The theatrical colors of the Star Wars trilogy — Page 7

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

DrDre said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

I have around 100-200 Star Wars technicolor print frames as well from different parts of the film, I bought them from an ebayer a while back when I was researching the colours for the Star Wars Semi-Specialised Edition V2.5, perhaps from the same person you did. They’re lovely to look at but as NeverarGreat said before, the colours can often be inconsistant from shot to shot, meaning every frame can’t be taken as exact colour references unless you plan to replicate those shot by shot colour inconsistancies. I have a feeling that every Star Wars technicolor print also looked a little different from each other colourwise as well.

I’m curious, are you planning to catalog these print scans just as a reference for everybody here to get a better feel for the colours of a Star Wars technicolor print or do you plan on creating a Star Wars release graded to a technicolor print similar to what NeverarGreat has been working on?

Wouldn’t these frames also look somewhat different when projected, accounting for the warmth of the bulb etc.? Can you really just look at the frames themselves?

kk650 said:

DrDre said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

I have around 100-200 Star Wars technicolor print frames as well from different parts of the film, I bought them from an ebayer a while back when I was researching the colours for the Star Wars Semi-Specialised Edition V2.5, perhaps from the same person you did. They’re lovely to look at but as NeverarGreat said before, the colours can often be inconsistant from shot to shot, meaning every frame can’t be taken as exact colour references unless you plan to replicate those shot by shot colour inconsistancies. I have a feeling that every Star Wars technicolor print also looked a little different from each other colourwise as well.

I’m curious, are you planning to catalog these print scans just as a reference for everybody here to get a better feel for the colours of a Star Wars technicolor print or do you plan on creating a Star Wars release graded to a technicolor print similar to what NeverarGreat has been working on?

It is true that technicolor prints vary in color. However, the original grading itself also suffers from inconsistencies, as poita also noted in the past. I personally feel these need to be replicated for a print restoration project. Of course in color grading a semi-specialised edition the goals are somewhat different, and I would go with what’s aesthetically appealing. I plan to catalogue them as a reference for everybody, but I’m also working on another project, which involves golor grading a print scan. However, to know more about that I advice you to contact williarob.

Handman said:

Wouldn’t these frames also look somewhat different when projected, accounting for the warmth of the bulb etc.? Can you really just look at the frames themselves?

I personally feel the best course is to first replicate what’s on the print (at daylight conditions 5500K), and then adjust them to the desired color temperature. As NeverarGreat noted the reports on these lamps are conflicting, and a 35mm collector I know told me, that in his experience the difference between 1970s lamps and modern lamps is very minimal when projecting a print. Both lamps are in principle daylight-balanced, but I also believe carbon lamps tend to burn at a lower temperature as they age (becoming warmer), so that’s also a thing to consider.

NeverarGreat said:

Handman said:

Wouldn’t these frames also look somewhat different when projected, accounting for the warmth of the bulb etc.? Can you really just look at the frames themselves?

There are conflicting reports on the temperatures of these lamps:

correlated color temperature of a white flame carbon arc lamp is 5800K

Xenon arc lamps are approximately 5400K

Which doesn’t make sense since Xenons are supposed to be bluer than the carbon arc lamps.

Here, David Mullen says this about it:
“They (Carbon Arc and Xenon projectors) should be similar (daylight-balanced, like around 5500K) but Xenons are slightly bluer I think. But since carbon arc projection is extremely rare these days, most people color-time for Xenon lamp projection. It’s a bit of an issue when projecting old Technicolor dye transfer prints with modern Xenon lamps since the print looks slightly cooler than it used to, since it was timed for carbon arc projection.”

Also mentioned somewhere is the fact that carbon arc light was notoriously inconsistent since the ends of the carbon would burn at a variable rate.

Here’s a source that gives strangely low numbers, but at least it shows that the carbon arc lamp is yellower: http://inspectapedia.com/electric/Bulb_Color_Temperatures.php
Projector bulbs, carbon arc (traditional movie) 3,900K
Xenon arclamps 4,100K
This image shows just how similar the lights are, if the 200K difference can be believed:

The difference is even less than the difference between two of those panels.

So to answer your question, it appears that you would need to adjust the color of technicolor film about 200K yellower to counteract the grading for carbon arc lamps.

