|
|
 
|
|
Author
|
Topic: Question about Dolby 3D
|
|
|
|
|
|
|
Julio Roberto
Jedi Master Film Handler
Posts: 938
From: Madrid, Madrid, Spain
Registered: Oct 2008
|
posted 02-03-2010 02:36 AM
In a nutshell, the micromirrors on the DLP chip can only reflect (or not) the light SOURCE, faking a grayscale. In a DLP projector, the (white-ish) light goes through 3 color filters, as we know, to fake the colors.
Micromirrors can't strip frequencies contained within the white light mix of the lamp that the ("relatively broad spectrum") color filters allow to pass through.
You could build a DLP projector with 3 narrow "RGB" filters that match ONE eye of the infitec scheme, but then you would need a dual-projector setup to use another projector specially made for the other eye.
Or build a special projector with 6 DLP chips instead of 3. Two each for the two different reds, greens, and blues.
So the problem is that the RGB color filters in the DLP allow for a range of blues, reds and green contained within the white of the lamp. Then, the micromirrors reflect THE WHOLE range of frequencies that passed through the color filters more-or-less, modulating the INTENSITY, not the frequency.
The mix of the 3 rgb INTENSITIES is what SIMULATES our eyes seeing different colors, giving the impression of different frequencies, but it's just a ("grayscale") mixture of 3 "fixed" frequencies (ranges) by the DLP assembly built-in color filters.
The white light of the xenon is broad-spectrum enough to get it to make two different (narrow) red, green and blue bands so that their mixture ALMOST (but not quite) include the DCI xyz gamut. And luckily the RGB filters on the DLP assemblies, being designed for maximun illumination, are broad spectrum enough to allow for each of the two reds/greens/blues to pass. So the filtering can occur either at the light source or after the light has gone through the projector (i.e. Dolby could've used an external color wheel similar to master image's polarized one). But the projector must be calibrated for the Dolby filters as well as the movies color corrected for the new, reduced, color space.
| IP: Logged
|
|
|
|
|
|
Julio Roberto
Jedi Master Film Handler
Posts: 938
From: Madrid, Madrid, Spain
Registered: Oct 2008
|
posted 02-03-2010 01:30 PM
The filter in the glasses indeed filter out all the light's frequencies except a (narrow) band of red, green and blue. This band is slightly different (and mostly "exclusive") on each eye.
Once again, it can't be "filtered out digitally" because the only thing you can do digitally is limit the INTENSITY (amplitude) of the SOURCE light, not the frequency or said light. The DLP engine is always going to output the same 3 frequency (bands) allowed by the built-in color filters and provided by the xenon spectrum. All you can digitally control, is HOW MUCH of each frequency hits the screen, not change the frequency itself.
When you see an image "digitally modified" to look "yellow", it's not (necessarily) "real yellow" that you are seeing, but a mixture of red-green-and-blue frequencies that, to the eye, look (about) the same as "real yellow". Again, you are not changing the frequency (bands) of the lights by the digital modulation, just the mixture of the three "fixed" frequency bands you are working with.
There are MANY ways to create "almost identical-looking yellows", and the DLP just chooses ONE of the methods and it CAN'T choose most of the others.
As long as the source lamp contains enough of the red, green and blue spectrum, all the other colors can be SIMULATED (not really obtained). If the xenon lamp didn't have any REAL yellows,i.e., attaining yellow would thus, by definition, be impossible. But if it has enough red-green-blue, you can produce a mix that would ALSO look yellow without the frequency or either the red, the green or the blue being, obviously, the same frequency as yellow. If you look at the image through a filter that blocks the yellow frequency, you would STILL see yellow because it allows the red-green-and-blue frequencies that, when mixed in the right amount, ALSO look yellow.
It's like using say REAL green ink to print a picture vs using Yellow/Magenta/Cyan/Black ink-jet. You print with the inkjet a picture of a forest, full of "green" (but using only yellow/magenta/cyan/black inks). Imagine you had a chemical that washes out ONLY "chemical green ink". You wash your picture and it still looks the same, green as before. That's before it LOOKS like it has green ink, but it has none at all. No matter how you "digitally manipulate your image", your printer just can not produce "real green ink", just a faked approximation using the inks provided by the cartridges.
The DLP projectors have to work with:
A) the color sprectrum of the xenon bulb (which is quite wide) and
B) The mixture of red-green-and-blue allowed by the filtering in the color prism assembly, which allow as much red, as much blue and as much green as possible.
No other way to control the light's frequency after this point "digitally". Only the intensity of each color channel (3 DLP chips) independently at a pixel level.
With film, I guess you could create a specialty (i.e. not off-the-shelf) formulation that could perhaps produce a projectable "Dolby 3D" picture by doing multipass processing or something, sort of like IB prints. Otherwise, the same problem exists. The color of each band of the film is "fixed" by the formulation of the emulsion and you expose the whole frame at the same time when you photograph it. With a color separation step, and filtering during developing, a formulation could be developed to contain two different layers (or a mix of two dye particles) of magenta/cyan/yellow, each very narrow spectrum and overlaped but exposed at different times (in two passes, each through a different filter), thus creating a "projectable Dolby 3D movie".
