Info: Star Wars - What is wrong and what is right... Goodbye Magenta — Page 12

Ronster said:

Bullshit. That is through a scaler.

Damn it.

So, Ronster, what is the exact model of your Samsung HDR 4k tv? You linked to a MU6100 calibration page on pg. 3.
Is MU6100 the model listed on the back of your TV?

Grayed out calibration options with different sources is not a feature, that sounds like a bug to me.

The MU6100 has a firmware update as recent as this month:
https://www.samsung.com/us/support/owners/product/2017-uhd-smart-tv-mu6100

Are you up to date on that?

It knows it is a PC, those options are available depending on what is connected to it. My TV is not on internet I only use it as a monitor.

I used the Nvidia control panel is much better than windows built in one.

Is using an HDR monitor really the best way to color correct non-HDR sources? I fully admit i have no idea, but it sounds to me like maybe you shouldn’t be using that as a color correction monitor.

dahmage said:

Is using an HDR monitor really the best way to color correct non-HDR sources? I fully admit i have no idea, but it sounds to me like maybe you shouldn’t be using that as a color correction monitor.

For a start I can’t even run the screen in native resolution as I have nothing 4K to put in or any HDMI 2.0 devices.

Rocky Horror is perhaps a good benchmark for the right time 1975 and how a good restoration should look.

This was quick in Gimp but Master Brightness and Contrast about +23 on both to bring this up to the same sort of level as Rocky Horror.

Before Brightness Contrast change. But Hue Shifted.

Original GOUT image color

As you can see I don’t want to do an awful lot to it anyway. This is basically the complete process of what I want to do. Apart from fixing the parts where the GOUT is no Good I plan to use the 97 Laserdisc.

Swingback on Cyan reveals Blue and Red Vader Chest Plate…Also notice the weird staining disappear on the left of the screen. This is about 45% out on hue for Cyan.

Here is my original Gout Shift

The process I am thinking of employing is the Purple swing back as well as the Brightness and contrast. This will be done via Selective Correction I also added some yellow in the Shadow here.

Basically I will do what ever I think it needs to get it looking better. Another Cyan Stain on the left mind you… It might be 45% in opposite direction this time?

Ronster said:
Caliberate a monitor to somone elses monitor is nonsense. I even read a forum on why companies do not calibrate before sending out to customers and how depending on the light in the room all this bollocks. You also have specification to contend with now like HDR color depth resolution and so on.

So, you are calling the most professional people and their recommendation nonsense. Interesting. You are telling us that Lucasfilm and THX made this whole monitor calibration process that is nonsense. You are telling us that Microsoft built this feature into Windows because it is nonsense.

RUBBISH!!!

Your view is nonsense. Every monitor needs to be calibrated so that reds are true red and blues are true blue. They need to be calibrated out of the box to correct for any oddities in manufacturing. They need to be periodically recalibrated to account for flaws that develop with age.

You calibrate to match for a purpose but in an environment where everyone has a different make / model and specification or type OLEd / LED / LCD / Plasma / projector whatever utterly pointless it is the content that needs to be tailored.

No you calibrate monitors and projectors so that the image is consistent and everyone sees the same thing. Otherwise you will get complaints that it is too yellow, too washed out, too blue, or the blacks are too crushed. While it is true modern monitors come close to an ideal calibration, it is still close. Anyone who works with colors professionally will tell you that a monitor must be calibrated. Why? Video is not the only area. Print is another. If you have a properly calibrated monitor, you can be assured that the color you pick in your office with match (within the known margin of error when switching between computer RGB and Print inks) that billboard and add campaign you were paid a million to do.

The content is not ever tailored. The content is supposed to be consistent. The content is the color master and the display is the end product. You don’t change your color master for a particular venue. Technicolor made a business out of consistent and accurate color reproduction. If they were given a mis-colored master to work from, all the prints would be wrong. You have your process ass-backwards.

you try calibrating an anti aircraft gun to be more like a machine gun never ever through calibration will you get the 2 things to match.