The Brightness of the bulb has very little to do with how the film should look more to do with how far you need to project an image through the air. The Bigger the cinema and a larger sized screen the brighter the bulb you would need. If you were in a small pokey cinema you might have a a less bright bulb but end up with a better brighter image than a very large cinema.

Think about distance and size of projected image vs the brightness of the bulb. but projected images will never look the same as anything on a TV screen ever, because projected images pass through air via light and TV screens display straight on to the display surface without any or very little atmospheric interference.

The temperature would be relative to how far the throw is of the projected image vs the temperature of the lamp before it hits the screen material (drop off) which is either reflective (slight gain) or permeable (greater loss).

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

I have around 100-200 Star Wars technicolor print frames as well from different parts of the film, I bought them from an ebayer a while back when I was researching the colours for the Star Wars Semi-Specialised Edition V2.5, perhaps from the same person you did. They’re lovely to look at but as NeverarGreat said before, the colours can often be inconsistant from shot to shot, meaning every frame can’t be taken as exact colour references unless you plan to replicate those shot by shot colour inconsistancies. I have a feeling that every Star Wars technicolor print also looked a little different from each other colourwise as well.

I’m curious, are you planning to catalog these print scans just as a reference for everybody here to get a better feel for the colours of a Star Wars technicolor print or do you plan on creating a Star Wars release graded to a technicolor print similar to what NeverarGreat has been working on?

It is true that technicolor prints vary in color. However, the original grading itself also suffers from inconsistencies, as poita also noted in the past. I personally feel these need to be replicated for a print restoration project. Of course in color grading a semi-specialised edition the goals are somewhat different, and I would go with what’s aesthetically appealing. I plan to catalogue them as a reference for everybody, but I’m also working on another project, which involves golor grading a print scan. However, to know more about that I advice you to contact williarob.

If I was doing such a print restoration project I would choose the technicolor frame with the best colour that I had from a certain scene and regrade the whole scene to be consistant with that frame. I know that the original Star Wars prints were supposed to be notoriously inconsistant colourwise from shot to shot but I think people watching films now have certain expectations when it comes to consistant colours in a scene on home releases. It’s an option that I think you and williarob should consider.

kk650 said:

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

I have around 100-200 Star Wars technicolor print frames as well from different parts of the film, I bought them from an ebayer a while back when I was researching the colours for the Star Wars Semi-Specialised Edition V2.5, perhaps from the same person you did. They’re lovely to look at but as NeverarGreat said before, the colours can often be inconsistant from shot to shot, meaning every frame can’t be taken as exact colour references unless you plan to replicate those shot by shot colour inconsistancies. I have a feeling that every Star Wars technicolor print also looked a little different from each other colourwise as well.

I’m curious, are you planning to catalog these print scans just as a reference for everybody here to get a better feel for the colours of a Star Wars technicolor print or do you plan on creating a Star Wars release graded to a technicolor print similar to what NeverarGreat has been working on?

It is true that technicolor prints vary in color. However, the original grading itself also suffers from inconsistencies, as poita also noted in the past. I personally feel these need to be replicated for a print restoration project. Of course in color grading a semi-specialised edition the goals are somewhat different, and I would go with what’s aesthetically appealing. I plan to catalogue them as a reference for everybody, but I’m also working on another project, which involves golor grading a print scan. However, to know more about that I advice you to contact williarob.

If I was doing such a print restoration project I would choose the technicolor frame with the best colour that I had from a certain scene and regrade the whole scene to be consistant with that frame. I know that the original Star Wars prints were supposed to be notoriously inconsistant colourwise from shot to shot but I think people watching films now have certain expectations when it comes to consistant colours in a scene on home releases. It’s an option that I think you and williarob should consider.

In my view that would be going into special edition teritory. The color inconsistencies are part of the original viewing experience, just like matte lines, the orange blob, etc. So, for a print restoration I would not try to improve the original presentation. I believe, the goal of a print restoration should be to restore the print to it’s original state.

So Dre, Is your idea to fine tune your GOUT regrade with these sources? Or are you looking to completely restart the timing process with these as reference?

