Jump to content

New Improvement to the Bluescreen Process


Recommended Posts

New Improvement to the Bluescreen Process

I have developed a new technique for the Bluescreen / Greenscreen editing process which will avoid problems of cross-contamination outside of the Mattes. This new technique will also make it possible to produce the combined Frame as a 1st generation copy of the original Camera Negative onto Intermediate Positive Film. It would also be possible to have some pure blue colour in the foreground image.

 

Male Matte Using the contact copying process, dim Blue Light (i.e. dim White Light through a Blue Filter) would be used to copy the original (developed) Negative onto Black & White Film. It is necessary that the Light be dim to ensure that there is no penetration of Light through the Yellow Dye Layer formed in the Negative Film Emulsion by the Blue Screen in the original Scene. Since you want every Halide Molecule in the Matte Film Emulsion to be exposed for the foreground, the exposure for this process will be a total over-exposure which will produce a solid 'Male Matte' from the 'Foreground Image'. This process doesn't need a Shutter, and you can expose more than one Frame at a time. You need to use as LOW a Speed of Film as possible to ensure as solid a Matte as possible -- 5 ISO would be preferable. At present, I believe that 80 ISO is the lowest speed available for Black & White Film. If possible, it would be best to have a glass plate in front of the two films to press them firmly together for the exposure. This would avoid cross-contamination along the edges of the Matte. If the Speed of the original Camera Negative is lower than 80 ISO, and the Scene of the Foreground Image is particularly bright, you could choose to use black paint to manually paint over the Matte on the developed Film. If there were any visible blue colour in the Foreground Image, you would just manually paint it over on the developed Matte Film, or you could use the subsequently produced Female Matte overlaid in front of the Negative to produce a second Male Matte made using brighter Light.

[NOTE: If you used a Green Screen in the original Scene, then the Blue Light and Black & White Film would be replaced with Green Light and Kodachrome 64 ISO Transparency Film. This would still be developed as Black & White Film -- not as colour Kodachrome. You could use regular Transparency Film instead of Kodachrome, but I wouldn't recommend it since, unlike Kodachrome, regular Transparency contains Dye Couplers in the Emulsion.]

 

Female Matte You would now place the developed Male Matte Filmstrip in front of the original Negative, and you would use White Light to copy it onto Kodachrome 64 ISO Transparency Film. Once again, this would be a total over-exposure which would produce a solid 'Female Matte' from the 'Bluescreen Background' Image (or Greenscreen). The Kodachrome would be developed as Black & White Film which would have two layers (Red and Green) of solid Black. This process would avoid any concerns about the Female Matte contaminating the Foreground Image.

 

The existing normal process to combine the Foreground and Background Filmstrips uses the Female and Male Mattes to first produce positive Filmstrips which, through the Optical Printer's prism, are combined and recorded on the Final Filmstrip by the Camera. However, I believe that a better process would be to combine the Foreground and Background Negatives separately onto one Interpositive Filmstrip using the contact copy method. First you would copy the 'Foreground / Bluescreen Image' Negative by placing the Female Matte Filmstrip in front of it. Then you would copy the 'Background Image' Negative by placing the Male Matte Filmstrip in front. I believe that you would find this contact method of duplicating / combining to be much higher quality than the standard prism method. This process should be easier than the existing process, and best of all it can be done on any old Movies which are being remastered since there are no required changes in filming technique. You can inform Film Technicians at the Studio Labs of this new technique, and if they wish to contact me with any questions, they can E-Mail me at tlmester@yahoo.ca.

 

It is important that the actual Blue Screen on the Set be solid blue, and also very well illuminated to ensure that it forms a solid Yellow Dye Layer on the Film Negative. You can verify the correctness of the Blue Screen by viewing it through Goggles with a Yellow Filter in front of the Goggle's Lenses. When viewing the Blue Screen through a Yellow Filter, you should see Black or a very pale Yellow.

Link to comment
Share on other sites

  • Replies 50
  • Created
  • Last Reply

Top Posters In This Topic

Hi Terry,

 

The problem I have with this method is I can never get a solid hi-contrast matte just off the blue negative without the help of a Red positive. How are you creating the matte purely off the blue channel? Adding the red positive I can get a solid core without the need for painting or any other tomfoolery.

