New Improvement to the Bluescreen Process

[Terry Mester]

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. ...

David, I never stated that prism-type Optical Printed Mattes are superior to Digital Printed Mattes. In my first post I advocated using Contact Printing. You are talking about the problems with the existing Optical Printer process, and I'm aware of and completely agree with you regarding these problems.

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.

Are you referring here to digitally printing the Mattes alone, or are you referring to digitally printing the combined Frame with FG & BG? If a digitally printed Matte works for Contact Printing, then excellent! It's a case of 'if the shoe fits, wear it'. However, when digitally printing the combined Frame, you're losing the original optical resolution and colour.

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.

Once again, I'm not suggesting going back to the old Optical Printer technique. However, as for the Optical Printer being "dead", what about amateur cinematographers (such as 16mm) who don't have access to an expensive Digital Printer? What do you say to them? A simplified matting technique will also help 16mm cinematographers.

[David Mullen ASC]

However, as for the Optical Printer being "dead", what about amateur cinematographers (such as 16mm) who don't have access to an expensive Digital Printer? What do you say to them? A simplified matting technique will also help 16mm cinematographers.

 

A lot more people can do composites on their home computers than own optical printers -- digital technology has brought compositing to the masses in a way that optical printing work could never do! Your average high school amatuer filmmaker is home doing chromakey composites routinely, whereas when I was in film school and actually took a 16mm optical printer class, we didn't even mess with chromakey stuff due to all the separation and matte work needed. Now chromakey composites are done every day by kids with laptops.

 

I assume you spend your time dreaming up these proposals because you think they have practical value, not as some sort of intellectual exercise. The truth is that very few independent filmmakers ever bothered to learn to do special effects in an optical printer -- they hired an optical house to put together those shots, due to the level of expertise and type of equipment needed. So if back in the past when digital compositing was not an option, most people did not take the time & effort to do their own compositing, why now that digital compositing on a home computer is an option, would more people today will suddenly want to non-digital composite work? I'm just trying to tell you that there is no market for your idea, which, by the way, I feel would lead to inferior composite work to boot. Compositing is much more than putting two image together with hold-out mattes, you need to able to control color & density of your color elements, density of the b&w mattes, ability to adjust the size and position of the matte at the micron level, etc. which is why it's a job for trained professionals. Sure, it could be done poorly by amatuers for fun, assuming an amatuer even had access to the printer and lab work needed, but then you're hardly getting good results compared to digital methods.

 

And yes, you can match color to a film element in a digital film-out, perhaps better than using a photochemical duplication method. Contrast build-up is a major problem of optical printer composites and you never get back the same contrast and color saturation as the original non-duped elements, so a lot of things have to be done to get around this, from how you photograph the elements to how you composite them. It's an artform as much as a science.

 

Proposing new non-digital compositing techniques today is like trying to sell people an improved typewriter to replace their home computer word processors.

[Leo Anthony Vale]

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.

 

White consists of blue, green and red.

So on a blue seperation blue and white are the same. If you view the negative through a yellow filter the blue backing appears black. But white also appears black.

So what you are printing is not a matte but a blue seperation. Both blue and white will be clear.

 

On a red seperation positive the blue is black, while the while the white is still clear. This is reversed on the

red seperation negative. So to get a matte, the blue pos has to be bipacked with the red neg.

 

But this is all moot. As David points out, the optically printing of travelling mattes has been displaced by digital compositing.

 

It's like knowing how Cinecolor prints were made. Fascinating but unpractical.

 

As mentioned elsewhere in the VFX section, JK printers are being discontinued. So it seems there's going to be even less opticalprinting don at home.

Edited March 17, 2007 by Leo Anthony Vale

[Stephen Williams]

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.

 

Hi,

 

Optical printers typically run at 3 or 4 frames per second. Remember they have to start & stop instantly.

 

Stephen

[Terry Mester]

White consists of blue, green and red. So on a blue seperation blue and white are the same. If you view the negative through a yellow filter the blue backing appears black. But white also appears black. So what you are printing is not a matte but a blue seperation. Both blue and white will be clear.

On a red seperation positive the blue is black, while the while the white is still clear. This is reversed on the

red seperation negative. So to get a matte, the blue pos has to be bipacked with the red neg.

 

But this is all moot. As David points out, the optically printing of travelling mattes has been displaced by digital compositing.

Hi Leo,

This is completely correct regarding 'normal exposure'. What I have suggested is 'total over-exposure' when producing the B&W Matte Negative (from the Camera Colour Neg.) which will turn "non-blue" into "black" and "semi-blue" into "essentially black" on a B&W Negative. The 'Matte Negative' is what would be used as the Matte. It's like shooting a light source with B&W Film. The B&W Negative would be totally black, and the Positive would be clear. With my idea it's important that the Blue Light (during exposure of the Matte Neg.) not be too bright so that it doesn't penetrate the Yellow Dye Layer formed on the Camera Negative by the Bluescreen on the Set. The Bluescreen forms an area on the Camera Neg. that filters only Blue Light, and therefore using Blue Light with 'normal' and 'reversal' developing (and over-exposure) you can produce the Male and Female Mattes.

Digital Composite Mattes have definitely taken over. However, it certainly would be interesting to know how Mattes from this new technique would compare to Digital Mattes, and if there were a discernible difference in quality between the two. Also, don't forget 65mm Film. Are there quality Digital Printers available to produce quality 65mm Digital Mattes?

Hi,

Optical printers typically run at 3 or 4 frames per second. Remember they have to start & stop instantly.