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

The blue highlights on Threepio are the reflection of the sky. This last image also lacks the reds that should be in the faces in the exterior Tatooine shots. The image has been over adjusted to make their skin tones look good when all the on set photos I’ve come across also show very red tones. Basically it has been over corrected. Though my one issue with DrDre’s new scan is that the darker areas seem over saturated to the point where detail is lost, but that could be the print itself. But the colors seem on the money to me. His color sense is excellent and he is using very good equipment to scan the frames.

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

I have around 100-200 Star Wars technicolor print frames as well from different parts of the film, I bought them from an ebayer a while back when I was researching the colours for the Star Wars Semi-Specialised Edition V2.5, perhaps from the same person you did. They’re lovely to look at but as NeverarGreat said before, the colours can often be inconsistant from shot to shot, meaning every frame can’t be taken as exact colour references unless you plan to replicate those shot by shot colour inconsistancies. I have a feeling that every Star Wars technicolor print also looked a little different from each other colourwise as well.

I’m curious, are you planning to catalog these print scans just as a reference for everybody here to get a better feel for the colours of a Star Wars technicolor print or do you plan on creating a Star Wars release graded to a technicolor print similar to what NeverarGreat has been working on?

It is true that technicolor prints vary in color. However, the original grading itself also suffers from inconsistencies, as poita also noted in the past. I personally feel these need to be replicated for a print restoration project. Of course in color grading a semi-specialised edition the goals are somewhat different, and I would go with what’s aesthetically appealing. I plan to catalogue them as a reference for everybody, but I’m also working on another project, which involves golor grading a print scan. However, to know more about that I advice you to contact williarob.

If I was doing such a print restoration project I would choose the technicolor frame with the best colour that I had from a certain scene and regrade the whole scene to be consistant with that frame. I know that the original Star Wars prints were supposed to be notoriously inconsistant colourwise from shot to shot but I think people watching films now have certain expectations when it comes to consistant colours in a scene on home releases. It’s an option that I think you and williarob should consider.

In my view that would be going into special edition teritory. The color inconsistencies are part of the original viewing experience, just like matte lines, the orange blob, etc. So, for a print restoration I would not try to improve the original presentation. I believe, the goal of a print restoration should be to restore the print to it’s original state.

Hmm, when I think of Special Editions, I think of Jabba in A New Hope, Greedo shooting first and Hayden Christenssen at the end of Return of the Jedi, I think of changed content, not consistant colour grading. I think trying to make the original unaltered theatrical cut of Star Wars more appealing to modern audiences is a worthwhile endevour and doesn’t make it a Special Edition. Does anybody here actually believe that if Disney does release the unaltered Original Trilogy, that they won’t make sure that they maintain colour/fleshtone consistency on a scene by scene basis, even if its not 100% accurate to the theatrical prints?

I think creating an appealing colour consistant version of the original unaltered Star Wars that becomes popular with the mainstream general public could actually be a great way of showing Disney that there’s real interest in these unaltered releases and make it that much more likely that they’ll be released as well when the inevitable 4k remastered Original Trilogy Special Editions come out.

yotsuya said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

The blue highlights on Threepio are the reflection of the sky. This last image also lacks the reds that should be in the faces in the exterior Tatooine shots. The image has been over adjusted to make their skin tones look good when all the on set photos I’ve come across also show very red tones. Basically it has been over corrected. Though my one issue with DrDre’s new scan is that the darker areas seem over saturated to the point where detail is lost, but that could be the print itself. But the colors seem on the money to me. His color sense is excellent and he is using very good equipment to scan the frames.

I’m not exactly sure what you think over corrected means Yotsuya. When I think of over corrected, I think of a frame that had a blanket blue tint like in the second frame and too much is removed in correction, making it too yellow, or a frame that is too contrasty and too much contrast is removed and the frame ends up looking flat and washed out. The first frame has a blanket red tint and the second has a blanket blue tint, both are oversaturated, with too much contrast for home viewing. You are aware that set photos are not 100% reliable when it comes to deciding what the colours/image dynamics in a film should be right? Those set photos get manipulated for publication and the original film footage also gets regraded in post.

It doesn’t matter how red the fleshtones are in set photos, what matters is getting as balanced colours as possible with the frame/transfer you’re working with and that for me means getting the contrast and saturation within normal levels for home viewing, getting fleshtones looking fairly natural relative to the light sources they’re exposed to and getting rid of any blanket tints like the blanket red tint in the first frame and blanket blue tint in the second frame, from the OT films we know the daytime sunlight on tatooine is supposed to be normal for a hot desert country, a warm yellowish colour, not tinted red or blue.