Ah, guns. You calibrate the gun not the bullets. Just as you calibrate the Monitor not what it displays. All guns should fire to hit the target. It doesn’t matter if it is a pistol, a machine gun, a sniper rifle, and anti-aircraft gun, a tank, a canon, or a big battleship gun. They are all calibrated so that the shell put in hits the target you aim at. Just the same way all monitors should hit the color that the source material is aimed at. Film makers don’t make their film and guess how the people will see it. They make their film with the expectation that when their characters talk about red, gold, green, purple, etc, that that is the color that shows on the screen. Lucas setup THX with picture and sound in mind so that when you experience a THX showing that it will always be the right sounds and the right colors. That is why you calibrate your damned monitor and not the source material The monitor is just a tool to show you what the filmmaker intended. You never calbirate the source to the type of screen it is shown on. Home theaters have had CRT’s, RBB projection, LCD projection, Plasma, florescent backlit LCD, fluorescent backlit LED, and true LED and in each case the makers of the product are striving to calibrate to produce the most lifelike colors. Good ones get close, bad ones are really bad. We took back one TV because the reds were too red and we could not reduced the glaring colors. I currently have a Samsung that I did a lot of tweaking on and it is the first TV I’ve had where I didn’t have to tweak the colors a lot.

So if you have been using uncalibrated monitors, you have been doing it wrong for years. You need to change what you are doing and calibrate your monitor to be sure it is producing accurate colors. If you don’t, not one of your color correction attempts is going to come out right.

You need to be sure that when you see these colors, these are the colors you are getting:

You need to be sure that when you see these colors, these are the colors you are getting:

This really does not hold up when you consider Vader’s chest plate is Blue yet it is green in some shots.

I ask you to be reasonable about how brutal this might be… I told you and you just want red faces Green chest plates so on. That is what you want no matter what I am saying you want it to look fucked up. I don’t.

45% shift in Cyan is huge… How the hell does that happen?

Please answer this rather than keep having a go at me otherwise I quit this site.

Ronster said:

You need to be sure that when you see these colors, these are the colors you are getting:

This really does not hold up when you consider Vader’s chest plate is Blue yet it is green in some shots.

This has nothing to do with having a calibrated monitor. This is due to the source you are working with, which is flawed.

Please answer this rather than keep having a go at me otherwise I quit this site.

Answer what? What you are saying is complete nonsense.

Ronster said:
It knows it is a PC, those options are available depending on what is connected to it. My TV is not on internet I only use it as a monitor.

I used the Nvidia control panel is much better than windows built in one.

Then you can install the firmware using a USB drive. You don’t need to be connected to the internet. Instructions on how to upgrade manually are included on that same link I posted above.

I understand that those options are grayed out for you depending on what is connected. I’m saying that I don’t think that is by design. I believe you are experiencing an error.

Again, what is the model of your TV? You know, the info listed here:

You keep straying off topic. The topic I keep addressing is to calibrate your monitor.

You have erroneously claimed that YCbCr is better, but that is just a different way of representing RGB color (allowing higher compression of the color and lower compression of the luma to give the illusion of a higher resolution image. It has nothing to do with the colors. That would be the bit encoding. In order to truly change the color space you have to switch from an RBG encoding to a CYM/CYMB encoding.

And HDR has nothing to do with color and everything to do with blacks and whites. With a backlit screen it is all about getting the black as dark as possible. OLED takes advantage of having 3 lights per pixel to achieve true black (as in no light at all for that pixel). The whites are just how bright the light can be. Again a feature of the display, not the source.

So the things you are citing as important are not really important at all. Not to a discussion about color correcting the original source material.

But let’s get back to what is wrong with the GOUT.

These are the stages of what film goes through for duplication. The exception is presentation prints which are generated directly from the negative, as are the dailies.

What you see on the interpositive is a general orange color. The image is lighter as well. Traces of this can be seen in the older home video of the OT. Things tend to be a touch orange and the contrast isn’t as high as it should be. This makes grays a touch red and skin tones a bit redder and blacks more gray. As you can see, this orange tint is reversed for the internegative and then canceled out on the final prints. For whatever reason, interpositives have often been used for telecines and a lot of telecines have this combination of orange and gray. In many shots in the OT, all teleciens look about the same. In others, they can be very different.

So to correct the GOUT to compensate for this trace of orange and gray, you really need a sample of what it should look like. Enter the Technicolor prints. Mike Verta was nice enough to post a sample of his scan. From that and the known green issues, I created a correction. Doing the same thing from frames of a different Technicolor print he acquired, DrDre did a correction. Technicolor is a CYM printing technique that doesn’t fade easily. Many old color movies exist for home viewing because a Techncolor print is available. There really is no other way to accurately find the original colors for A New Hope. By the time of the SE in 1997, the negative had faded badly and inconsistently and they used Lucas’s own Technicolor print to restore the colors for the SE. You have said you like the colors of the SE Telecine. That is all well and good, but those aren’t very reliable colors.

