AVS Forum banner
Status
Not open for further replies.
1 - 20 of 92 Posts

·
Registered
Joined
·
307 Posts
Discussion Starter · #1 ·
I thought I share the results of my eX(1)periment :)

I found an old 46mm polarized filter I had kicking around and I thought: hmm... this looks awfully like the lens size of my X1.

Now, I' ve read on this forum that some use filters on other PJs, but mostly FL, not PL.

So I held this filter in front of the X1's lens and.....I've got a much dimmer image of course. But wait, it also looked so much better somehow.

First of all the black bars of the unused 4x3 portions of the screen became so much darker, that they are hardly visible even without masking.

I increased the brightness slightly, cranked the contrast up to around 70' pumped up the color to 68 or so, adjusted the color balance as it tends to get to much green boost with the filter and...Voila!

The picture is so much more film-like and 3D looking.

The overall light output has been significantly reduced so the room has to be completely black, but the image looks so much less"digital" and so much more like CRT or film. I guess "analog" is after all much more pleasing to our eyes and ears, isn't it?

I played some LOTR, Driven (great color and excellent dvd transfer BTW), Monsters Inc and Dinasaur. My wife said it looks like IMAX now - so tri-dimentional and life-like! Now how's that for WAF?

But best of all the rainbows have virtually disappeared. I mean 90% less of them. From moving objects, white credits, anywhere.

I just have to figure out how to mount the filter, as the X1 lens has no thread. Try it if you can and see if it does the same for you. For me it was simply astounding.

piesek
 

·
Registered
Joined
·
388 Posts
Good work. This may be a significant discovery affecting the use of business projectors for HT


A polarizing filter rotates on its mounting ring which screws onto a camera lens. With sunlight and a camera it works best in one orientation. It gets rid of reflections coming off water or snow. I don't know how the light coming out of a pj is oriented, so, you may get different results with different orinetations of the filter. Do any any optical experts have an opinion on this?
 

·
Registered
Joined
·
3,534 Posts
I've used a circular polarizer on my LT240. I measured it to be roughly equivalent to a 3x ND filter. As expected, black level dropped too. I wouldn't say the image was "better". Just dimmer.


Rainbows didn't go away for me, but then I'm oversensitive. On an HT1000, I had to drop the contrast down to 40 (from 128) before rainbows completely disappeared. That's roughly the same as going from 100 IRE to 30 IRE on an 84" diagonal image (no filters, 128 contrast)...a huge drop in brightness. Of course, my white wall is >1 gain.
 

·
Registered
Joined
·
307 Posts
Discussion Starter · #4 ·
I noticed a great difference between the image with the filter and just an image with less brightness/contrast.

To reduce the rainbows on the X1 without the filter I had to turn the contrast down to about 40. But than the image was totally lacking "punch" or...contrast. It was washed out and un-defined.

Increasing or decreasing the brightness didn't do anything to rainbows except making the image more or less "muddy" or "foggy"

Now, with the filter I can crank the contrast (white level) up to over 70 and since the brightness (black level) gets reduced, because the image is globally dimmer the actual CR (contrast ratio) seems to increse.

And the raibows disappear! I mean any anbient light ruins the film-like effect completely, but I only watch the movies in the dark anyway.

It is hard to describe, but trust me it is different.

piesek
 

·
Registered
Joined
·
307 Posts
Discussion Starter · #6 ·
I'm not sure if it is a circular polarizer. When looking through the filter it actually changes while rotating it. However turning it while on the lens doesn't seem to make any difference on how the image looks. It looks just as good no matter what's the position of the ring...

The label on the ring reads: 46mm PL filter, it is made in Japan by Optex.

piesek
 

·
Registered
Joined
·
13 Posts
I have the same filter here. I am going to test it with my X1... Have you tried with the white segment on (Presentation mode)?


Pat
 

·
Registered
Joined
·
307 Posts
Discussion Starter · #9 ·
Quote:
Originally posted by patbut
I have the same filter here. I am going to test it with my X1... Have you tried with the white segment on (Presentation mode)?

Pat
Nope, I only tried it with a heavily modified Film preset. (brightness at 55, contrast at seventy something, I don't remember the color settings now, but It needed lots of boost in color and reduction in green gain.

I didn't attempt to eyeball the greyscale yet as I need to attach the filter somehow. Which leads to the next issue:

Quote:
Originally posted by sunam
X1 fit 62mm filter.

I use UV filter for protectiong of lens.
Well, I guess you're right, the outer (focus) ring of the lens assembly seems to be around 60mm, but 46 is almost perfect to fit over the actual lens - just enough clearance not to touch the glass.