Thanks Stephen for this info. This would be a safe speed. The big problem with Optical Printers is the prism which busts up the Light Rays. The Optical Printer was developed in the early 1930s for B&W Film, and is fine for B&W. Colour Light of course is much more complicated. Even current contact copying techniques don't maintain full colour quality.

 

By the way David, Optical Printers need to be kept by Labs for use in Black & White Movies. A Digital Printer cannot even remotely provide a resolution comparable to B&W Film.

[David Mullen ASC]

Also, don't forget 65mm Film. Are there quality Digital Printers available to produce quality 65mm Digital Mattes?

By the way David, Optical Printers need to be kept by Labs for use in Black & White Movies. A Digital Printer cannot even remotely provide a resolution comparable to B&W Film.

 

There is plenty of digital restoration work being done for b&w movies, and Technicolor has recently been digitally creating b&w separation masters for color digital intermediate work (i.e. recording red, green, and blue separation masters created from the digital master out to 35mm b&w fine-grains using a laser recorder).

 

You need to get your terminology correct. First of all, you use "film recorders" to transfer digital files to film, and you don't use film recorders to do compositing work -- that's done earlier using computers with compositing software. The finished composite is recorded out to film using the film recorder.

 

Considering that many IMAX movies now use film recorders, the notion that a film recorder can't handle the resolution of 4-perf 35mm b&w is nonsense considering that they can handle 15-perf 65mm!

 

The main reason that most b&w restoration work is still done using optical techniques is mainly that it's faster and cheaper, not that it's better, and you don't have issues like restoring faded color dye layers to deal with. But "Dr. Strangelove" just went through an expensive 4K digital restoration effort.

http://blog.digitalcontentproducer.com/bri...ct-for-cineric/

 

You're under some mistaken notion that digital post work can't be done at the full resolution that film contains. This is nonsense -- the only limiting factor is time and money. Many people choose to work at half the resolution (2K) for digital post work as a cost saver, but 4k-to-6K, which is the range necessary to capture 35mm film's resolution, is becoming more affordable and thus will become more commonplace.

[Terry Mester]

There is plenty of digital restoration work being done for b&w movies, and Technicolor has recently been digitally creating b&w separation masters for color digital intermediate work (i.e. recording red, green, and blue separation masters created from the digital master out to 35mm b&w fine-grains using a laser recorder).

When you talk about restoration work on old B&W Movies, are they using the original Camera Negative, Inter-Positive, Inter-Negative, or the final (3rd Generation) Positive? Chances are, they're using a final 3rd Gen. Positive, and the resolution quality will be much lower than the Cam. Negative. The Blue Light (linear) Resolution for a 22mm space would range from 5238 Light Rays to 6136 Rays across the 22mm Width of the original Cam. Neg. Frame. This is higher than 4K which is 4096 linear Pixels, and Light is "random".

You need to get your terminology correct. First of all, you use "film recorders" to transfer digital files to film, ... The finished composite is recorded out to film using the film recorder.

Thank you for correcting my terminology. I don't want to get into semantics on this Thread, but, whoever chose to name as "Film Recorder" the device for 'printing' Digital Video Files onto Film, deserves a scolding. With every other device I've heard of, a "recorder" inputs something into the device, and to output from the device is called a "printer" for printing or a "speaker" for speaking or an "amplifier" for amplifying, etc. This device should be called a "Digital Printer" or "Digital Film Printer" in line with the properly named "Optical Printer".

Considering that many IMAX movies now use film recorders, the notion that a film recorder can't handle the resolution of 4-perf 35mm b&w is nonsense considering that they can handle 15-perf 65mm!

The digitally printed IMAX Movies are not even remotely meeting the full optical resolution potential of the IMAX Film Frame -- period. One of the benefits of Film is that it doesn't have to be filled full of light -- including digitally printed Light Pixels. You can digitally print onto IMAX format at any low resolution, and it will work. I noted above the resolution potential of the B&W Camera Negative.

The main reason that most b&w restoration work is still done using optical techniques is mainly that it's faster and cheaper, not that it's better, and you don't have issues like restoring faded color dye layers to deal with. ... You're under some mistaken notion that digital post work can't be done at the full resolution that film contains. This is nonsense -- the only limiting factor is time and money. Many people choose to work at half the resolution (2K) for digital post work as a cost saver, but 4k-to-6K, which is the range necessary to capture 35mm film's resolution, is becoming more affordable and thus will become more commonplace.

I agree completely that the Optical Printer is not the best method for copying Film. Contact copying is obviously better. I'll report back as soon as I've had a professional Lab test out my idea for improving contact copying. (I've been reading that the Contact Printer is run at speeds faster than 2000 feet per minute! This is not exactly what you would call quality printing work. You can't blame Film for poor printing practices.) The 4K Digital Film Recorder just covers the maximum potential of 3965 Red Light Rays that can fit into a 22mm space (Regular 35mm). However, 4K (per 22mm) doesn't meet the resolution potential of Green and Blue Light. Even a 6K Film Recorder wouldn't meet the Red Light resolution potential of 65mm. Additionally, as noted above the digital film print is "linear" -- not "random" like real Light! Furthermore, what is the colour content of the Digital Film Recorder: 24-Bit? 24-Bit colour only amounts to a meager 256 Shades per Primary Colour. If you're filming a Movie with 500 ISO Film, then Digital Colour content may appear adequate. However, if you're using 50 ISO Film, how does Digital Colour look by comparison? I wish 25 ISO were available -- how would Digital Colour compare to that?

