Mechanical dynamic iris designs will be improved and always add to a CR value no matter how high or low it is in native form. Problem is that they are not as fast as they could be - make noise - and wear out over time.

Enter the 'fluid iris' which is a transparent and independant section of the lens assembly containing a viscous conductor fluid which when fed an electrical charge can darken 'instantly' blocking light flow. Nice eh?

The change to black from it's transparent form in speed over a mechanical iris is over 400% faster.

The iris has 10 rings ( which are invisible to the naked eye ) which are like tree rings eminating from the center of the lens assembly and can be darkened from the inside out or the outside in.

Heres the part that really caught my attention:

This technology is already underway for release in the commercial digital cinema market. Benefits?

-The entire lens can be designed as an automatic ND filter adjusting itself frame by frame. This would produce an almost unlimited grey scale paremeter which would put everything designed today back to the stone ages.

-The different levels of light blockage depend on the 1-1000 different grey levels.

Would this technology be able to paddle down to the home theater market one day?

This is a technological marvel to produce massive light out and high CR at the same time for commercial digital theaters.

Lots of very technical forum members who could chime in with there opinions on how home projectors could benefit?

Yeah that does sound cool. I wonder how opaque the current material is. I have seen something similar with window shading several years ago.

This technology is already underway for release in the commercial digital cinema market. Benefits?Sounds pretty cool. I've mentioned before that I think the commercial digital cinema market could benefit from something that home theater would have a problem with. That is, digital cinema plays the same stuff over and over again and they actually play few things per year (one projector may only show a couple dozen or less movies per year). So, those iris positions could be pre-encoded by a professional. That would allow being more aggressive with the darkening, while at the same time avoiding problem situations, since the professional could choose not to close the iris.

It is kind of like look-a-ahead except even better since a professional could make a choice. In the above case ads and other content could just have the dynamic system disabled, but since lights are usually on during ads and on/off CR isn't critical there, that shouldn't be a problem.

Digital cinema does have the issue of exit signs and other lighting even during the movie, but even given those things I think they could benefit from going much higher than 2k:1 on/off CR from the projector.

Home theater projectors could benefit from the above, but I think 2 panel or 4 panel projectors could do even more as far as CR. Might lose some light from the extra panel, but could benefit from not having to either dim or brighten the whole panel (so bright in one part of the image could be done at the same time as very dim in another part).

--Darin

This is interesting news to me but why bother with this if could control light at the source itself.

This will happen with LEDs + dmd combo in the not too distant future. Bob Wiliams replied to a question of mine saying LEDs could replace the function of the DI and intelligent lamp control. The downside according to him was that LEDs could not yet provide enough brightness and that they are still expensive.

In the digital cinema market I see the active fluid as having a stronger position but I would suggest a control at the light source even here. If lasers can not be made powerful enough for digital cinema the active fluid might be the way to go. However, I hope those rings really are invisible.

This is still not the same as real native contrast and will inevitably bring whites down for some dark scenes or even crush them.

Interesting.

I wonder if there are optical side effects like diffraction at the edges of the rings.

Also with a superfast voice coil actuated DI like Planar has developed I wonder how much better this could really be.

"why bother with this if could control light at the source itself. "

A smaller iris increases CR.

noah katz
Yes, a more closed iris would increase the native contrast to some degree. However the really efficient systems seem to be only the dual iris systems with one iris before the dmd and one after the dmd.
It will amaze me this magical fluid can take the abuse of the massive light output of a digital cinema projector.

I would like to find out the chemical nature of the fluid. I have no idea at this point. But there are individuals who I can ask.

No I am not as smart as them.:D

Mechanical dynamic iris designs will be improved and always add to a CR value no matter how high or low it is in native form. Problem is that they are not as fast as they could be - make noise - and wear out over time.

Enter the 'fluid iris' which is a transparent and independant section of the lens assembly containing a viscous conductor fluid which when fed an electrical charge can darken 'instantly' blocking light flow. Nice eh?

The change to black from it's transparent form in speed over a mechanical iris is over 400% faster.

The iris has 10 rings ( which are invisible to the naked eye ) which are like tree rings eminating from the center of the lens assembly and can be darkened from the inside out or the outside in.