I guess If you used a thin stripe of a two-sided tape placed on the filters thread it would stick just nicely to the inside (zoom) plastic ring on the X1.

I have to try it tonight. But if you had a filter to fit over the whole thing it would be easier (and safer)

piesek
 

·
Registered
Joined
·
807 Posts
given that DLP works by selectively moving little mirrors in the DLP array, and that a polarizer's job is to kill reflections (not the whole story) -- I'm not sure that this is a slam dunk ...


I would imagine that rotating the polarizer would have some sweet spot which would trail off as the polarizer is rotated.


I'm not questioning the validity of Piesek's improvements but am suggesting that it might not work the same for each of us w/o finding that sweet spot.


I will try to get a 62mm polarizer and see for myself! -- I just wish I could figure out how this could help.


keep the details coming,


Mark
 

·
Registered
Joined
·
807 Posts
OK, I looked for polarizing filters in an Edmund Scientifics Optics catalog and found a couple of candidates --


mounted and unmounted. the unmounted have only 30% transmisivity, which sounds like too much of a brightness cut.


the mounted ones come w/ an interesting note: 2 can be screwed together to make a variable neutral filter.


So. I'm going to order 2 62 mm mounted filters.


I'll play w/ 1 to see what the polarizer alone does and see if I can replicate piesek's good results.


if not then I'll try that variable neutral filter ...


the one thing I haven't figured out is how to mount the filter -- sunam, how did you do it ?


each mounted filter is 10mm thick so I'm wondering how to keep it in place without resorting to epoxy.


suggestions?


Mark
 

·
Registered
Joined
·
307 Posts
Discussion Starter · #13 ·
Quote:
Originally posted by mpilon
given that DLP works by selectively moving little mirrors in the DLP array, and that a polarizer's job is to kill reflections (not the whole story) -- I'm not sure that this is a slam dunk ...

Mark
Let me make one thing clear here:

I didn't have any comparison between polarizing filter and neutral density filter for example. All I had on hand was a polarizer that I used and it yielded good results. I said that in my first post: the picture is significantly dimmer. Perhaps this is the only advantage of having this filter in front of the lens. However, that still returns a vast improvement in picture quality once the contrast and color gain are increased. The appearance of rainbows is also significantly diminished. (Remember that this is not the same as simply decreasing X1's brightness setting.)

There may be no magic in this, as it's seems obvious that simply a dimmer picture will produce less rainbows.

So in other words there maybe no "sweet spot" or no mysterious quality this filter has. But with the filter on the image IS better. What's better about it? Well, I wanted my picture to look as close to the actual movie experience as possible - and if simply dimming the image is the way - that does for me.

Some people are excited about the bright, "crisp" and ultimately unnatural picture digital PJs or plasma screens produce. It was that hard to describe "glow" in whites and milky blacks that made the image look artificial. With the filter it's much closer to what you would experience with a crt PJs - and they are much dimmer, right?

I wanted to have a "home theatre" and not "home bestbuy showroom" with neon green grass, glowing pink faces and super bright highlights looking like fluorescents. And I NEVER use it with any ambient light.

Try it first if you can before ordering expensive filters- my solution isn't universal and quite subjective so you may be very disappointed or very happy as I am right now. In any case I cannot guarantee it!

piesek
 

·
Registered
Joined
·
1,665 Posts
Well, I too find this potentially VERY interesting! piesek, I do also want to know if your filter is linear or circular polarizer.


What you have observed may actually make sense theoretically, as DenisT eluded. Unlike the LCD projector (in which the LCD panel is sandwiched between two perpendicularly oriented linear polarizers), a DLP projector does not use any polarizer in the optical engine as far as I know. However, the light from the lamp may be partially polarized by being reflected by the DMD micromirrors (just as the reflected sun light is). If so, use of a linear polarizer in an "optimal" orientation (or perhaps even a circular polarizer) might cut "stray light" significantly while mostly passing the light correctly reflected by the DMD, leading to a true enhancement in contrast...


Am I smoking something??? :D
 

·
Registered
Joined
·
3,534 Posts
I measured 30% transmission out of my circular polarizer. It did not reduce rainbows significantly for me, but then my >1 gain wall was probably the culprit. I believe ambient light, screen gain, and projector brightness affect rainbows more than anything inherent in polarizing the image (aside from image brightness and reduced hotspotting).


But then again, I never tried a linear polarizer. :)
 

·
Registered
Joined
·
499 Posts
a few weeks ago some one on here had the same sort of experience w/his L300. I'm wondering if its worth the 30 bucks to try it out. Is this only a DLP thing?
 