 

As a matter of interest, what is the monetary cost of these Digital Film Recorders compared to a Continuous Contact Printer? What are their upkeep / maintenance costs?

Edited March 19, 2007 by Terry Mester

[David Mullen ASC]

As a matter of interest, what is the monetary cost of these Digital Film Recorders compared to a Continuous Contact Printer? What are their upkeep / maintenance costs?

 

They don't even do the same things. Film recorders transfer digital data files to film, usually to a 35mm internegative. You still need to make contact prints off of that. Since it's not a choice between a film recorder versus a contact printer, a comparison doesn't make sense.

 

Besides, the average person isn't going to own either. Post houses own film recorders and labs own contact printers (although some post houses own labs and some labs own post houses). And film recorders require the existence of labs to process the output negative and then make prints off of the negative.

 

You also have to factor in that a film recorder sits in the corner of a room, is the size of a xerox machine, and is operated by a single person -- and a contact printer takes up a large room, needs plumbing, water, chemistry, waste disposal, etc. so they aren't even comparable devices! It's like asking me to make a comparison between my refrigerator and my car -- what's the point? I need both, they do separate things, etc.

 

I think your confusion comes from your use of the word "digital printer" as if now movie prints are rountinely being made by some sort of digital printer, like at a still photo lab.

 

Sure, a film recorder can do that, record a positive image to print stock, but considering it would take a couple of days to make a feature-length print and it would cost at least $50,000 per print, obviously film recorders are not normally being used to make release prints, just the negatives for printing from. A contact printer can make a feature-length print in a few minutes for about $1000, so clearly there's no comparison. The film recorder is not designed to replace the contact printer.

 

You seem to be living in some other world where they are hard boundaries between film and digital post technology, when we have been living in a hybrid of both worlds for a decade now.

 

I suspect you think there is some way to turn back the clock and return to a time of more film-based technologies only for post-production, but it simply isn't going to happen. For one thing, young people entering filmmaking these days are almost more heavily engaged in digital technology than the industry they hope to enter. Look at the move towards digital intermediates for color-correction of movies. That's almost entirely been pushed by filmmakers who want to be able to use the same digital color-correction tools they have been using for video-finished productions (for decades) for theatrically-released productions, rather than be limited by traditional photochemical RGB color printer timing controls. And they want to do chromakey composites digitally too and preview the effect on a computer or digitally project it, be able to make changes, before recording it out to film.

 

Motion picture processing and printing (contact or optical) is not something that amatuers or independent filmmakers normally do themselves -- they use labs, they use optical houses, etc. Any new method of doing composite work photochemically would only matter to the people who are still doing photochemical composites, the few small efx houses that have not retired their optical printer. And few of them have usually have contact printers too -- they go to a lab for that service. They may try and use their optical printer as a contact printer using bipack methods for certain types of effects. Restoration companies may have a range of printers and even their own processors for doing photochemical restoration, but usually they wouldn't be involved in creating composites for new efx shots.

[Patrick McGowan]

You're under some mistaken notion that digital post work can't be done at the full resolution that film contains. This is nonsense -- the only limiting factor is time and money. Many people choose to work at half the resolution (2K) for digital post work as a cost saver, but 4k-to-6K, which is the range necessary to capture 35mm film's resolution, is becoming more affordable and thus will become more commonplace.

 

I am still trying to wrap my head around the endless aspects of digital post production. I guess I understand that 4k to 6k can capture film's resolution accurately, but what about the digital colorspace? I mean isn't that something that suffers when these digital files are scanned back to film?

[Will Earl]

Furthermore, what is the colour content of the Digital Film Recorder: 24-Bit? 24-Bit colour only amounts to a meager 256 Shades per Primary Colour.

 

Okay, scanning film into a digital form is a huge and fairly technical topic, so I'm bound to get a detail wrong or simplify it too much. Doing a search may help, keywords include... cineon, dpx, log, kodak.

 

Film is scanned into the computer as 10bit Cineon files, these files store the information in a logarithmic form. There are three zones in a log file, the 'toe' region defines the values of the negative which are below 2% black ref, this area is where the film grains barely respond to light. Next we have the straight line part of the s-curve, which represents most of the information for a properly exposed shot. And then we reach the shoulder of the curve which represents all values above 90% white ref, such as bright highlights. The 'toe' and the 'shoulder' of the curve are reasonably flat and are assigned less values in the cineon file because there is very little preceivable information in those areas of the negative, in the 'toe' we may see some grain movement in the shadows and in the 'shoulder' we have a nice roll-off on our highlights.

[David Mullen ASC]

I don't know what he means by "256 shades per color in 24-bit" but 10-bit LOG RGB is something over 1000 luminence values per channel. I don't know what he means by "shades". You get other colors by mixing red, green, and blue -- not within red, green, and blue individually. As far as I know, 24-bit per color is more than 256 luminence "steps" in each color.

 

As for any loss, there is always some loss in every process. There is loss in high-speed continuous contact printing -- you don't view a negative, you view a print of the negative.

 

And since most release prints are made from a dupe negative, made from an IP, made from the original negative, there is even more loss.

 

So for some films that have gone through a D.I., they spend the money to digitally record multiple "original" ESTAR negatives for printing, so every release print is only one generation removed from the original. So any "digital" loss in that scenario is smaller compared to the loss of resolution and color from using an IP/IN step.