 

M = (B - 1) + R

 

Using this old formula I get a very good colour difference matte - actually I get the very same matte results as this...

 

M = B - Max (R, G)

 

Cheers,

Will

Link to comment
Share on other sites

The problem I have with this method is I can never get a solid hi-contrast matte just off the blue negative without the help of a Red positive. How are you creating the matte purely off the blue channel? Adding the red positive I can get a solid core without the need for painting or any other tomfoolery.

Hi Will,

The existing method definitely works. I just think it's more complicated than need be. I want to make sure I understand your terminology correctly: the "blue negative" you're referring to is the 'foreground / bluescreen' Camera Negative, and the "red positive" is a red Male Matte of the 'Foreground Image'? If so, I'm simply proposing that the red positive be replaced with the Male Matte.

 

I'm interested to know what Type and Speed of Film you are using to create the Male Matte, Female Matte, and your Red Positive? Also, are your Matte Films being developed using the "reversal" process? With this method I've suggested, there is no reversal development required. As regards painting, I don't think there would be any problems using 80 ISO Matte Film with a 50 ISO Cam. Negative. I've only suggested using paint if someone wanted to be 100% careful. If you used blue in the foreground scene, then that part of the Matte would need to be painted over.

Link to comment
Share on other sites

Hi Terry,

 

I'm not actually using film but replicating the process and the math in Shake.

 

When I refer to the 'blue negative' I'm taking the blue channel information from the RGB image and reversing it. That's the first part of the math: B - 1. I then take the information from the red channel (as a positive) and add it to the result. The idea here is there is little bit's of blue in most things which is going to filter through and so we need a way to difference them out - hence we use the red channel to do this, there should be less red in the blue screen and more in the foreground actor so when we add it to the blue negative it increases the difference between the actor and the bluescreen, which gives us a fairly decent matte.

 

To create the hi-contrast I can either slap a contrast node into the mix or for more exact control an expand node (although doing this optically you wouldn't have the level of control given by an expand node) to clip off some of the values to get a better black and white matte.

 

The M = (B - 1) + R method had been around for awhile. I've never actually tried it as a keyer because M = B - Max (R, G) is much more common.

Link to comment
Share on other sites

I'm not actually using film but replicating the process and the math in Shake.

.....

The M = (B - 1) + R method had been around for awhile. I've never actually tried it as a keyer because M = B - Max (R, G) is much more common.

Hi Will,

I believe you're referring to "Chroma Key" with Video Editing. The new process I've suggested is strictly with Film. Due to the complications of the current Optical Film Bluescreen process, it has been mostly replaced by Video Chroma. My purpose for trying to simplify the Optical process is for cinematographers who don't want to use the Video Chroma process. The Video process is definitely easier, but you lose the resolution and colour quality on the Film.

Link to comment
Share on other sites

Hi Terry,

 

No I'm not referring to a 'Chroma Key' method, I'm referring to a 'colour difference' method of matting which uses the same basic math in both the film and digital worlds, so much so that I grabbed the M = (B - 1) + R direct from a reference book with a section on optical visual effects. I just wrote the equation down based off the very instructive examples I was looking at.

 

A 'chroma key' matting technique works by taking a colour value and finding similar values to create the matte, it doesn't take into account the difference between the FG and the BG. A 'colour difference' matte takes into account the difference (hence the name) and as such I don't see how you can pull the needed matte based purely off one colour channel - which is what your trying to do. Until you get passed this problem your never going to get decent quality composites even if you solve the generation loss that occurs in optical compositing.

 

Also it is very rare in VFX to use the 'chroma key' method for creating mattes.

 

Please could you provide evidence or examples?

 

Cheers,

Will

Link to comment
Share on other sites

Thanks Will for the Weblink example. Yes, this is a 'Digital Version' of the 'old Optical Process'. The old Optical Process is rather cumbersome which is my thinking behind my new idea. Regarding the example, I believe that the light hitting the Greenscreen was too bright which results in White Light (containing R,G&B) being reflected off of the Screen. (As I noted above respecting a Bluescreen, if you look at it through a Yellow Filter you should see Black or a very dim Yellow. If you see too much Yellow, then the Bluescreen is too brightly lit!) Also, the Camera recording that image probably used the "Bayer CCD" which causes further problems you don't get with Film. If you look closely at the final composite (#7), you can see green along the edges of the man's outline -- "hair" is an impossible image to perfectly separate from the background green colour. (I personally would recommend a blue screen rather than green.)