Heres the part that really caught my attention:

This technology is already underway for release in the commercial digital cinema market. Benefits?

-The entire lens can be designed as an automatic ND filter adjusting itself frame by frame. This would produce an almost unlimited grey scale paremeter which would put everything designed today back to the stone ages.

-The different levels of light blockage depend on the 1-1000 different grey levels.

Would this technology be able to paddle down to the home theater market one day?

This is a technological marvel to produce massive light out and high CR at the same time for commercial digital theaters.

Lots of very technical forum members who could chime in with there opinions on how home projectors could benefit?Where can I get this for my projector

This is still not the same as real native contrast and will inevitably bring whites down for some dark scenes or even crush them.

That's what I was thinking too.

I would like to find out the chemical nature of the fluid. I have no idea at this point. But there are individuals who I can ask.

No I am not as smart as them.:D

I could be totally off base....

But from the descriptions above it sounds like it's an LCD panel that is organized in rings instead of in pixels. So as you turn on more and more of the outer rings the "iris" get smaller and less light gets through.

Then again I'm probably totally wrong about all that.

Mac: Not sure but you actually have a great idea believe it or not.:)

But from the descriptions above it sounds like it's an LCD panel that is organized in rings instead of in pixels. So as you turn on more and more of the outer rings the "iris" get smaller and less light gets through.I was wondering that too, but I think that would mean having to polarize the light to get it to work well and one reason 3 chip DLPs are so efficient (bright) is because they don't have to polarize the light and lose some in the process. I wonder if there is another material that will block light well when charged one way, but let it through when not, that doesn't rely on polarization.

--Darin

Somehow I'm not too crazy about the idea.

The simpler the light path is and the less amount of "junk" we put in it, the better.

Keep it simple, remove the band-aids, and do it right (0 lumens at 0IRE at the lightsource).

Keep it simple, remove the band-aids, and do it right (0 lumens at 0IRE at the lightsource).0 lumens at 0IRE at the lightsource isn't simple for projection, given that part of the image can be video black (0IRE) and part of it can be other things. Having the lightsource go to zero if the whole screen is video black, but something higher if there are any pixels that are non-video black isn't simple. It is complicated because now you have to handle either clipping or pulsing, much like a dynamic iris. I like the idea of things like LEDs that can go off, but they don't magically make everything simple. If something like this liquid iris is a band-aid, then I would say so is using an LED lightsource like a dynamic iris. Probably a band-aid I'll want because I care most about the end result and not making it simple.

How do you propose that they get to 0 lumens at 0IRE for huge screens (like 30' wide+)?

--Darin

Darin, my main objection to "tricking out" the lens has to do with the difficulty and high cost to have a simple, high-quality, optical lens. As a photography hobbyist, I know how expensive a good lens can be...

Add zoom, (cost) large aperture (more cost) to be able to transmit lots of light and now we want to add stuff (LCD rings?) to this in order to constrain the light transmission frame by frame.

My point is that this will soon be available using pulsed LEDs as we have discussed before.
I'd much rather have that, that anything that would "screw" with the optics.

As far as lighting up huge screens, you're right, there's nothing in the near future.
Maybe a "light-farm" of LED arrays will be bright enough within the next few years.

My feeling is that if "fluid dynamic irises" see the light of day (pardon the pun), Darin will be the 1st one to point out their weaknesses and the newly discovered artifacts that they cause... ;)

My point is that this will soon be available using pulsed LEDs as we have discussed before.
I'd much rather have that, that anything that would "screw" with the optics.
Unless these LEDs provide very large native CRs (large enough that they can blackout and not have visible artifacts), then I still want manual irises (one before the DMD and one after for DLPs) in order to get higher native and dynamic on/offs with dimmer images for some material, and brighter when I want that (like 6-12 ft-lamberts for movies with the lights out and 20-40 ft-lamberts for some video games, lights on viewing, or other cases where I want much brighter images).

I doubt individual LEDs are going to provide large enough native CRs with DMDs that closed down irises wouldn't still be useful to improve the native CRs.

And I'm sure I will discuss both the positives and negatives that I see different techniques bringing. If these fluid irises refract the light in ways that cause misconvergence on the screen (like chromatic aberration) then solid iris pieces that move may be better suited for combining with LEDs.