 

In terms of "richness" of color, I can tell you that the prints for "Astronaut Farmer" are more saturated and contrasty than if I had printed the original Fuji Eterna photography, because we were able to add that look digitally. So shots made in flat, overcast weather could be boosted to look more like the shots made in sunny weather.

 

I'm not saying there are no problems or quality hits with digital scanning and recording, because there are. There is the primary problem of the fact that most of the work is done at 2K when it should be done at 4K. And there is the creative problem, i.e. given digital color-correction tools that allow you to make changes within the frame (the most radically being in "Astronaut Farmer" that we turned green trees into yellow trees to match the ones that turned yellow in the Fall, sometimes we had to turn yellow trees back into green trees) is that you can generate digital artifacts that are not film-like. But given all the powerful tools of digital color-correction, it's hard to not use them creatively or to use them to fix and match shots, even improve missed focus, that normally would not be fixable in normal printing.

 

Personally I prefer the look of straight contact printing of negative, but the trouble is that a certain percentage of shots in a movie often are not original negative; they may have been duped for optical effects like dissolves (studios often won't allow A-B roll-cut negatives but require a single-strand negative) or contain visual effects. And you may have shot in a format that doesn't allow contact printing, like Super-16 or 2-perf 35mm or 3-perf / 4-perf Super-35, so you're weighing an optical printer conversion through dupe elements to a digital conversion.

 

And if you choose to use an optical printer to convert a Super-35 movie to a projection format, now the entire movie has gone through an IP/IN generation, and any duped shots in the movie made using an optical printer, like for dissolves, already went through an IP/IN generation and thus are double-duped.

 

As for whether to time a movie digitally versus through contact printing, assuming I shot in a standard 35mm projection format like 1.85 or anamorphic 2.35, ignoring the efx shots, transitional devices, etc. generally contact printing looks better because of the compromises introduced by digital intermediate (2K being the primary one), but creatively, you may find that the digital method will yield superior results in achieving some unique look that can't be achieved in traditional photochemical color-timing. But once you toss in format conversion as well because you shot in a non-projection format, then the argument for doing it digitally becomes greater.

 

All these processes, contact printing, optical printing, digital intermediates laser recorded out to internegatives, they all impart a look, a texture, but it's not about just one process always being preferred or superior over the other -- there are real-world issues to consider.

[Leo Anthony Vale]

This is completely correct regarding 'normal exposure'. What I have suggested is 'total over-exposure' when producing the B&W Matte Negative (from the Camera Colour Neg.) which will turn "non-blue" into "black" and "semi-blue" into "essentially black" on a B&W Negative. The 'Matte Negative' is what would be used as the Matte. It's like shooting a light source with B&W Film. The B&W Negative would be totally black, and the Positive would be clear. With my idea it's important that the Blue Light (during exposure of the Matte Neg.) not be too bright so that it doesn't penetrate the Yellow Dye Layer formed on the Camera Negative by the Bluescreen on the Set. The Bluescreen forms an area on the Camera Neg. that filters only Blue Light, and therefore using Blue Light with 'normal' and 'reversal' developing (and over-exposure) you can produce the Male and Female Mattes.

 

What you are describing is basically the "white screen process". Where the foreground is filmed against a bright white background so that the background on the negative is very dense and the foreground cann be printed to a high con matte. Over exposing the matte will cause it to bleed, so there will be white fringing.

This is the process used in the scene of King Kong breaking open the gate in the Cyclopean wall and the flying shots of George Reeves in the 'Superman' TV series.

 

So it can be used in B/W and color. But the regular method of making mattes in the standard blue screen processes will give a cleaner fit.

 

I dont think you'll see the flaw until you actually try making mattes with your system.

 

But it's all moot anyway.

[K Borowski]

Terry: while your intentions are good and noble, your logic is flawed. You're trying to improve a process that was obsolete in the late '80s.

 

I'd say that the biggest improvement that could be made along the lines of optical printing would be a printing head (I think someone here conjured up an idea for an LCD lightsource in contact with the negative) that could render a lot of the same effects as a DI without killing the image resolution, so you can do many (not all) of the color tweaks possible with a DI during regular contact printing of your film.

 

What else? Maybe an estar camera negative with a thinner base to double camera loads or halve the weight/size of magazines for a given amount of shooting time.

 

Color accuracy in fiml could use a lot of improvement, as could dye stability. Intermediate stocks are due for an improvement.

 

There's a whole lot of areas where film still has quite a bit of room for improvement, but, sorry, you aren't going to revive analog special effects no matter how hard you try. This is one area where digital (now I mean digital COMPOSITING, not CGI; CGI is sh-it) is on top in the game.

 

At the same time, there is something beautiful about all of the Star Wars/Star Trek films, even the Star Trek Original Series. I really think when they went back and redid some of the beautiful matte paintings from the original series to make them look more realistic, they did a disservice to the show.

 

I think then that there could also stand to be some more improvements in the technology for scanning and manipulating analogue matte paintings an dactual models.

[Terry Mester]

I don't know what he means by "256 shades per color in 24-bit" but 10-bit LOG RGB is something over 1000 luminence values per channel. I don't know what he means by "shades". You get other colors by mixing red, green, and blue -- not within red, green, and blue individually. As far as I know, 24-bit per color is more than 256 luminence "steps" in each color.

...

In terms of "richness" of color, I can tell you that the prints for "Astronaut Farmer" are more saturated and contrasty than if I had printed the original Fuji Eterna photography, because we were able to add that look digitally. So shots made in flat, overcast weather could be boosted to look more like the shots made in sunny weather.