A 'chroma key' matting technique works by taking a colour value and finding similar values to create the matte, it doesn't take into account the difference between the FG and the BG. A 'colour difference' matte takes into account the difference (hence the name) and as such I don't see how you can pull the needed matte based purely off one colour channel - which is what your trying to do. Until you get passed this problem your never going to get decent quality composites even if you solve the generation loss that occurs in optical compositing. ... Also it is very rare in VFX to use the 'chroma key' method for creating mattes.

I'm interested to know, why is it not possible with the Digital Process to separate out from the picture the "pure green" Pixels (i.e. Pixels having almost no red or blue overlaid), or is this the 'Chroma Key' Process? If you could separate the "pure green" Pixels -- which I guess is the Chroma Process -- you would have your Female Matte, and then you automatically have your Male Matte. What do you personally think of the Chroma Process, and why is it not used?

 

Regarding my idea, you need to remember that using 'real Light' with Film is different from an electronic Digital Process. Under my idea only Blue Light and over-exposure is used to create the Male Matte on Black & White Film. Using B&W Film avoids some of the problems you'll get with a Digital Process. Once you have the Male Matte, producing the Female Matte is very easy since there would be no colour separation involved.

Link to comment
Share on other sites

[/size]

I'm interested to know, why is it not possible with the Digital Process to separate out from the picture the "pure green" Pixels (i.e. Pixels having almost no red or blue overlaid), or is this the 'Chroma Key' Process?

 

If the greenscreen had no red or blue values in it then the initial matte would have a lot more contrast between the FG and BG, as well as having much better range in the grey areas (hair, edges).

 

What do you personally think of the Chroma Process, and why is it not used?

 

It's too indiscreet and it's binary, it takes a value and then looks for a range of similar values and then marks them as transparent. This isn't good for dealing with edges, hair, shadows, glass, etc.

 

Regarding my idea, you need to remember that using 'real Light' with Film is different from an electronic Digital Process. Under my idea only Blue Light and over-exposure is used to create the Male Matte on Black & White Film.

 

Yes, it is different. Working with optics and film introduces a lot more variables into the process then I would care to replicate. Please post some examples or evidence to show that your way will work. As it stands your proposed method is well intention, but you've yet to provide any evidence to suggest it actually works.

 

If you look closely at the final composite (#7), you can see green along the edges of the man's outline -- "hair" is an impossible image to perfectly separate from the background green colour. (I personally would recommend a blue screen rather than green.)

 

I think I mentioned that the composite still needs some fine-tuning. It was put together within half-an-hour, and a pretty good matte for something that people normally pay lots of dollars for specialised plugins to do (granted they also do other things that haven't been included in my shake script). And as regards to hair, it doesn't matter what type of colour screen you use, the colour is still going to mix in with the hair. The type of screen only reflects what type of colours you may have in the FG, blue jeans don't work too well on bluescreen stages - unless you like rotoscoping.

 

There is a reason why compositing switched from being an optical process to a digital one. I like your spirit in trying to improve the optical process, but I don't think you'll ever get the results you would with a digital composite.

Link to comment
Share on other sites

  • Premium Member

Terry, you need to post frames of your final composites using this technique, and show how well it handles different edges of objects.

 

Also, optical compositing introduces generational loss, and going to an IP doesn't help because you still need to make a dupe negative from that IP for printmaking.

 

Most professional optical printer supervisors were more than happy to switch to digital compositing because it both looks better and it is more flexible.

Link to comment
Share on other sites

Hi Will, The following is the very vague description I've found for the Bluescreen Optical process. I'm not completely certain about its use of terminology. It mentions a "blue filter" which I assume means blue to filter out Red and Green Light, but perhaps they mean a Yellow Filter to filter out Blue Light. It mentions "red and green filters" which I assume are to filter out Blue Light -- in which case a Yellow Filter would be better, but are they actually talking about Cyan and Magenta Filters to filter out Red and Green Light. In the process described below, the Mattes would have to be created from the Inter-Positive instead of the Negative.