I’m confident that there is no blanket tint in that frame I posted and that the brightness/contrast/saturation settings are correct in the context of the blu-ray’s colour scheme because a single setting was used to grade the vast majority of the film, including this whole scene with Ben. White scenes on Tantive IV and other scenes from the film like the ones below were used to carefully calibrate that single overall setting and remove any blanket tints on the blu-ray and get the best balance of colours possible:

Blu-ray 1:

Star Wars Semi-Specialised Edition V2.6 Regraded 1:

Blu-ray 2:

Star Wars Semi-Specialised Edition V2.6 Regraded 2:

Blu-ray 3:

Star Wars Semi-Specialised Edition V2.6 Regraded 3:

Blu-ray 4:

Star Wars Semi-Specialised Edition V2.6 Regraded 4:

Blu-ray 5:

Star Wars Semi-Specialised Edition V2.6 Regraded 5:

Blu-ray 6:

Star Wars Semi-Specialised Edition V2.6 Regraded 6:

Blu-ray 7:

Star Wars Semi-Specialised Edition V2.6 Regraded 7:

Blu-ray 8:

Star Wars Semi-Specialised Edition V2.6 Regraded 8:

Blu-ray 9:

Star Wars Semi-Specialised Edition V2.6 Regraded 9:

Blu-ray 10:

Star Wars Semi-Specialised Edition V2.6 Regraded 10:

Blu-ray 1:

Star Wars Semi-Specialised Edition V2.6 Regraded 1:

Blu-ray 2:

Star Wars Semi-Specialised Edition V2.6 Regraded 2:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

I have around 100-200 Star Wars technicolor print frames as well from different parts of the film, I bought them from an ebayer a while back when I was researching the colours for the Star Wars Semi-Specialised Edition V2.5, perhaps from the same person you did. They’re lovely to look at but as NeverarGreat said before, the colours can often be inconsistant from shot to shot, meaning every frame can’t be taken as exact colour references unless you plan to replicate those shot by shot colour inconsistancies. I have a feeling that every Star Wars technicolor print also looked a little different from each other colourwise as well.

I’m curious, are you planning to catalog these print scans just as a reference for everybody here to get a better feel for the colours of a Star Wars technicolor print or do you plan on creating a Star Wars release graded to a technicolor print similar to what NeverarGreat has been working on?

It is true that technicolor prints vary in color. However, the original grading itself also suffers from inconsistencies, as poita also noted in the past. I personally feel these need to be replicated for a print restoration project. Of course in color grading a semi-specialised edition the goals are somewhat different, and I would go with what’s aesthetically appealing. I plan to catalogue them as a reference for everybody, but I’m also working on another project, which involves golor grading a print scan. However, to know more about that I advice you to contact williarob.

If I was doing such a print restoration project I would choose the technicolor frame with the best colour that I had from a certain scene and regrade the whole scene to be consistant with that frame. I know that the original Star Wars prints were supposed to be notoriously inconsistant colourwise from shot to shot but I think people watching films now have certain expectations when it comes to consistant colours in a scene on home releases. It’s an option that I think you and williarob should consider.

In my view that would be going into special edition teritory. The color inconsistencies are part of the original viewing experience, just like matte lines, the orange blob, etc. So, for a print restoration I would not try to improve the original presentation. I believe, the goal of a print restoration should be to restore the print to it’s original state.

Hmm, when I think of Special Editions, I think of Jabba in A New Hope, Greedo shooting first and Hayden Christenssen at the end of Return of the Jedi, I think of changed content, not consistant colour grading. I think trying to make the original unaltered theatrical cut of Star Wars more appealing to modern audiences is a worthwhile endevour and doesn’t make it a Special Edition. Does anybody here actually believe that if Disney does release the unaltered Original Trilogy, that they won’t make sure that they maintain colour/fleshtone consistency on a scene by scene basis, even if its not 100% accurate to the theatrical prints?

