4 min read
With the advent of 6P (six primary) laser projection systems that generate a 3D image by using two distinct sets of laser primaries (two sets of wavelengths for the Red, Green and Blue primary colors) a curious discussion started taking place. Namely the question: “is each ‘eye’ in a 6P mode DCI compliant?”
Typically, my answer to this is “can you watch 3D with one eye closed?” – suggesting that the above question is irrelevant – as long as one has two eyes, a functional stereoscopic vision, and the net result is DCI compliant, one needn’t worry.
However, this question has made me realize that the situation with 3D is more complex now – laser itself is a new projector technology and brings more 3D technologies with it.
So let’s start from the beginning…
A cinema projector has a light source, a light integration rod, a color splitting prism and three DMD® (digital micro-mirror device) chips, one for each primary color. The light from the light source is mixed and homogenized in the integrator rod and then it passes through the prism that optically separates the red, green and blue beams of light onto a dedicated DMD chip. So each chip will create its own part of the image, and then the three images (red, green and blue) will be combined again in the projection lens to make the correct image with all its colors.
The key physical principle at work here is Grassmann’s Law which states that you can make any color by a linear combination of two other colors (so in case of an RGB color ‘triangle’ the projector can make any color in within it by mixing R, G and B in the right proportion). Yellow for example is a mixture of some red with some green, and white is a mixture of red, green and blue in the correct proportions, etc.
Stereoscopic 3D gives your brain the illusion of seeing a 3D image, by presenting each eye with an image whose perspective is slightly shifted from the other eye’s image. (Raise a finger in front of your eyes, and alternate closing each eye while looking at your finger and you’ll see what I mean). The amount of shift (separation) will determine the amount of 3D depth one perceives.
In principle, one needs a way to project two separate images, and separate the images in a way that each eye receives only the image intended for it. Otherwise both images will mix into each eye, and rather than a 3D effect, you will see a blurred image. The existing methods of separating the images are by color, by polarization, and by alternately blocking the unwanted image using active shutter glasses.
Two images can be projected for example in sequence, by using a single DLP® projector or in parallel at the same time, using two DLP® projectors.
The principle of color 3D is that two images are presented, each with slightly different sets of RGB color primaries. These two sets of primaries can be optically filtered by glasses (the left-eye glass filtering one image and the right-eye glass filtering the other image) and so your brains recombine or “fuse” the two images to perceive a 3D effect. The different color in a 6P laser projector is made by combining carefully selected wavelengths: a set of three wavelength bands in the red, green, and blue region for the left eye, and a set of slightly different RGB wavelength bands for the right eye, so that the 3D glasses can efficiently separate them. In this process, next to creating a 3D stereoscopic image, the brain also merges the two different colors to a single color that is a mixture of the two according to Grassmann’s law. This is the same principle for creating any other color using the R, G and B primaries.
Now back to the original question: “does each eye need to be independently DCI compliant?”
The answer is no. By putting glasses on and watching with both eyes, a person reconstructs the 3D stereoscopic image, and in case of Color3D, this is done by mixing the color of the left and right eye images to a new mixture. So, it is this combination that needs to be DCI compliant, not each eye’s image individually.
But still, you would argue, if one can make two eyes DCI compliant, why wouldn’t you do so, if only for the sake of a marketing argument?
It turns out there are very solid reasons for an optimum choice of wavelengths in a laser projector.
The image quality argument would push for extending the color gamut as much as possible, which means short green wavelengths and (very) long red wavelengths.
The next question however is the economical viability of these wavelengths, as well as the impact this choice has on the projector brightness, and on balancing towards the correct white point.
Whether you see light (and how bright it appears) depends on its wavelength. IR and UV light is not visible (= zero lumens) but both can burn your eyes. A green wavelength of 555nm is the brightest among all wavelengths, for a given optical power. So the further one departs from 555nm (in either direction), the more laser power one needs to project to generate the same perceived brightness (lumens of light). Very long red wavelengths (650nm or longer), and much shorter green wavelengths (e.g. 525nm) that is to say, ‘wide color gamut primaries’, are much less efficient in terms of lumens per optical watt, meaning that you need to stuff your projector with costly lasers that consume more electricity and require ever more cooling.
And finally, due to the same argument – not all wavelengths ‘weigh’ the same towards perceived luminance, if the wavelength choice is wrong, it will produce the wrong white point. As a result, the projector will lose more lumens and dynamic range in order to color-correct to the specified white point.
The Barco flagship laser projector series uses the above knowledge to produce the world’s first and only integrated laser 6P single projector solution, that is at the same time internally despeckled so it also works with silver screens and other 3D systems. In addition, it is the most energy-efficient RGB laser projector on the market, providing around 40% lower power consumption per lumen, than other high brightness Xenon and RGB laser projectors.