 

"The bluescreen shot was first rephotographed through a blue filter so that only the background is exposed. A special film is used that creates a black and white negative image ? a black background with a subject-shaped hole in the middle. This is called a 'female matte'.

The bluescreen shot was then rephotographed, this time through a red and green filter so that only the foreground image was cast on film, creating a black silhouette on an unexposed (clear) background. This is called a 'male matte'.

The background image is then rephotographed through the male matte, and the bluescreen shot rephotographed through the female matte. An optical printer with two projectors, a film camera and a 'beam splitter' combines the images together one frame at a time. This part of the process must be very carefully controlled to ensure the absence of 'black lines'. During the 1980s, minicomputers were used to control the optical printer. For The Empire Strikes Back, Richard Edlund created a 'quad optical printer' that accelerated the process considerably and saved money. He received a special Academy Award for his innovation."

All I have suggested with my idea is to create the Mattes directly from the Negative. Since the FG will contain Blue Light, I have simply suggested placing the Male Matte in front of (or behind) the Negative when creating the Female Matte. This would prevent any exposure of the FG on the Female Matte Film. Another option to create the Female Matte would be to use Blue Light to expose the FG, and then 'reverse develop' the Matte Film which would produce the Female Matte from the BG. This should be an improvement over using Red & Green Light.

Terry, you need to post frames of your final composites using this technique, and show how well it handles different edges of objects.

Also, optical compositing introduces generational loss, and going to an IP doesn't help because you still need to make a dupe negative from that IP for printmaking.

Most professional optical printer supervisors were more than happy to switch to digital compositing because it both looks better and it is more flexible.

I completely agree with you David that the Digital Matte process is considerably easier than the old Optical Process, and that's why Digital has taken over. The prism method of combining FG & BG only makes the Optical Process worse. I know that there is optical generational loss. That's why I propose this technique of making the Mattes from the original Negative. There's really no difference between what I propose and the old process other than that I propose using the Negative to avoid adding any extra generations, and I propose using the Male Matte to produce the Female Matte. Although, as I note above, by using 'reversal developing' it wouldn't even be necessary to use the Male Matte for making the Female. I'm afraid that I don't have Darkroom facilities at my disposal, and so I can't readily undertake such delicate tests. No tests are needed to know in advance that a 1st Generation copy will be higher quality than 2nd Generation and so on. Avoiding extra generational copies is a main purpose behind my proposal. There of course is inherent optical degradation with the standard contact copying process, but that's a separate issue from making Bluescreen Mattes.

Link to comment
Share on other sites

  • Premium Member

No one is going to take a proposal seriously if you haven't even tried it yourself yet. Lining up matte elements in an optical printer and eliminating lines and fringing is a bigger problem than the generational loss issue.

 

You need to complete a composite using the traditional optical printer technique, your modified one, and a digital composite, and show comparisons, frame enlargements, etc. to prove any sort of superiority to your proposed technique.

Link to comment
Share on other sites

The 'Colour Difference' method used to create mattes in the optical days of VFX explained, hopefully this makes the process easier for people out there to understand...

 

http://www.earlyworm.co.nz/film/colour_dif...nce_matting.htm

 

This is not the color difference process, but rather the older color seperation processs.

The color difference was devised by Petro Vlahos and first used in 'Ben Hur'.

In the color difference a synthetic blue seperation is made by combining the green seperation and a difference matte showing the difference between the blue and green

 

http://www.google.com/patents?vid=USPAT3158477

Link to comment
Share on other sites

 

seperations. with the synthetic green seperation the blue backing becomes black, or almost.

Thus a cover matte isn't needed on the foreground, or a thin one is used.

This allows the use of transparent objects and fine edges.

 

Vlahos's Ultimatte keyer uses the same principal.

 

---El Pedante

Link to comment
Share on other sites

seperations. with the synthetic green seperation the blue backing becomes black, or almost.

Thus a cover matte isn't needed on the foreground, or a thin one is used.

