AVS Forum banner
1 - 11 of 11 Posts

·
Registered
Joined
·
21,464 Posts
Discussion Starter · #1 ·
Rather than post under an existing machine thread, I would like to start a specific thread on how a manufacturer could enhance the contrast (various types) of a single DC4 projector that uses three LEDs for the lighting source. Assuming the present use of PWM LED dimming (255 steps available for each LED) and further improvements tohow implemented, I assume that using irises could yield further improvement and\\or lessen artifacts. Cleary dynamic or adjustable static irises could be placed before the chip (say near the LEDs), somewhere after the chip up to the lens, or in the lens itself near the entering end. What would be expected effects of the placement positions and whether static adjustable or dynamic?
 

·
Registered
Joined
·
114 Posts
I think that a manual Iris would be a nice addition so to optimize contrast or light output for each setup or ocassion.

I might be wrong, but I don't beleive an automatic iris would add much to a PWM LED projector.

This is an electrionic system vs a mechanical one. After the right algorithms are found and chips for controlling the LEDs get produced in enough quantities/ by different providers the benefits will be enough to prevent an automatic iris to be implemented.

LED dimming is very fast, absolutly silent, and will have greater reliability than a mechanical iris.

Black frame insertion can be another benefit when LEDs became enough bright as to spare some output.


Disclaimer: I am not expert in LEDs and I am basing my opinion in that PWM dimming the LEDs doesn't induce color temperature changes / aberrations.
 

·
Registered
Joined
·
729 Posts
I do not think a dynamic iris will be needed in LED. Am I missing something? Even when LED becomes brighter you can design in the ability to light up only some of the LEDs in a color pack instead of all. This already happens in some designs in case a LED burns out the other color packs will turn off a LED to adjust and not destroy the color balance. A an adjustable static iris would be a plus in adapting a large projector to a small screen, assuming there is not enough electronic control and such a thing as a large bright LED projector.
 

·
Registered
Joined
·
23,188 Posts

Quote:
Originally Posted by Jim Burns /forum/post/16935982


A an adjustable static iris would be a plus in adapting a large projector to a small screen, assuming there is not enough electronic control and such a thing as a large bright LED projector.

And the adjustable static iris can bring higher native on/off CR as the white level is brought down, so preferable to just dimming the LEDs more in my opinion. The native on/off CR is important to the kind of dynamic on/off CR they can get without too many bad side effects. They may be able to go beyond the general 3x to 5x range for native on/off CR to dynamic on/off CR, but until they can go to total black without bad side effects I would say that there are advantages to increasing native on/off CR.


To Mark's question, I would go with something along the lines of the Marantz 11S2 or Sharp 20k, but with a little more control. Maybe like the BenQ 10k with 2 irises that I think would open and close together. Basically put one adjustable iris at the proper spot in the lens. Then put one at the proper spot before the DMD and as the lens iris is closed also close the previous iris in a way that matches up well with it. Perfection is basically where the light that makes it through the first iris makes a shape in the plane of the lens iris for white that matches the opening there, or something close to that.


Irises like the above could be used dynamically, but for the time being I think it would probably be better to just make them static during video and let the user pick the positions. Maybe with just a single slider that moves both irises even though the user may think only one iris was moving.


Given that BenQ did something like this with the 10k maybe they could take a step back from the dynamic irises they did with the 5k and 20k (which I think they did a poor job with) and use the basic 10k system for an LED projector.


--Darin
 

·
Registered
Joined
·
23,188 Posts
To followup what I was saying about 2 irises working together working well with single chip DLPs I figured I would give some numbers for the 11S2 from Greg Rogers' review. There are 3 iris choices that are basically both open, one closed down with one open, and both closed down. Iris 3 is most open. Here is what Greg got for lumens and CR for each mode with one setup (close throw with high lamp):


Iris 3: 842 lumens, 3340:1 on/off CR

Iris 2: 426 lumens, 5350:1 on/off CR

Iris 1: 379 lumens, 8100:1 on/off CR


People may notice that closing one iris down lowered lumens quite a bit, but also increased on/off CR a fair amount. Iris 3 had almost twice the lumens of iris 2, but iris 2 had 66% higher on/off CR. So, less than a 1:1 relationship percentage wise. But the other thing people may see is that once one iris is shut down there isn't much cost to closing the other one down (when the projector is designed properly), but there is good benefit to the on/off CR. In this case iris 2 only has 12% more lumens than iris 3, but iris 3 has 50% more on/off CR. This is why I feel that once you close one iris down, you might as well close the other one down (at least with single chip DLP). And if the irises had more choices then the iris in the lens could have been closed down partway and an internal iris closed down part way until they let 426 lumens through and the on/off CR would have likely been higher than 5350:1.