I think creating an appealing colour consistant version of the original unaltered Star Wars that becomes popular with the mainstream general public could actually be a great way of showing Disney that there’s real interest in these unaltered releases and make it that much more likely that they’ll be released as well when the inevitable 4k remastered Original Trilogy Special Editions come out.

Restoring a print and preparing an official release are two different things in my view. First of all contrast and saturation are different for a print than for a home video release. Secondly the color grading for home video releases are generally somewhat different from the theatrical release, and optimized for home view. This is not the objective of restoring a print. I’m of the opinion it’s oke for the person restoring the print to interpret the colors in the absence of accurate references, but if accurate color references are available, I don’t think it’s the restorer’s job to attempt to improve the colors, which is generally a pretty subjective endeavour. Also, having examined the frames themselves, I don’t believe the color imbalances are such, that anyone watching the films would notice. Consider the fact that home video releases such as the JSC suffer from color imbalances as well, and I only noticed those, when I directly compared frames from different shots.

Ps. I really like the v2.6 grading. It’s a great improvement over the previous version. Great job!

NeverarGreat said:

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

I have around 100-200 Star Wars technicolor print frames as well from different parts of the film, I bought them from an ebayer a while back when I was researching the colours for the Star Wars Semi-Specialised Edition V2.5, perhaps from the same person you did. They’re lovely to look at but as NeverarGreat said before, the colours can often be inconsistant from shot to shot, meaning every frame can’t be taken as exact colour references unless you plan to replicate those shot by shot colour inconsistancies. I have a feeling that every Star Wars technicolor print also looked a little different from each other colourwise as well.

I’m curious, are you planning to catalog these print scans just as a reference for everybody here to get a better feel for the colours of a Star Wars technicolor print or do you plan on creating a Star Wars release graded to a technicolor print similar to what NeverarGreat has been working on?

It is true that technicolor prints vary in color. However, the original grading itself also suffers from inconsistencies, as poita also noted in the past. I personally feel these need to be replicated for a print restoration project. Of course in color grading a semi-specialised edition the goals are somewhat different, and I would go with what’s aesthetically appealing. I plan to catalogue them as a reference for everybody, but I’m also working on another project, which involves golor grading a print scan. However, to know more about that I advice you to contact williarob.

If I was doing such a print restoration project I would choose the technicolor frame with the best colour that I had from a certain scene and regrade the whole scene to be consistant with that frame. I know that the original Star Wars prints were supposed to be notoriously inconsistant colourwise from shot to shot but I think people watching films now have certain expectations when it comes to consistant colours in a scene on home releases. It’s an option that I think you and williarob should consider.

In my view that would be going into special edition teritory. The color inconsistencies are part of the original viewing experience, just like matte lines, the orange blob, etc. So, for a print restoration I would not try to improve the original presentation. I believe, the goal of a print restoration should be to restore the print to it’s original state.

I thought so too for a long time, and graded the first half of the film to mirror the inconsistency. At some point I realized that the effect was too distracting and revised it to be consistent from scene to scene.

The problem with changing the color from shot to shot within a scene is that you’re no longer trying to sustain the illusion of a world within the film, but rather you are destroying the illusion of a world in order to create the illusion of a film. Yet the intent of the film is to create the illusion of a world, and the film was never intended to be inconsistent. Theoretically, each scene could have been consistent in terms of color, and this makes it different than matte lines, which couldn’t have been eliminated with 70’s technology.

Granted, my project is ‘the Special Edition in the best possible light’, so I have had to make peace with these sorts of interpretive changes.

I agree with you to a large extend, but when I look at the technicolor frames, I don’t see the level of inconsistency that would destroy the illusion of another world. I would agree with respect to the Mike Verta scans, which are all over the place, but the actual frames I have, don’t show this level of inconsistency. In fact I was surprised how balanced the frames look.

NeverarGreat said:

DrDre said:

NeverarGreat said:

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

It will be interesting to compare your regrade (and others) to the calibrated frame scans, once I’ve digitized them. I have a number of frames of this exact shot. I will also be getting a number of shots of the entry to Mos Eisley scene next week. I’m very interested what that scene looks like. Particulary the color of R5-A2 (orange or yellow?). Once I’ve started the scanning process I will create a separate thread for color references.