This allows the use of transparent objects and fine edges.

 

Vlahos's Ultimatte keyer uses the same principal.

 

Thanks for picking me up on this Leo,

 

On further reading I did notice that there was a difference in methodology used between the colour seperation and colour difference progress - the reason why I originally referred to it wrongly was because it was giving me the exact same results as what a modern colour difference keyer does M = B - Max (R,G). In both cases their finding the difference between the FG and BG to create a matte.

 

From what I can gather from the patent Vlahos was originally using M = B - G to create his mattes, he's obviously updated the formula in the advent of digital compositing.

 

Cheers,

Will

Link to comment
Share on other sites

New Improvement to the Bluescreen Process

I have developed a new technique for the Bluescreen / Greenscreen editing process which will avoid problems of cross-contamination outside of the Mattes. This new technique will also make it possible to produce the combined Frame as a 1st generation copy of the original Camera Negative onto Intermediate Positive Film. It would also be possible to have some pure blue colour in the foreground image.

 

Male Matte Using the contact copying process, dim Blue Light (i.e. dim White Light through a Blue Filter) would be used to copy the original (developed) Negative onto Black & White Film. It is necessary that the Light be dim to ensure that there is no penetration of Light through the Yellow Dye Layer formed in the Negative Film Emulsion by the Blue Screen in the original Scene. Since you want every Halide Molecule in the Matte Film Emulsion to be exposed for the foreground, the exposure for this process will be a total over-exposure which will produce a solid 'Male Matte' from the 'Foreground Image'. This process doesn't need a Shutter, and you can expose more than one Frame at a time. You need to use as LOW a Speed of Film as possible to ensure as solid a Matte as possible -- 5 ISO would be preferable. At present, I believe that 80 ISO is the lowest speed available for Black & White Film. If possible, it would be best to have a glass plate in front of the two films to press them firmly together for the exposure. This would avoid cross-contamination along the edges of the Matte. If the Speed of the original Camera Negative is lower than 80 ISO, and the Scene of the Foreground Image is particularly bright, you could choose to use black paint to manually paint over the Matte on the developed Film. If there were any visible blue colour in the Foreground Image, you would just manually paint it over on the developed Matte Film, or you could use the subsequently produced Female Matte overlaid in front of the Negative to produce a second Male Matte made using brighter Light.

[NOTE: If you used a Green Screen in the original Scene, then the Blue Light and Black & White Film would be replaced with Green Light and Kodachrome 64 ISO Transparency Film. This would still be developed as Black & White Film -- not as colour Kodachrome. You could use regular Transparency Film instead of Kodachrome, but I wouldn't recommend it since, unlike Kodachrome, regular Transparency contains Dye Couplers in the Emulsion.]

 

Female Matte You would now place the developed Male Matte Filmstrip in front of the original Negative, and you would use White Light to copy it onto Kodachrome 64 ISO Transparency Film. Once again, this would be a total over-exposure which would produce a solid 'Female Matte' from the 'Bluescreen Background' Image (or Greenscreen). The Kodachrome would be developed as Black & White Film which would have two layers (Red and Green) of solid Black. This process would avoid any concerns about the Female Matte contaminating the Foreground Image.

 

The existing normal process to combine the Foreground and Background Filmstrips uses the Female and Male Mattes to first produce positive Filmstrips which, through the Optical Printer's prism, are combined and recorded on the Final Filmstrip by the Camera. However, I believe that a better process would be to combine the Foreground and Background Negatives separately onto one Interpositive Filmstrip using the contact copy method. First you would copy the 'Foreground / Bluescreen Image' Negative by placing the Female Matte Filmstrip in front of it. Then you would copy the 'Background Image' Negative by placing the Male Matte Filmstrip in front. I believe that you would find this contact method of duplicating / combining to be much higher quality than the standard prism method. This process should be easier than the existing process, and best of all it can be done on any old Movies which are being remastered since there are no required changes in filming technique. You can inform Film Technicians at the Studio Labs of this new technique, and if they wish to contact me with any questions, they can E-Mail me at tlmester@yahoo.ca.