 

I'm not saying there are no problems or quality hits with digital scanning and recording, because there are. There is the primary problem of the fact that most of the work is done at 2K when it should be done at 4K. And there is the creative problem, i.e. given digital color-correction tools that allow you to make changes within the frame (the most radically being in "Astronaut Farmer" that we turned green trees into yellow trees to match the ones that turned yellow in the Fall, sometimes we had to turn yellow trees back into green trees) is that you can generate digital artifacts that are not film-like. But given all the powerful tools of digital color-correction, it's hard to not use them creatively or to use them to fix and match shots, even improve missed focus, that normally would not be fixable in normal printing.

 

And if you choose to use an optical printer to convert a Super-35 movie to a projection format, now the entire movie has gone through an IP/IN generation, and any duped shots in the movie made using an optical printer, like for dissolves, already went through an IP/IN generation and thus are double-duped.

 

All these processes, contact printing, optical printing, digital intermediates laser recorded out to internegatives, they all impart a look, a texture, but it's not about just one process always being preferred or superior over the other -- there are real-world issues to consider.

In terms of Digital Colour, "Bits" is referring to Computer Memory which comes in Bytes (8 Bits). Since there are 3 Primary Colours, 24-Bits per Pixel allocates 8-Bits per Colour. 8 Bits of Computer Memory can record up to 256 different numbers (0-255). 10 Bits of Memory can record up to 1024 numbers (1K). Each number represents a pre-determined Shade of Primary Colour. "Shades" are simply the "brightness" or "intensity" levels of each Primary Colour (R,G&B). If you increase or decrease the brightness of only one Primary Colour Light Wave, the 'colour' and 'contrast' of the combined RG&B Light Ray is changed. The fact that 10-Bit colour can record 1024 different Shades per Colour in the Computer Memory does not mean that the CCD Pixel Sensors, which scan the original Film into the Computer, can accurately produce 1024 different Voltages to represent 1024 different Shades. What if it can only produce 800 or 500 or 300 different Voltages. You have no way of knowing. CCD Light Sensors cannot provide the colour accuracy of Film. The Computer will just arbitrarily assign the R,G&B Shades for each Pixel -- whether it's accurate or not! Furthermore, the linear CCD Pixel Sensors will divide and chop-up one Light Ray between 2 or even 4 Pixel Sensors which distorts and perverts the original Light. If you wish to see the difference in quality between Film and Digital, make a 8x10 paper Photograph from a Camera Negative Frame, and make a Digital Picture from a computer scan of that same Frame. On the paper copies you can then easily observe the deficiencies of Digital.

 

I was wondering, in Astronaut Farmer did you make a Digital IN for the whole Movie or just the special effects Scenes? Also, what was the general Film Speed used? 200 ISO and higher Camera Negative will suffer worse colour degradation during duplication than 100 ISO and lower. There's a rule of thumb that should be followed: whatever Speed of Film you start out with, it's all downhill from there.

 

The issue of films shot in Super35 is also pertinent. Last summer I had recommended to Kinoton and Cinemeccanica that they develop a Projector that can run both Regular and Super 35mm Movies. This would basically only require the Lens to be movable sideways by 1.47 mm. For Super35 Movies the DTS Time Code could be replaced by an electric pulse each time the Shutter opens. The two Digital Sound Tracks might need some adjustments with Super35. This would end the need to convert Super35 to Regular35. However, they indicated a need for development cost funding from Hollywood.

 

I don't disagree with you at all about the conveniences of digital editing. I have no doubt that the younger generation of filmmakers wants all digital because they have been spoiled by its conveniences. With the 4K Recorder, some digital editing here and there doesn't concern me. However, if the whole IN is being printed digitally, then this is not good. The unfortunate side-effect of the Digital Film Recorder is the lazy propensity to replace more expensive real special effects with cheaper CGI effects. The Studios are spewing out Movies with phoney-looking CGI special effects, and they don't really care about the quality of the Theater experience. They expect the movie-going public to put up with it. Theater attendance will suffer as a result. This is a direct consequence of the ability to make digital film prints.

 

What you are describing is basically the "white screen process". Where the foreground is filmed against a bright white background so that the background on the negative is very dense and the foreground cann be printed to a high con matte. Over exposing the matte will cause it to bleed, so there will be white fringing....

This white screen process is very interesting. I didn't know there was an equivalent to Bluescreen for B&W Movies. Do you know what Speed of Film was generally used to shoot Bluescreen Scenes? I personally would recommend 50 ISO to limit bleeding.

[K Borowski]

I was wondering, in Astronaut Farmer did you make a Digital IN for the whole Movie or just the special effects Scenes? Also, what was the general Film Speed used? 200 ISO and higher Camera Negative will suffer worse colour degradation during duplication than 100 ISO and lower. There's a rule of thumb that should be followed: whatever Speed of Film you start out with, it's all downhill from there.

 

The issue of films shot in Super35 is also pertinent. Last summer I had recommended to Kinoton and Cinemeccanica that they develop a Projector that can run both Regular and Super 35mm Movies. This would basically only require the Lens to be movable sideways by 1.47 mm. For Super35 Movies the DTS Time Code could be replaced by an electric pulse each time the Shutter opens. The two Digital Sound Tracks might need some adjustments with Super35. This would end the need to convert Super35 to Regular35. However, they indicated a need for development cost funding from Hollywood.