It would be akin to a Vectograph print, which yields "projectable 3D polarized 35mm print" on an unmodified projector without any filters.
| IP: Logged
|
|
Mark J. Marshall
Film God
Posts: 3188
From: New Castle, DE, USA
Registered: Aug 2002
|
posted 02-03-2010 03:52 PM
Ok, we're getting pretty technical here, and I admit that I'm out of my league a little bit but I'm trying to understand this.
Whereas DLP adds light from three basic colors to produce the picture, film actually subtracts (or blocks) light to produce the picture. So explain to me where I'm going wrong here.
If you start off with clear film, ALL of the frequencies from the xenon bulb are making it to the screen.
The yellow, cyan and magenta layers can't ONLY be blocking out their respective yellow, cyan and magenta frequencies. If they were, then all of the rest of the frequencies of light would be passing through to still hit the screen, yes? If that's true then why can't we get the emulsion to block the very specific and narrow bands of red, green, and blue light for the Dolby glasses?
I don't care about DLP here. Talking about film.
| IP: Logged
|
|
Julio Roberto
Jedi Master Film Handler
Posts: 938
From: Madrid, Madrid, Spain
Registered: Oct 2008
|
posted 02-03-2010 06:21 PM
Same thing with film, but with substractive mode.
The film has three FIXED FREQUENCY BANDS "color filters". These are "tuned" for MAXIMUN color range ("broad frequency") ONE for cyan, ONE for magenta and ONE for yellow. Each band may be (and often is) compossed of several layers in modern film, but let's say they were only one layer for simplicity sake.
Infitec filters strive for MINIMUN (or hardly overlapping) *SIX* frequency bands, two for red, two for green, two for blue.
If you wanted to manufacture a "Dolby 3D film", you would have to find a way to modulate (change) the INTENSITY on EACH of the SIX frequencies, not just the three light-sensitive "layers" it usually has.
If your film emulsion allowed for "all of the reds from the lamp" to pass through ... how are you going to "modulate" ("print") 2 *DIFFERENT* "grayscale" images on the same light-sensitive layer? And even more, how are you going to block a certain frequency from the lamp if all you can do is block a larger or smaller AMOUNT of light, thus modulating intensity only, not frequency?
A special film could be manufacture with "six" dye particles configured to allow for a mix of two narrow bands of red, green, and blue light to pass through the filter. But it would also need layers to allow for light sensitivity during exposure ONLY to those bands, since the same part of the film is going to get hit by TWO different bands to print each one of the images, one for each eye.
In another words, we are back to a specialty-made film that would need, basically, twice the number of (important) layers of current film. It's doable, but not off-the-shelf.
The short answer is: when you modulate a grayscale image into a color layer, all you are doing is deciding how MUCH of the light goes through that layer, not what frequencies get blocked. That's decided by the "layer" itself and it's fixed. And if you only have 3 layers (rgb or cmy, doesn't matter), that's 3 frequency bands you can modulate. We need 6 for Infitec 3D scheme. Thus either a rotating filter with two filters to "double up" the bands of colors from the light source, or a specialty made projector with 6 DLP chips/6 film emulsion layers or dual-projection (3 bands each).
| IP: Logged
|
|
Christopher Seo
Jedi Master Film Handler
Posts: 530
From: Los Angeles, CA
Registered: Jun 99
|
posted 02-03-2010 09:09 PM
The fact that film dyes do not block a narrow frequency of light is just the problem with implementing a Dolby/Infitec 3D process on film.
Each dye blocks out a pretty large frequency band of light - about one-third of the visible spectrum, for obvious reasons if a three-dye process is to be used. Each frequency band roughly corresponds to a primary additive color. Yellow dye blocks mostly blue light, cyan dye blocks mostly red light, and magenta dye blocks mostly green light. Thus the sensation of seeing yellow on film is produced not by seeing yellow frequencies of light per se (though that happens, of course), but by seeing light that is 'not blue'. This includes an entire range of light frequencies spanning green, yellow, orange, and red. Take a look at the Spectral Dye Density Curves on p. 6 of the Kodak Vision Print Film brochure. Color is noted as wavelength, which you can interpret with the color spectrum on Wikipedia.
But notice with the Kodak dye curves how non-ideal the dyes are. There is some unevenness, and a lot of overlap, which wouldn't work for Dolby 3D. As Julio pointed out, there would need to be six dyes and their response curves would have to be close (so that there are two that look yellow, two that look cyan, etc.), but the curves could not touch at all - any overlap would manifest as ghosting. Look at these (apparent) Infitec response curves here and you'll see what I mean. Complete separation between each curve is needed.
I've been told by a Dolby tech that the filters on the glasses require 40 or so layers to work properly. I think that developing an emulsion version of this would be very daunting, especially with color-coupler processes which can only use certain dyes to begin with.
| IP: Logged
|
|
|
|
|
|
|
|
|
|
All times are Central (GMT -6:00)
|
|
Powered by Infopop Corporation
UBB.classicTM
6.3.1.2
The Film-Tech Forums are designed for various members related to the cinema industry to express their opinions, viewpoints and testimonials on various products, services and events based upon speculation, personal knowledge and factual information through use, therefore all views represented here allow no liability upon the publishers of this web site and the owners of said views assume no liability for any ill will resulting from these postings. The posts made here are for educational as well as entertainment purposes and as such anyone viewing this portion of the website must accept these views as statements of the author of that opinion
and agrees to release the authors from any and all liability.
|
Home
![[Smile]](smile.gif)
![[Roll Eyes]](rolleyes.gif)
Printer-friendly view of this topic