes that it is meant to be blue tinted). In these cases it is essential to have some kind of faithful reference for color information.

Oftentimes, directors and DP's are involved in transfers to make sure the telecine colorist gets it right, has the precise subtlties and is informed of all the color intentions. But sometimes you also get into cases where directors and DP's have unreliable memories but everyone assumes they are the ultimate authority, or they flat-out change their minds about how they want something to look. And in other cases they know exactly how something looked in theaters or on the negative but they really wanted some other effect but couldn't achieve it and will now use digital tools to get the color effects that were intended but never implemented.

Its hard to always get it right, and that last issue introduces a philosphical dillemma for "purists".

Re: Fake colors on the restoration process?

That's when a director-approved dye-transfer reference print would come in handy.

Re: Fake colors on the restoration process?

I've always been a strong advocate of this. Unfortunately this isn't really even an option anymore (wasn't Apocalypse Now Redux printed on Technicolor dye-transfer or something? I don't know where they would have gotten that done; I know one place in China used to make dye-transfer prints but I thought these basically became obsolete in the early 80's).

Re: Fake colors on the restoration process?

There was a renaissance of dye transfer in the late-90s, but I don't know how many houses are handling it anymore. It appears that Technicolor stopped dye transfer printing shortly after its acquisition by Thomson.

Re: Fake colors on the restoration process?

There were a few dye transfer prints made for the U.S. premiere of Pearl Harbor. There was an article about the new dye transfer in American Cinematographer mag, but apparently now all of Technicolor's Dye Transfer machinery is in storage.

Re: Fake colors on the restoration process?

hmm....in the context of DVD mastering (which is low res), that's true
but film emulsions do have "adjacency effects", or the so called "mackie line", which is caused by diffusion in the development stage. So when the enlargement is strong enough (like for 16mm 8mm) or when the emulsion-effects (like grain, diffusion etc.) are big enough to be seen in 35mm (like in eary eastmancolor) you can sometimes see something that looks quite like edge enhancement, only more pleasant and not so harsh.

It can also be done on purpose with B&W film, using high acutance developers.

Also, Technirama films, used zirconium for a tanning effect on the negative, making a kind of relief image, which created edge effects in prints, making it look slightly sharper than non-technirama negatives.
Tanning is also used in classic B&W darkroom