 

I don't disagree with you at all about the conveniences of digital editing. I have no doubt that the younger generation of filmmakers wants all digital because they have been spoiled by its conveniences. With the 4K Recorder, some digital editing here and there doesn't concern me. However, if the whole IN is being printed digitally, then this is not good. The unfortunate side-effect of the Digital Film Recorder is the lazy propensity to replace more expensive real special effects with cheaper CGI effects. The Studios are spewing out Movies with phoney-looking CGI special effects, and they don't really care about the quality of the Theater experience. They expect the movie-going public to put up with it. Theater attendance will suffer as a result. This is a direct consequence of the ability to make digital film prints.

 

This white screen process is very interesting. I didn't know there was an equivalent to Bluescreen for B&W Movies. Do you know what Speed of Film was generally used to shoot Bluescreen Scenes? I personally would recommend 50 ISO to limit bleeding.

 

YES, the whole film Astronaut Farmer went through a DI, at least from what I've seen here on the forum. Fortunately, they were at least able to manipulate it all at 4K before downrezing to 2K at the final out put stage.

 

The same problem exists with getting digital images onto photographic paper. The average digital printer can only resolve 250-300DPI, as opposed to at least 400 usable lines from a good optical print from a fine-grained negative.

 

The problem isn't that people are spoiled, it's that people aren't learning how to light and filter and are having to fix a lot of problems in post. I heard part of the commentary for the film "Master and Commander" which went through the digital intermediate process, and some of the stuff they were talking about making mistakes with, like gross underexposure, was kind of worrisome to me, because they were essentially using the process to hide mistakes they shouldn't have made in the first place.

 

There's no doubt in my mind that digital is far easier for color correction and color consistancy purposes. I bang my head against the wall every weekend trying to keep fleshtones consistant across different film and paper emulsions in different lighting conditions, and it can be hell on Earth sometimes getting the colors and contrast just where you want them with color. Don't get me wrong, when the analog color film/paper system clicks, it beats the piss out of digital imaging, but when it doesn't, argggh is it frustrating going back and realizing "hey her skin is way too yellow in this shot, must have been because of a stray cloud", or "ahhh poop, I know why the colors are different for part of the group than the rest: some were getting some sunlight, some WEREN'T getting supplemental lighting illumination, an some weren't getting supplemental illumination and were getting sunlight at the same time. So I can totally relate to where David is coming from with having a hard time with roaming clouds affecting color balance. It's always frustrating when a *stray cloud* ruins a good take, which is what would have been the case in the past, they'd have to go back and reshoot and just throw that take out. I really hate when DI is used overzealously, Technicolor's "The Aviator" one was particularly infamous, both for "smudging" the whole film to the point that I thought it had been projected out of focus, and hwo they totally didn't get the look of Kodachrome and 2 & 3-strip technicolor right. When digital can do something that a filter cannot do, a corrected lighting ficture cannot do, then I am all for it as an interesting new creative route, but it has plenty of obvious visual quality drawbacks compared to optcial printing.

 

Let's not forget that, since few if any studios make laser recordings of every copy of every distribution print directly, but rather make several "internegatives" for making prints, that there is further generation loss, I'd estimate 15-20% of image information, which theoretically brings the actual resolution of a 2K-output film projected in a theatre down to a resolution lower than just getting an HD TV signal hooked up at home.

 

So I do think it is important to either try to maintain more optical printing through perhaps improved optical means, like that LCD design, or introduce a standard taht can actually resolve *more* information than 35mm so that quality loss will be effectively completely eliminated for two or three generations worth of copying.

 

Regards,

 

~Karl

[David Mullen ASC]

Actually, the 2-strip Technicolor look of "The Aviator" is a good example of where using a D.I. makes more sense than any other technique, whether or not you thought it was too extreme. You can't use a camera filter to create that effect and even pulling YCM separations in an optical printer and dropping the blue record isn't going to quite be the same (although it may be workable.)

 

As for additional loss from IP/IN steps, the solution would be to both do all-4K for D.I.'s and then digitally record out as many "original" negatives as needed for making mass release prints. Then you'd actually get better 35mm prints in the theater than you currently get even with non-D.I. movies that go through an IP/IN.

 

It's certainly possible in theory since 4K post is dropping in price and laser recording is also getting cheaper (there have been some big HD features that have done this to save some resolution, make a bunch of original printing negatives.)

 

Intermediate stock is pretty good these days -- I just think there will always be limits in quality when you copy a piece of film twice, unless the intermediate steps were in 65mm or something, but even that wouldn't stop the loss of color saturation that happens, or the flattening out of the blacks. Although that can be fixed by insisting that all release prints use Vision Premier print stock.

 

Yes, projecting contact-printed Super-35 would look great (you'd have to have digital-only soundtracks though), it's just never going to happen. There would only be about 100,000 35mm projectors worldwide to convert, that's all...

[Sam Wells]

A lot more people can do composites on their home computers than own optical printers -- digital technology has brought compositing to the masses in a way that optical printing work could never do! Your average high school amatuer filmmaker is home doing chromakey composites routinely, whereas when I was in film school and actually took a 16mm optical printer class, we didn't even mess with chromakey stuff due to all the separation and matte work needed. Now chromakey composites are done every day by kids with laptops.

 

I owned a JK optical printer at one time, it was a pretty good machine (albeit "slow" but optical printing is not exactly real time <_< .)

 

I still see interesting work, especially in experimental films, done with optical printing but this is typically very 'optical' intervention (sorry to use artworld jargon du jour) -- illuminating the source off-axis, and so on.