I have around 100-200 Star Wars technicolor print frames as well from different parts of the film, I bought them from an ebayer a while back when I was researching the colours for the Star Wars Semi-Specialised Edition V2.5, perhaps from the same person you did. They’re lovely to look at but as NeverarGreat said before, the colours can often be inconsistant from shot to shot, meaning every frame can’t be taken as exact colour references unless you plan to replicate those shot by shot colour inconsistancies. I have a feeling that every Star Wars technicolor print also looked a little different from each other colourwise as well.

I’m curious, are you planning to catalog these print scans just as a reference for everybody here to get a better feel for the colours of a Star Wars technicolor print or do you plan on creating a Star Wars release graded to a technicolor print similar to what NeverarGreat has been working on?

It is true that technicolor prints vary in color. However, the original grading itself also suffers from inconsistencies, as poita also noted in the past. I personally feel these need to be replicated for a print restoration project. Of course in color grading a semi-specialised edition the goals are somewhat different, and I would go with what’s aesthetically appealing. I plan to catalogue them as a reference for everybody, but I’m also working on another project, which involves golor grading a print scan. However, to know more about that I advice you to contact williarob.

If I was doing such a print restoration project I would choose the technicolor frame with the best colour that I had from a certain scene and regrade the whole scene to be consistant with that frame. I know that the original Star Wars prints were supposed to be notoriously inconsistant colourwise from shot to shot but I think people watching films now have certain expectations when it comes to consistant colours in a scene on home releases. It’s an option that I think you and williarob should consider.

In my view that would be going into special edition teritory. The color inconsistencies are part of the original viewing experience, just like matte lines, the orange blob, etc. So, for a print restoration I would not try to improve the original presentation. I believe, the goal of a print restoration should be to restore the print to it’s original state.

I thought so too for a long time, and graded the first half of the film to mirror the inconsistency. At some point I realized that the effect was too distracting and revised it to be consistent from scene to scene.

The problem with changing the color from shot to shot within a scene is that you’re no longer trying to sustain the illusion of a world within the film, but rather you are destroying the illusion of a world in order to create the illusion of a film. Yet the intent of the film is to create the illusion of a world, and the film was never intended to be inconsistent. Theoretically, each scene could have been consistent in terms of color, and this makes it different than matte lines, which couldn’t have been eliminated with 70’s technology.

Granted, my project is ‘the Special Edition in the best possible light’, so I have had to make peace with these sorts of interpretive changes.

I agree with you to a large extend, but when I look at the technicolor frames, I don’t see the level of inconsistency that would destroy the illusion of another world. I would agree with respect to the Mike Verta scans, which are all over the place, but the actual frames I have, don’t show this level of inconsistency. In fact I was surprised how balanced the frames look.

Fair enough, I look forward to seeing what you come up with 😃

I think it’s awesome to see the many different interpretations of the Star Wars color grading, yours, kk650’s, yotsuya’s, Swazzy’s, and many others. The quality level is just amazing!

kk650 said:

yotsuya said:

kk650 said:

DrDre said:

This is a very rough approximation, and is definitely not an accurate color reference, but the Mike Verta scans look something like this:

…whereas the print itself looks something like this:

On my main monitor the top image looks too red and the bottom image looks too blue, you can see it in the highlights on C3PO. There’s a very noticable difference in colour between those two frames, at least to my eyes. Both are oversaturated as well IMHO.

For those interested in comparing, here’s roughly that frame from the new Star Wars Semi-Specialised Edition V2.6 I just uploaded to tehparadox:

The blue highlights on Threepio are the reflection of the sky. This last image also lacks the reds that should be in the faces in the exterior Tatooine shots. The image has been over adjusted to make their skin tones look good when all the on set photos I’ve come across also show very red tones. Basically it has been over corrected. Though my one issue with DrDre’s new scan is that the darker areas seem over saturated to the point where detail is lost, but that could be the print itself. But the colors seem on the money to me. His color sense is excellent and he is using very good equipment to scan the frames.

I’m not exactly sure what you think over corrected means Yotsuya. When I think of over corrected, I think of a frame that had a blanket blue tint like in the second frame and too much is removed in correction, making it too yellow, or a frame that is too contrasty and too much contrast is removed and the frame ends up looking flat and washed out. The first frame has a blanket red tint and the second has a blanket blue tint, both are oversaturated, with too much contrast for home viewing. You are aware that set photos are not 100% reliable when it comes to deciding what the colours/image dynamics in a film should be right? Those set photos get manipulated for publication and the original film footage also gets regraded in post.