So either you rely on the Technicolor samples, or you do some research and see if you can find references that will help. I did both. I have read the research on the correct blue for R2. I have seen the parts with my own eyes to confirm it (the original units were more purple). I have studied flesh tones. I have extensive experience mixing paints for art and the correct way to make skin tones involves yellow and red to get the right peachy orange color. I have watched movies by the actors made at different times by different studios. There is a nice array of natural skin tones. Some are paler and pinker. Some are darker and tanner. Some are redder. Some places add to the red tone (our blood is red after all and the only way our skin is naturally not at least somewhat pinkish is if we are sick. Jaundice makes us yellow. What I see in your corrections (and in my early attempts) are people who should go see a doctor immediately for jaundice. One thing I did (and keep doing) is to compare my skin tone to others. Just in our office meetings I am presented with a nice variety of skin tones. When I keep those images in mind as I color correct, I have to keep the yellow to a minimum. Too much and the image is ruined and the people look sick. Too little and they all look sunburnt. And I found that Owen should look a bit singed in all his shots. Most of the humans on Tatooine should look like they’ve been in the sun. The people should be decidedly redder than the rocks (they didn’t film any of Tatooine at Red Rocks in Colorado after all) and the rocks should be a nice tan color.

And what I’ve found in all that research has led to my color correction and my near agreement with DrDre. I think he takes a bit too much red out of some shots. You take way too much red out and the DVD/HD/BR left too much yellow out.

I hate to argue with you, Poita, but I don’t think the colors in Star Wars/A New Hope are as bad as you think. I think what you are referring to are the Technicolor prints which I have seen are quite messed up. But with the process winding down and Star Wars being the last film with commercial Technicolor prints, it isn’t a surprise that they screwed up the colors. And I don’t think you are correct about the telecines. While it is true that a telecine operator can tweak things on the fly, the overall consistency between the various transfers indicates that is not what we are seeing. Team Negative 1 and Puggo have both transferred multiple 35 mm and 16 mm prints to arrive at the Silver Screen and Puggo Grand presentations. And starting with the Moth3r bootleg transfer (apparently done from a release print not an interpositive) and going through the old pan & scan releases and the widescreen releases (multiple in each country and in at least 6 countries and from old and new interpositive prints), all the transfers show are marked similarity and a stark difference from the Technicolor prints. That many different telecine operator cannot all have done exactly the same corrections on the fly. That color timing has to be from the prints themselves. That the 35 mm and 16 mm scans agree with that says to me that the Technicolor color timing is very screwed up. The sources appear to vary between original May 1977 prints (Puggo Grand US, Moth3r, and the early US/UK telecines), later 1977 prints (all the non Technicolor 35 mm prints and early foreign telecines), and the fresh interpostives done in the mid to late 80’s. But the results are all pretty close. That body of fairly consistent color timing in all those transfers vs. what we see in the Technicolor scans to me indicates that the Technicolor prints, while low fade and high resolution, are not representative of the color timing of the optically generated prints. The scene I have taken note of is where Ben and Luke are talking in the canyon. It is so washed out in the technicolor scans while it is so vibrant in every other scan and telecine.

The only way this could get any better is if mverta suddenly appeared to add to the discussion.

Very informative posts ^^ and I enjoyed reading them.

poita said:

I have spoken to Mark Wileage who did the colour for the SE, and I have an almost unfaded Kodak print here, it isn’t very different at all to the Tech IB other than the slight green in the IB being absent.

Mark did color for the 2004 DVD release - I worked at Technicolor at the time and his description of the conditions and timeframe were pretty dismal and surprising. It’s amazing the DVDs look as good as they did - Did he do the SE as well?

Well if you ignore anything that has to do with not calibrating your monitor. Then yes, this thread is great.

Maybe it should be retitled - Information on Star Wars color timing.

Here is the thing, when I look at the timeline of how the movie was released and the telltale signs that the process left that were archived in the home media releases, I can’t reconcile that with what is being said.

There were 5 edits to the movie between the 16 mm US release (the main source for Puggo Grand) and Moth3r bootleg and the fresh print used for the Definitive Edition/Faces/2006 GOUT. The obvious one is the opening credits. But 3 FX shots were also changed as were the end credits. Now, I’ll ignore the changed opening sequence because we all know when and why it was changed. So far, all the 35 mm prints that this community has found and used, be they optical transfers or Technicolor prints, are have the other 4 changes. The key to understanding when those 4 changes were made lies in the release sequence after the initial release (which was 35 and 70 mm with only the Dolby encoded stereo track and the 70 mm 6 track). Both Puggo Grand and Moth3r have a stereo mixdown mono track instead of the real mono track created later in the year. By the time they made the international release, they had the mono track done and the edits had been made. Not one international version has been found without the 4 changes. However, all the US/UK (as in English) home video releases between the Moth3r bootleg and the Definitive Edition are missing the 3 FX changes.