 

I couldn't dream of doing the things I'm doing on a G5 now with _any_ optical printer. (And they're still a lot of work). The contrast build up as David mentions _alone_ would kill it. Let alone 12 image layers at once. And still maintaing control over color.

 

I'm all for photochemical. I'm all for getting it in the camera. I like reversal film even, WYSIWYG.

 

But really, once you cross the border out of the purely photographic encounter and into composite image creation, it's something different, you're _not_ dealing with that native 'track of the light encountered' - to me this is a cross discipline, something akin to painting.

 

I'm doing some things now I've wanted to do for 25 years, now I can.

 

-Sam

[Will Earl]

The Studios are spewing out Movies with phoney-looking CGI special effects, and they don't really care about the quality of the Theater experience. They expect the movie-going public to put up with it. Theater attendance will suffer as a result. This is a direct consequence of the ability to make digital film prints.

 

I don't know of any VFX house (big or small) who 'spews' out phoney-looking CGI, every effort is put into making each shot believable and of the highest quality, if the effects don't work then it's not through a lack of caring.

 

If you look at most of the studios big money winners within the last year or so, then you'll see a lot of films which have gone through a DI process, low audience attendance isn't a direct result of digital film-outs.

 

It's great that you've thought about improving image quality, but it's about now in your thinking stage that you need to introduce some pratical considerations into the mix. What your suggesting (if it works) has a lot of artistic restrictions or limitations placed on it with regards to filmstock used and shooting conditions. You also haven't fully weighed up the cost and time it would take, spending a week to improve the quality of a film print for little to no perceivable gain might not be the best way to improve the overall quality of a shot or the film.

[Terry Mester]

Intermediate stock is pretty good these days -- I just think there will always be limits in quality when you copy a piece of film twice, unless the intermediate steps were in 65mm or something, but even that wouldn't stop the loss of color saturation that happens, or the flattening out of the blacks. Although that can be fixed by insisting that all release prints use Vision Premier print stock.

David, do you know whether the Light source in Contact Printers uses AC or DC Electricity? This would definitely make a difference in print quality. The Human Eye cannot see the difference between AC / DC light, but DC Electricity would definitely provide better quality light for copying Films.

Yes, projecting contact-printed Super-35 would look great (you'd have to have digital-only soundtracks though), it's just never going to happen. There would only be about 100,000 35mm projectors worldwide to convert, that's all...

I was only proposing that they re-design a "new" Projector -- not convert existing Projectors in Theaters which is much too complicated. For their part, Theaters will never spend money to buy a "Super35 only" Projector because there are so few Super35 Movies being filmed. However, if through attrition they could replace Regular35 Projectors with dual Regular/Super35 Projectors, then within a few years time each Cineplex outlet would have at least one Projector capable of running Super35. This would make it safe (and cheaper) for filmmakers to start filming with Super35. A Super35 Print is basically only 75% as long as Regular35. The Studios would save a lot of money by being able to send out Super35 Film Prints. They could send out one-third more Prints using the same amount of film as Regular35. It would make a lot of sense for them to pay the costs of developing a new dual Projector.

[Terry Mester]

There's no doubt in my mind that digital is far easier for color correction and color consistancy purposes. I bang my head against the wall every weekend trying to keep fleshtones consistant across different film and paper emulsions in different lighting conditions, and it can be hell on Earth sometimes getting the colors and contrast just where you want them with color. Don't get me wrong, when the analog color film/paper system clicks, it beats the piss out of digital imaging, but when it doesn't, argggh is it frustrating going back and realizing "hey her skin is way too yellow in this shot, must have been because of a stray cloud", or "ahhh poop, I know why the colors are different for part of the group than the rest: some were getting some sunlight, some WEREN'T getting supplemental lighting illumination, an some weren't getting supplemental illumination and were getting sunlight at the same time. ...

Hi Karl,

You sound like a perfectionist like myself. "Good enough" of course is never good enough, but sometimes you need to accept "great enough". 'Real Light' itself is "imperfect", and this is why Photography and Cinematography will also always be imperfect. Only a Portrait Studio can provide you controlled conditions. When working outside, you need to take a whole lot of Pictures, and then you're more assured of getting enough good ones. If necessary, you can cheat by scanning your Negative into your Computer, and making a colour-corrected Digital Print. When you're on a tight deadline, you sometimes need to make compromises -- especially when you have all those competing Digital Photographers offering their cheap alternatives to your clients. The client will have to decide if they wish to settle for cheap photography.

I don't know of any VFX house (big or small) who 'spews' out phoney-looking CGI, every effort is put into making each shot believable and of the highest quality, if the effects don't work then it's not through a lack of caring. ...

It's great that you've thought about improving image quality, but it's about now in your thinking stage that you need to introduce some pratical considerations into the mix. What your suggesting (if it works) has a lot of artistic restrictions or limitations placed on it with regards to filmstock used and shooting conditions. You also haven't fully weighed up the cost and time it would take, spending a week to improve the quality of a film print for little to no perceivable gain might not be the best way to improve the overall quality of a shot or the film.

Hi Will,

I realize that there are many compromises that need to be made to get a Movie finished on time and on budget. The financing behind the project is key to what you can and cannot do. A guy involved in financing told me that I should cut the costs for my own Screenplay down to about $12 Million. I could get down to $20M but not $12M. I personally would rather not make my Movie than make a cheap Movie. Each filmmaker has to decide for themself what compromises they're willing to accept. The contact copying process definitely has room for improvement, and this would benefit all cinematographers. For example, colour and contrast degradation would be reduced if the copy quality is improved.