 

It is important that the actual Blue Screen on the Set be solid blue, and also very well illuminated to ensure that it forms a solid Yellow Dye Layer on the Film Negative. You can verify the correctness of the Blue Screen by viewing it through Goggles with a Yellow Filter in front of the Goggle's Lenses. When viewing the Blue Screen through a Yellow Filter, you should see Black or a very pale Yellow.

 

What you described here sounds like the old Disney Sodium Vapor matte process which was great in its day. Just look at Mary Poppins--super matte work there. Used a yellow screen and sodium vapor lights lighting the screen to give somewhat the improvements you're talking about here. It was interesting because they only produced one camera to use this process. The prism inside could never be replicated.

 

I really don't understand why someone would use any of these methods now though with the digital tools available.

 

http://www.answers.com/topic/sodium-vapor-process

Edited by Richard Andrewski
Link to comment
Share on other sites

By the way, it looks like Vlahos came up with the sodium vapor process also for disney. Here's an excerpt of a paper I found online. It was from a paper on bluescreen techniques by some guys from Apogee which was the split off group headed by John Dykstra when George Lucas moved Industrial Light and Magic to San Francisco:

 

"Petro Vlahos' sodium vapor system is felt to be the best. It results in a matte image of the same size as the foreground subject image, whereas the ultraviolet and infrared mattes are distorted in size by virtue of their respective positions at opposite ends of the visible spectrum. Because the matte image is formed on a separate piece of film, there is no need to break the color negative down into separations, nor to lose any portion of the spectrum in order to gain a matte. There is also no problem of fringing, and glass objects as well as fine details like hair can be faithfully matted."

 

Obtained here:

 

http://www.digitalgreenscreen.com/smpterbs.html

Link to comment
Share on other sites

Before anyone picks up on my terrible math. I forgot to add an absolute function, also I'm basing my calculations on a normalized 0..1 range. So here's the corrected formula for 'colour seperation' mattes...

 

M = abs (B-1) + R

Link to comment
Share on other sites

What you described here sounds like the old Disney Sodium Vapor matte process which was great in its day. Just look at Mary Poppins--super matte work there. Used a yellow screen and sodium vapor lights lighting the screen to give somewhat the improvements you're talking about here. It was interesting because they only produced one camera to use this process. The prism inside could never be replicated.

 

I really don't understand why someone would use any of these methods now though with the digital tools available.

 

http://www.answers.com/topic/sodium-vapor-process

 

No. This is nothing like the sodium vapor process. The prism that Mr.Mester mentions is not in the camera, but in an optical printer and is used for combining images in a single pass instead of two passes.

Since the foreground and background are in seperate heads, they image size and position can be adjusted for a better fit.

 

The prism in the camera for sodium vapor seperates the foreground and the sodium vapor lit backing onto seperate films. Color neg for the Fg, PXN for the backing. Disney used a Technicolor Three-strip camera.

Link to comment
Share on other sites

Sorry I wasn't clear. I was just talking about the idea of reducing passes not the actual technical method to achieve it. The disney process apparently got matte plate and live action at the same time out of the camera without having to create other passes. At least they don't mention having other passes. Optical printer was still required later though for combining elements as it would be for all of these types of matte work. I'm not aware of any way to cut out optical printer without using digital techniques.

 

Yes I can imagine the stage was quite hot too although sodium vapor (today at least--not sure about early 60's version) is a more efficient at watts per lumen than tungsten.

Link to comment
Share on other sites

How would your process handle a nun in a white habit riding a white Vespa,

or a mostly white jetliner?

There is an important reason for combing a blue negative with a red positive for making the matte.

Hi Leo,

I'm not certain what problem you're specifically thinking of? Are you thinking of Blue Spill from the Bluescreen reflecting off of the Nun's or Vespa's white colour? Such reflection is obviously a big problem. It should be able to be solved on Set by placing the Bluescreen further back from the Subject or by using a darker shade of blue for the Screen. If you look at the Subject on Set through a Blue Filter, you will be able to tell if there's any Blue Spill. As far as my idea for making the Mattes, Blue Spill would have minute affect because of the use of over-exposure of the Matte Film -- an option not readily available with the Digital process. Since the Bluescreen will form a solid Yellow Layer (Filter) on the original Negative, and as long as the Blue Light used to produce the Matte Film is not too bright, it won't penetrate that Yellow Filter. It will therefore only over-expose the FG (Male) image on the Matte B&W Film, and through 'reversal developing' of B&W Film this same process can also be used to make the Female Matte. If you had an area in the FG image that was high in Blue colour content, then I think it would be easier to just paint over that area of the Male Matte (using a magnifying glass to see of course) instead of going through the complicated colour separation process. If you have a case of high Blue colour in the FG, you can also use the Male Matte in front to produce the Female Matte on Transparency Film instead of B&W.