·
Registered
Joined
·
4,735 Posts
Quote:
Originally posted by sushi
What you have observed may actually make sense theoretically, as DenisT eluded. Unlike the LCD projector (in which the LCD panel is sandwiched between two perpendicularly oriented linear polarizers), a DLP projector does not use any polarizer in the optical engine as far as I know. However, the light from the lamp may be partially polarized by being reflected by the DMD micromirrors (just as the reflected sun light is). If so, use of a linear polarizer in an "optimal" orientation (or perhaps even a circular polarizer) might cut "stray light" significantly while mostly passing the light correctly reflected by the DMD, leading to a true enhancement in contrast...


Am I smoking something??? :D
I don't think you are going to be able to up CR with these (excluding projector setting modifications). I understand what you are saying about stray light getting polarized differently, but if that happened and could be taken advantage of then the DLP manufacturers would have missed out on a very easy way to up CR.


Bob Williams posted about trying the polarizers on LCDs. He said that the green tends to be oriented 90 degrees offset from red and green, so it doesn't look like the polarizer would help for that technology. I tried one of these on my D-ILA for fun and found the same thing. It threw the color balance off on that machine.


Most probably already know this, but here is one way that a filter can help CR. If your projector is too bright for you and you then turn down the contrast, that will hurt your CR. But using a filter to decrease all light won't affect the CR. Same thing with a screen. If you buy a Hi-Power and decide that it is too bright, a filter is generally a better alternative than decreasing the contrast setting.


--Darin
 

·
Registered
Joined
·
4,735 Posts
Quote:
Originally posted by KingofOld
a few weeks ago some one on here had the same sort of experience w/his L300. I'm wondering if its worth the 30 bucks to try it out. Is this only a DLP thing?
Not if you make sure you get a neutral density filter. The part about rainbows is, but if you feel that your black levels are too high and your images are too bright, then it is worth a shot. Now, if you feel that your black levels are too high and your projector isn't bright enough, then the filter isn't going to help that situation. We only have so much CR to play with on any projector and the filter isn't going to change that. Also, if you like the high bulb setting more than the low bulb setting, then these filters are probably not for you.

Here is a set for $15 (plus shipping) that could help you determine if you like the look with different amounts of filtering. I think the lowest filtering in this pack is 50%, though.


--Darin
 

·
Registered
Joined
·
807 Posts
peisek,


I'm not arguing with you, just trying to reconcile the bits I know with your good resutls. the part that I'm having the hardest time understanding is that rotating the filter doesn't change anything. maybe the orientation of the light after it's reflected off of the dlp array isn't tied to 1 orientation [don't know the correct term].


circular vs. linear: given that reflections dim/brighten as he rotates the filter, I take that to mean it's linear.


I'll pop $36 (or so) for the edmund filter.


sunam, thanks for the picture -- after looking at it I'll definitely try 62mm.


thanks all,


Mark
 

·
Registered
Joined
·
307 Posts
Discussion Starter · #20 ·
Quote:
Originally posted by darinp
I don't think you are going to be able to up CR with these (excluding projector setting modifications). I understand what you are saying about stray light getting polarized differently, but if that happened and could be taken advantage of then the DLP manufacturers would have missed out on a very easy way to up CR.

--Darin
I agree that the "absolute" CR is not higher . But a "relative" or perceived CR definitely is. What do I mean by this? Well let's say to make the image fairly good looking in the film mode and to reduce the rainbows and the overall "glow" as I like to call it I have to turn down contrast (white level) to about 40.

Now the brightness should be set to around 50-53 - anything lower than that and you crush shadow detail, anything higher - you lose the decent black level. Now, with the filter to match the calibrated black level I had to up the brightness (black level to about 56-58 -that's 3-5 clicks only. Yet I was able to boost the contrast (white level) to about 70 (!) and I still saw MUCH LESS rainbows than before. The blacks are MUCH deeper and the whites are about equal in output.

So I'm sure the "distance" between the darkest blacks and the brightest whites of any image has infact increased, eventhough the picture could be described as "dimmer". I'm not sure if this makes much sense, but it works for me. Also check this quote from KBK:
Quote:
Originally posted by KBK
dimmer is the trick they use to kill rainbows. But.. what I have found, is when increasing CR alone.. that rainbows dropped in noticability.
and this one:
Quote:
Originally posted by KBK



Why do I say that Rainbows go DOWN with increased CR? Becasue I have always achieved higher CR by bringing black level down.. Not by cranking output. So output remains the same, But....black level gets stupid.

Oh, and btw I also have much less visible mosquito noise, which drops significantly already when the white segment is off going into film mode.This enhances it even more - no crawlies anymore.

piesek
 
1 - 20 of 92 Posts
Status
Not open for further replies.
Top