 

How would you rate the CGI in the latest Superman movie and this new 300 movie? The special effects in the original Superman movies looked pretty damned real. I don't question how hard animators work to produce CGI. It's tedious frustrating work, and I wouldn't have the patience to do it. The task being given to CGI animators is impossible, but they still do a fantastic job. The root of the problem is that, even with 10-Bit colour, you cannot make CGI look like real Light. Only real Light looks real. There are many areas where the Studios opt for CGI instead of small-scale models which would probably not cost any more than CGI. Take an "explosion" for example, it's pretty hard to replicate fire in a computer. Hollywood Studios are filthy rich, but old Sumner and Rupert still want to cut costs to make more money.

[Leo Anthony Vale]

This white screen process is very interesting. I didn't know there was an equivalent to Bluescreen for B&W Movies. Do you know what Speed of Film was generally used to shoot Bluescreen Scenes? I personally would recommend 50 ISO to limit bleeding.

 

In which case, I'm assuming you're not aware of the "black backing" process either.

Thus you don't quite get the real difference between them and the bluescreen process.

 

The bleeding has nothing to do with the film stock used for photographing the blue screen sceen.

This is a characteristic of the lab stocks.

 

How familiar are you with various lab stocks?

 

Contrasty print stocks have less laditude than camera stocks.

Exposure and processing will affect the size and edges of the mattes. Having the mattes fit perfectly is one of the most important aspects of gettig bluescreen to work, but getting a perfect fit for the mattes is very tricky and is where the effect most frequently falls apart.

 

Doing it digitally is for more efficient.

[Terry Mester]

In which case, I'm assuming you're not aware of the "black backing" process either.

I've actually thought of the scenario of using a "black box" background instead of Blue / Greenscreen. While it could eliminate the need for creating a Female Matte, it prevents the ability to create the Male Matte.

There is one situation where a Bluescreen background would cause a problem. Filming a light source such as 'fire' will suffer blue seepage since it's partly transparent. However, you could film the fire in total darkness which would eliminate the need to create a Female Matte. The Male Matte would have to be manually (or digitally) painted. Since the fire is bright and partly transparent, it's not necessary for the Male Matte to come right up to the outer edge of the fire.

[Michael Nash]

There is one situation where a Bluescreen background would cause a problem. Filming a light source such as 'fire' will suffer blue seepage since it's partly transparent. However, you could film the fire in total darkness which would eliminate the need to create a Female Matte. The Male Matte would have to be manually (or digitally) painted. Since the fire is bright and partly transparent, it's not necessary for the Male Matte to come right up to the outer edge of the fire.

 

Effects such as fire, smoke, water and so forth are commonly done against black for this reason.

[Leo Anthony Vale]

I've actually thought of the scenario of using a "black box" background instead of Blue / Greenscreen. While it could eliminate the need for creating a Female Matte, it prevents the ability to create the Male Matte.

 

It doesn't prevent the ability to create a matte.

Look at the homonculi scene in 'The Bride of Frankenstein', Whale's 'The Invisible Man' and 'The Incredible Shrinking Man'.

All black backing and the mattes are nowhere near perfect, but the scenes are gems of lining up action and timing.

 

Your bluescreen process is using the principle of the black and white backing techniques; throw lots of light onto the matte film to overexpose it.

While the blue backing combines opposite colors and tones to cancel out foreground detail and copy that for the mattes.

 

Go over the link to the Vlahos patent and study the diagrams to see how that works..

Edited March 22, 2007 by Leo Anthony Vale

[K Borowski]

Actually, the 2-strip Technicolor look of "The Aviator" is a good example of where using a D.I. makes more sense than any other technique, whether or not you thought it was too extreme. You can't use a camera filter to create that effect and even pulling YCM separations in an optical printer and dropping the blue record isn't going to quite be the same (although it may be workable.)

 

Sorry, but I disagree that DI was best, because they spent all of that money on having someone make it "look like Technicolor" and then it didn't! I only wish I got paid so well for failing so miserably.

 

They made the colors look "bad" and "funky". Maybe that's all some people associate with the old stocks, but I defintiely think there is an incredible pallate there, sort of like what you'd get with Kodachrome, only even more punchy. For 2-strip specifically, well, it's kinda cyan and orange are your only two colors. I could take a picture right now, using just that information and come up with a close proximity to what the process looked like digitally.

 

Now if I had money though, again probably a whole hell of a lot, I'd find an old Technicolor camera, probably there are so many collecting dust laying around that they don't know what to do with them, and bipack film just like they did back in the day. Find some panchro B&W stock, maybe try to have Kodak put the same sort of colored filtration coating on the back of the one strip, and run it through.

 

Hollywood nowadays is understandably nervous about non-standard technology, but if you run a couple of test with something and it works, they really should go with stuff like that instead of the tired cliche of just timing 35mm negative film to look like it, which it never could or will.

 

My point is, when you have access to the technology that made the stuff (I think 2 strip could actually be done on several standard-issue period 35mm cameras IIRC), and it was a proven technology with quantifiable characteristics, why not go with the actual technology instead of faking it. Hell, it'd probably be cheaper, even coating dyes on the film, to shoot 2-strip than get the same effect through DI.

 

People need to get their hands dirty more in Hollywood, take some risks, be creative again! Formulaic methods of doing things are never conducive to the creative process, so no wonder film sets can turn out such garbage nowadays.