--Darin

I wonder what artifacts the pulsed LEDs will introduce?

Darin,

"Unless these LEDs provide very large native CRs ..."

Not sure exactly what you mean, but don't LED's instantly extinguish to essentially zero?

But they probably don't even need to; similar to panels in series, the net CR will be that of the LED X that of the panel.

Darin,

"Unless these LEDs provide very large native CRs ..."

Not sure exactly what you mean, but don't LED's instantly extinguish to essentially zero?Probably, but so can a dynamic iris. But you can't close a dynamic iris all the way for blackouts without bad side effects unless you have enough native CR to keep things like bad fluctuations in the images from being annoying. Like from going from a blackout to black with a small amount of white. From the Planar implementation I would say 3x the native CR can be done with a good implementation with very little visible site effect. I think Sony may be closer to 4 or 5x the native CR, but beyond that I think it gets very difficult to provide improvement without too many problems, unless things are done in zones (which neither dynamic irises nor single LEDs provide).
But they probably don't even need to; similar to panels in series, the net CR will be that of the LED X that of the panel.If the LEDs light up the whole panel then it is like a 1 pixel panel in series with the main panel, so not all that much different than a dynamic iris as far as the multiplier from native CR that is reasonable to expect with good algorithms. If there were hundreds or thousands of LEDs with overlapping zones, then that would be more like two panels in series, where they aren't 1:1 mapped, but were there is still some resolution to the first panel. I doubt this would work that well without quite a few zones. Not sure how many that is, but somebody said that the Samsung flat panels with something like 128 LEDs behind didn't have enough zones. The Brightside stuff with almost 2000 LEDs behind seemed to have enough zones from what I saw.

BTW: I may not have been clear. I meant native CR of the projector before using the LEDs in a dynamic way (kind of like a dynamic iris).

--Darin

There have been hints that LEDs somehow can be used to illuminate zones of the dmds but I have never understood how that can be done given the size of a LED campared to the size of the dmd. If that is possible there is great potential and if not LEDs can function DI style "1 pixel in series".

Perhaps if dmd/leds are fast enough you could vary the led brightness while sequentially building the image from the dmd. Instead of just sequentially building by color like is currently done, you could sequentially build a single color by brightness and then continue on to the next color.

There have been hints that LEDs somehow can be used to illuminate zones of the dmds but I have never understood how that can be done given the size of a LED campared to the size of the dmd. If that is possible there is great potential and if not LEDs can function DI style "1 pixel in series".

How about an array of LEDs focused on the DMD by the use of a lens? Then it doesn't matter if the LEDs are bigger than the DMD.

How about an array of LEDs focused on the DMD by the use of a lens? Then it doesn't matter if the LEDs are bigger than the DMD.

For a 1080p that's either ~2 million LEDs (white) or ~6 million LEDs (RGB), if there is a 1:1 ratio between pixels and LEDs. This is probably not feasible with current tech for home applications.
Aligning LEDs pixels to pixels would be an increasingly larger problem as you approach 1:1, but probably something along the lines of the Samsung 81 series LCDs with 'local LED dimming' might be feasible as it had 96 zones.

A couple of thousand of LEDs would probably be enough to make the design expensive and complicated.


.

For a 1080p that's either ~2 million LEDs (white) or ~6 million LEDs (RGB), if there is a 1:1 ratio between pixels and LEDs. This is probably not feasible with current tech for home applications.
Aligning LEDs pixels to pixels would be an increasingly larger problem as you approach 1:1, but probably something along the lines of the Samsung 81 series LCDs with 'local LED dimming' might be feasible as it had 96 zones.

I have no idea how they would implement it. Just answering Ohlsons question about the size of the LED compared to the DMD.

By the way. If you have a whole 1920x1080 array of LEDs then you wouldn't need the DMD anymore. Just put a lens in front of the array and project on the screen.

I'm not so sure you would gain that much by local dimming "Samsung style" either. I believe much of the degradation in intrascene contrast is caused by reflections in the optics and the DMD, which you still would have when using local dimming.

A couple of thousand of LEDs would probably be enough to make the design expensive and complicated. .

I wouldn't argue with that. :cool: