Quote:
Originally Posted by
Stoney Jackson 
Looking into my lens, the black opening looks like a backwards capital letter "D". is everyone's like this? What is the purpose of this design. I noticed some light leakage on the left of the screen too. Doesn't really bother me though as I plan to make the wall the screen will go on black.
Disclaimer: The information in this post reveals certain internal workings and structures related to the Qumi projector. I will under no circumstances take any kind of responsibility wether legally or financially for presenting this information in a public forum or how this information is used by the public. I only present this because it represents the facts and aids in discussion about the projector. I have never been associated with Vivitek or any company related to them.
With the short legal rambings out of the way, here goes. I had 2 Qumi projectors bought directly from retail and then performed a teardown to learn about the internal workings.
The inverted D you mentioned forms the aperture as seen into the exit pupil of the projection lens. The exit pupil is the side that faces the outside so it can be seen directly by everyone. Most people associate apertures as being a perfectly round circle, but this condition does not need to be strictly enforced because most optical systems have focal points at the electronic that are usually some kind of rectangular surface (ie. 35mm film, ccd sensor, microdisplay, you name it)
The aperture shape of the Qumi lens can be seen below:
Within the lens

The actual part

An additional picture for those interested is viewing the lens from the entrance aperture. This is only possible via disassembly, and I estimate the F number is between 2.1 ~ 2.3 (the F number as far as I know has not been publically disclosed by Vivitek)

Anyhow, all the stuff I presented above can only prevent stray light from affecting the picture, and is not the root cause of the light leakage. The inverted D is only one of the attempts to fix the stray light problem.
The real culprit is the right angle prism in the picture below

Those familiar with DLP projection systems will probably know about TIR prisms and RTIR prisms. The right angle prism used in the Qumi looks nothing like the prisms used by more traditional DLP projector architectures. From my perspective, the selection of the prism type works and the prism type chosen is much cheaper than traditional TIR and RTIR types. But as a trade off, it creates a mess with massive amounts of stray light that cripples the ANSI contrast performance if it is not treated properly. This is due to LED light sources being less well collimated than ideal (LEDs are even worse than bulbs) and there is substantial reflection of stray light in the on-state.
The physical mechanism why this is so can be explained by the use of total internal reflection in DLP systems. In addition to the TIR effects, there are also frustrated total internal reflection (FTIR) issues as the 45 degree surface on the prism is undoubedtly AR coated (this causes additional light leakage). Those curious can do the research as to why


I won't speculate wether it was a conscious design decision by Vivitek, but I guess they probably didn't fully anticipate the particularities of a LED light sources when designing the system.
At the end of the day, consumers can only fix this issue using the ghetto mod as shown by others above. To really fix the root cause of the problem requires a huge engineering investment, and thus it is best left to the original manufacturer to figure it out

PS. An engineering solution to the problem is possible