 

I believe some of the other problems with Bluescreen can be easily solved. I don't know what Speed of Film would normally be used in Camera to film the Bluescreen / Greenscreen Scenes, but I absolutely would advise using the lowest available -- 50 ISO. If you use a higher Speed Film, when making the Mattes you are definitely going to suffer penetration of your chosen colour Light (Blue or Green) through the Yellow Filter it produces on the Camera Neg. Some of the problems of fringing and black lines could very easily be caused by too high a Speed of Film. Jonathan Erland's article mentions that fringing and lines are contributed to by Mattes that are 2nd or lower generation copies. I believe that another problem was that the prism Optical Printer uses two separate Film Gates (in two Projectors), and you can be certain that the dimensions of those two Gates are not 100% alike. This would result in the two projected images not being in perfect alignment which means fringing or black lines. Another problem would be the speed of the Optical Printer which I assume would be at 24 f/s. The faster the Film is going through the Printer, the greater the propensity of the Film slightly bending, and this would result in fringing or black lines. Due to the degradation of Light caused when it goes through a prism, I suggest using the contact copying method which uses only one Film Gate. As I had mentioned in my other Thread back in January, it would be better with contact copying to have the films pressed together with a glass plate in front. With this improvement to the contact copying method, and using 1st Generation Mattes, the problem of fringing and black lines should become undetectable in the realtime Movie. Arthur Widmer developed the Bluescreen technique in the early 1950s, and with it began the effort to make special effects look completely real. Just look at Cecil B. DeMille's Ten Commandments. Digital processes may be easier, but there's one big problem with them: they don't look real. Just look at the latest 'Superman' compared with the original Movies, or look at this new '300' Movie. The CGI are so obviously not real. Hollywood has now gone back in time to the pre-Bluescreen era when the attitude was that the audience should accept non-reality since it's just a Movie. I personally reject this mentality, but this doesn't mean that old Optical methods cannot be improved upon. You need to think outside the box.

 

I want to note two specific problems with the Disney Matting technique. Firstly, the camera prism degrades the main image being filmed. Light Rays are busted up into a million pieces when they travel through glass -- especially angled glass, and you don't want to have anything between the Subject and the Film except for the Camera Lens. The second problem is that the Matte Film used a separate Film Gate, and this means that there will not be perfect alignment between the image on the Camera Negative and the Matte Negative. You could also see the alignment problems with the tri-neg. Technicolor Movies. Since the main Bluescreen process clearly worked in other movies, this setup wouldn't be worth the complications.

Link to comment
Share on other sites

  • Premium Member

You're confusing the issue of CGI (Computer Generated Imagery) with digital compositing.

 

No serious efx person thinks that there are any superior results to be gained from compositing in an optical printer over digital compositing. Optical printer travelling matte composites nearly always stuck out from the surrounding non-duped footage, unlike many digital composites today that go by in a movie without the viewer being aware of it, unless it is something clearly that has to be an effect. You can scan and record back to film an completed composite with the same contrast and grain structure of the original film elements, unlike with an optical printer composite which will never be the same generation as the surrounding non-duped elements.

 

This proposal of yours will go nowhere because no one wants to go back to the days of dealing with multiple hold-out mattes, wedge testing, etc. And not all composites only have a foreground / background element to sandwich together. Chromakey composite work using an optical printer is a dead, period.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Forum Sponsors

Broadcast Solutions Inc

CINELEASE

CineLab

Metropolis Post

New Pro Video - New and Used Equipment

Gamma Ray Digital Inc

Film Gears

Visual Products

BOKEH RENTALS

Cinematography Books and Gear



×
×
  • Create New...