For this second in this series on the Pico Projector “use model” I want to discuss the issues with ambient light. A front projector, like most pico projectors today, have to deal with the fact their best case “black” is limited by the the existing ambient light.
I have been all over the world with pico projectors for about 4 years with different light outputs and in different lighting environments. You might say, I am a “pre-early-adopter.” I’m also very big on “user interfaces” having started work on one of the earliest bitmap graphics processors aimed at graphical user interfaces, the TMS34010, before there was a Mac®, LISA®, or MS-Windows® (the TMS34010 was used as an accelerator for Windows 1.0 and the second generation TMS34020 came close to being used in the Color Macs).
One thing I know with certainty is the “use model” model of what I call “shooting on the wall” like a traditional conference room projector does not work very well. First, there is the problem that typical rooms don’t have a nice white or even solid light color wall and if do happen to have a nice screen, then more than likely they have a 2000+ lumen projector aimed at it already. Second, you have the problem that with most pico projectors there is no “zoom” feature so that even if there was a wall to project on, it may be too far away to be usable with a sub 100 lumen projector. Third is the problem of ambient light, or what I call the “Dracula effect” and the need to rid yourself of the ambient light.
Consider the table above from the paper “Projector Images and Room Light” (there are a papers with similar information if you do a search but this table was conveniently in lumens per square foot).
So, you go around and close all the blinds, turn off most of the lights, but there is a still some light so people can see what they are doing. From the table in a dimly lit room, there are 1 to 2 lumens per square foot. Which means if you project an image that is one square foot, “black” is 1 to 2 lumens. So if you are trying to use a 10 lumen projector in a dim room, you’re starting with best case of 10:1 or 5:1 contrast regardless even if the contrast “spec” of the projector was infinite. Without getting into all the math, it turns out that the contrast of the projector is not much of a factor in the overall contrast when the ratio of the projected light to the ambient is less than say 30:1; this is why it is silly to talk about 1,000’s to one contrast in a 10 to 20 lumen pico projector. If you want to project a 50-inch diagonal image with a 16:9 ratio (say the wall is too far away or you want to show a lot of people), then you have nearly 10 square feet to fill, you better have blocked almost any other light hitting the screen.
The bottom line is that the “shoot on the wall” model with the projected lumen levels that can be expected (more on what can be expected in later posts in this series) from an “embedded” pico projector even with some massive improvements isn’t practical. All the demos with 80” or 100” images have to be done in very dark environments that people don’t normally like to be in.
The fact that light per unit area is inversely proportional to the square of the diagonal is one of the keys. My experience is that humans, even those that are scientist and engineers, tend to estimate linearly and underestimate the effects of square law. For example if we project a 16:9 image on an 11-inch sheet of paper, the light will less than half a square foot or 1/20th the area of the 50-inch diagonal image.
Conclusion: A sub 30 lumen projector is more suited to filling a sheet of paper than a wall.
To be continued . . .
In the follow-on posts in this series, I plan to write about the use of cameras for input and what can reasonably be expect in the way of brightness from an embeddable projector once laser have been perfected.