I've tried to explain elsewhere why I think this is the case with single chip DLPs, but I can do it again if anybody wants.


--Darin
 

·
Registered
Joined
·
1,499 Posts

Quote:
Originally Posted by darinp2 /forum/post/16936126



Iris 3: 842 lumens, 3340:1 on/off CR

Iris 2: 426 lumens, 5350:1 on/off CR

Iris 1: 379 lumens, 8100:1 on/off CR


--Darin

If the same thing could be done with the upcoming BenQ W3000, if you apply the same ratios as the Marantz' to the BenQ's speculated specs, it would look like this:


Iris 3: 600 lumens,2800:1 on/off CR + DB with 4X Multiplier = 11200

Iris 2: 303 lumens, 4485:1 on/off CR + DB with 4X Multiplier = 17940

Iris 1: 270 lumens, 6790:1 on/off CR + DB with 4X Multiplier = 27160


Even if they were to add only 1 iris, that would be great. If some one has any influence or contact with BenQ, get this done! Especially since the cost to add them is minimal I believe.
 

·
Registered
Joined
·
152 Posts

Quote:
Originally Posted by mark haflich /forum/post/16937679


emember I amtalking about all types of contrast.It will seemtome that PWM and LED dimming could not increase ANSI while an iris in the lens good.

If a lens iris can improve ANSI contrast at all, then its full open position might as well be set right there. This has nothing do do with dynamic irising.


Basically there is signal light and there is stray light passing through the plane of the iris. If the signal light is compact in that plane and the stray light is widespread (as was he case with darinp2's iris plane), then you can craft an iris to pass the signal light and preclude much of the stray light.


The mentioned stray light is of two kinds. Some of it arises from the signal light. This is probably mostly due to interreflections between the glass-air surfaces occuring before the iris plane. Reducing this improves ANSI contrast. Other of the stray light arises from the non-signal light. Reducing this improves the FOFO contrast.


darinp2 primarily improved the FOFO contrast by crafting a new iris. It did this by hardly reducing the Full On luminance. Obviously the lens iris and the projector illumination system should be designed in tandem. After that, LED dimming can do everything that adjustable irising can do.


All these discussions ignore the optical considerations on irising. Reducing the lens aperture reduces some aberrations while increasing diffraction. There is generally an optimum aperture which balances these effects. It is bad optical engineering to use a lens iris to adjust brightness. Better to dim the LEDs.
 

·
Registered
Joined
·
23,188 Posts

Quote:
Originally Posted by mark haflich /forum/post/16937679


emember I amtalking about all types of contrast.It will seemtome that PWM and LED dimming could not increase ANSI while an iris in the lens good.

The Vivitek has an iris in the lens now that probably gives a little bit better ANSI CR than a manual iris in the same spot would if all the way open, but a little worse ANSI CR than a manual iris would if closed down to its minimum. But I doubt that ANSI CR differences would be that great from the iris positions. The on/off CR differences due to iris position are probably much greater.


--Darin
 

·
Registered
Joined
·
23,188 Posts

Quote:
Originally Posted by dcouzin /forum/post/16938043


darinp2 primarily improved the FOFO contrast by crafting a new iris. It did this by hardly reducing the Full On luminance.

Just to be clear, I reduced the light for 100% video level quite a bit with the things I did to get significantly better on/off CR with an Optoma H79.
Quote:
Originally Posted by dcouzin /forum/post/16938043


It is bad optical engineering to use a lens iris to adjust brightness. Better to dim the LEDs.

By this did you mean in a dynamic way? Because using a lens iris to reduce the lumens in a way that improves native on/off can be much better than using the LEDs to reduce the lumens just because of the native on/off CR improvement. Of course, matching it with another iris to improve on/off CR is even better. But there is the point from full open to a little bit closed where a lens iris can improve the on/off CR quite a bit also since getting those last lumens can be pretty expensive from a native on/off CR standpoint. I think InFocus has generally gone with very open lenses to get as many lumens as reasonble until recently, but at the expense of native on/off CR.


--Darin
 
1 - 11 of 11 Posts
Top