Also, being an avid movie fan, I have encountred troubled telecines in the past. The transfer for one of Hao Miyazaki’s films carried a bad orangy tint. They quickly did a corrected version and blamed it on the source being an interpostive print that wasn’t fully corrected. The information I’ve found on the Definitive Edition LD release indices it was done from a new interpositive print. The color is off on the DE/Faces/GOUT US/UK version in a manner that is consistent with an interpositive print (slightly orange and low contrast).

And back to those earlier US/UK and Japanese releases that have the original version of those 3 FX shots, by the time the home video was being made, the negative had already been changed so these cannot have been done from a print just for Telecine as any new print from the negative would match the international releases (and also the later 35 mm prints and the print the DE was made from as well as the SE which still has two of those scenes). From my understanding, an interpositive matches the parameters described and doesn’t have to be struck separately. Also, being a lover of old movies, all of the old Technicolor movies that saw relseases on home video (tape, LD, and DVD) were telecined from release prints, so telecine operators are obviously familiar with using many types of sources and can work with what they are given.

Which brings me back to some of the scenes from the technicolor prints that bother me. Particularly that scene where Ben and Luke are talking in the canyon. There is little color to recover in that scene and yet so many of the transfers show a nicely blue sky behind Ben. That scene and those like it (which seem to be mostly on Tatooine) are timed very differently between the Technicolor version and all the other versions. To me that indicates that when they created the 3 color master that starts the Technicolor process, they did not get the timing settings right. And they didn’t get it right in an oddly consistent way. Plus there is the green tint that we already know about. If they screwed up the timing in one place, maybe they did on the whole thing. That is why I say careless. Once the master is made, it is hard to screw up the rest of the process unless you fail to get the dyes the correct colors. But as a result of what I have seen from Mike Verta’s samples, DrDre’s samples, the 4k77 no-DNR version (which sticks to the Techncolor asthetic), and the samples of raw scans of both 4k77 and SSE, I don’t trust the colors on the Technicolor prints. And I bet if you could project one side by side with an optical print you would see the same odd timing. I don’t see how a telecine operator could pull that much color if it wasn’t already there. And it wouldn’t be just one. By my count, ANH has had more than 13 separate telecines from around the world. The scene of Ben and Luke and the canyon looks nothing like the Technicolor in any of them. That is way too many different telecine operators making the exact same (and nearly impossible) correction.

I will admit that the telecines are not identical. Some are more yellow, some more blue, some a bit orange. The JSC has a very orange Threepio on the opening sequence in a bluish corridor. I wasn’t arguing that. I was saying that what they point to for the colors are remarkably similar and that level of similarity over that many different copies from that number of different prints (there are details that indicate the THX foreign language transfers were new prints so that leads to at least 10 separate prints) points to very little tweaking during the telecine process.

And I am aware that the projection process changes things. If your goal is to relive how the movie was projected when you watch it on your TV, then correcting to the projected version makes sense. But that involves tweaking for blub color and a lot of variables that lead to different interpretations. My experience with photos (scanning from negatives, from prints, reprinting negatives, color correcting, repairing, etc) over the last 25 years says that the oddities I’m seeing in the Technicolor scans are not just because the other versions have been corrected or that there is some oddity in the scan that doesn’t show up when projected. I have enough experience to know that isn’t how it works, not to mention a lot of examples from the work done by our members. It is a flaw in the Technicolor print. It is washed out, low contrast, and the color seen in other transfers cannot be recovered from what is in the Technicolor print. Unless you experienced a Technicolor print back in the 70’s, that isn’t how you saw Star Wars. And the whole point of Technicolor prints was that you didn’t need special equipment. Ideally, after the era of Natalie Kalmus, a movie should look the same whether it is a Technicolor print or optical print. But if you have the last movie to get Technicolor prints and someone didn’t do things right, then you can end up with a superior process producing inferior prints. I think that is the case. Everything I’ve found points to that.

yotsuya = ronster part 2, never give up dude. maybe someone will listen.

deathstar1138 said:

yotsuya = ronster part 2, never give up dude. maybe someone will listen.

Nah, yotsuya actually grasps the core concepts at play, whether you agree with his conclusions or not.

ChainsawAsh said:

deathstar1138 said:

yotsuya = ronster part 2, never give up dude. maybe someone will listen.

Nah, yotsuya actually grasps the core concepts at play, whether you agree with his conclusions or not.

This