It doesn’t matter how red the fleshtones are in set photos, what matters is getting as balanced colours as possible with the frame/transfer you’re working with and that for me means getting the contrast and saturation within normal levels for home viewing, getting fleshtones looking fairly natural relative to the light sources they’re exposed to and getting rid of any blanket tints like the blanket red tint in the first frame and blanket blue tint in the second frame, from the OT films we know the daytime sunlight on tatooine is supposed to be normal for a hot desert country, a warm yellowish colour, not tinted red or blue.

I’m confident that there is no blanket tint in that frame I posted and that the brightness/contrast/saturation settings are correct in the context of the blu-ray’s colour scheme because a single setting was used to grade the vast majority of the film, including this whole scene with Ben. White scenes on Tantive IV and other scenes from the film like the ones below were used to carefully calibrate that single overall setting and remove any blanket tints on the blu-ray and get the best balance of colours possible:

Blu-ray 1:

Star Wars Semi-Specialised Edition V2.6 Regraded 1:

Blu-ray 2:

Star Wars Semi-Specialised Edition V2.6 Regraded 2:

Blu-ray 3:

Star Wars Semi-Specialised Edition V2.6 Regraded 3:

Blu-ray 4:

Star Wars Semi-Specialised Edition V2.6 Regraded 4:

Blu-ray 5:

Star Wars Semi-Specialised Edition V2.6 Regraded 5:

Blu-ray 6:

Star Wars Semi-Specialised Edition V2.6 Regraded 6:

Blu-ray 7:

Star Wars Semi-Specialised Edition V2.6 Regraded 7:

Blu-ray 8:

Star Wars Semi-Specialised Edition V2.6 Regraded 8:

Blu-ray 9:

Star Wars Semi-Specialised Edition V2.6 Regraded 9:

Blu-ray 10:

Star Wars Semi-Specialised Edition V2.6 Regraded 10:

Blu-ray 1:

Star Wars Semi-Specialised Edition V2.6 Regraded 1:

Blu-ray 2:

Star Wars Semi-Specialised Edition V2.6 Regraded 2:

By over correcting I mean that there is a problem with the source and in the process of correcting the problem you take it too far the other way. So the above frames correct the magenta of the BR but leave the frames far yellower than our most accurate sources would indicate. In essence, exchanging one bad color for another (though the yellow is less offensive to the eye). DrDre has posted some awesome color corrections that are generally agreed to be the best the posters to this thread have ever seen the colors. When you overlay the original BR with the above correction, you get something far more pleasing and I feel more accurate.

The Bull’s Eyed Womprat said:

So Dre, Is your idea to fine tune your GOUT regrade with these sources? Or are you looking to completely restart the timing process with these as reference?

Yes, I think I will be using these references to fine tune the GOUT regrade. Although I might go for a unifying color grading for all the different sources we have.

I would like to add, that I believe it’s important to preserve the original theatrical print colors (or one of them) without interpretation, or subjective taste. I think once those are preserved, we can have a fair discussion about which colors are intentional, and which might be erroneous or otherwise.

DrDre said:

I would like to add, that I believe it’s important to preserve the original theatrical print colors (or one of them) without interpretation, or subjective taste. I think once those are preserved, we can have a fair discussion about which colors are intentional, and which might be erroneous or otherwise.

I like your idea of basing it on the technicolor prints and correcting the GOUT as a template. I am eagerly awaiting your more accurate scans as I am doing the same thing and I currently have based it on Mike Verta’s samples with some corrections for the flaws he reported. I think this project can really help with the restoration of TESB. I’ve gone under the assumption (not always safe, but I like the results I see from it) that the 1993 Definitive Edition (the GOUT’s source) used the same process for all three films using interpositive prints. So I’ve applied exactly the same settings to all three films and it works very well. Following Poita’s current work on TESB, there are no prints of it that truly preserve the original colors so correcting the ANH GOUT to match the Technicolor frames and applying the results to TESB might be the best color source we have for it.

That’s an interesting approach. I think having good print references will certainly help to understand the processes, that were used to create the various home video releases.