OK... recently I have come to the conclusion that an overlooked part of our quest for higher lumenage is the differing transmissisivity of our LCDs... a quick example: SupraGuy has measured his at 8.5%... I've measured mine at 4.6% .... which means all else being equal, his lcd will make a pj 85% brighter than mine..sigh
I'm not sure the best way to measure this though... not sure if collimated light will make a difference, LCD powered on or off, if on, just a white screen? measure 2-3 spots before and after and average? any ideas are welcome.... so..to get it started:
-------------------------------------------------------------------------------------
Mikyd
LCD: DAEWOO f227b 17", 450:1 contrast
Panel: HT17E12-200 which I think is actually Hyundai, though I could be wrong
LCD powered off, measured in center of lcd, back of sensor aginst LCD before,sensor touching LCD after

Lux Before LCD:3300
Lux After LCD: 154
LCD % = 4.6%
-------------------------------------------------------------------------------------
Rox
LCD Hami 8"
LCD powered off, using the sun as a light source, sensor touching LCD

LCD % = 5.5%
---------------------------------------------------------------------------------------
SupraGuy:
CMV529A = 8.5% (non-collimated,non-directed light source)
LG1750 = 5.5% approximately

Samsung 740N(17"):
taken at the projection screen:
Lcd in pj: 430 lumens
LCD removed:5000 lumens = 8.5%
We have our current winner... both in 15" and 17" varieties...can this guy pick a winner or what... got any lottery numbers for us?
---------------------------------------------------------------------------------------
jonjandran:
15.4" Wuxga.
Antiglare stripped.
Rear Tac and reflective layers stripped.
T15 pulse strike Ushio.
Pro Reflector.
Reflector/Light/220mm/650mm/Lcd/triplet
middle of LD, on x100
Lamp side with back of meter touching Lcd : 913
Triplet side with front of meter touching Lcd : 98
= 10.7 %
the new winner....
---------------------------------------------------------------------------------------
jonjandran:
Benq FP-567S
Antiglare and rear Tac was stripped off.
Sunlight test. Meter at x100
Before Lcd / against glass : 320
After Lcd / against glass : 25
Conclusion : 7.8%
--------------------------------------------------------------------------------------
Simul8r:
Cornea 15"
With LCD off and unplugged
50 58 49
61 75 60
53 60 50
Average 57.33 = 186.32 ANSI
With LCD panel removed and just 317mm fresnels in place with same area projection
889.6 / 970.8 / 809
1103.6/ 1302.4 / 983.6
1009.8/ 1095.4 / 902.8
Average 1007.44 = 3274.18 ANSI
Divide 186.32 by 3274.18 = 5.7%
---------------------------------------------------------------------------------------
MichaelJ:
Hami 8" with AG, sensor touching lcd surface
Average over 3 readings gives me 6.2%
Specs are 300:1
250Cd/m²
and the backlight draws 7w
---------------------------------------------------------------------------------------
ElbeGhast:
Gateway FPD 1500 w/AG removed 16.28%
15.1" TFT,200:1 contrast,250 cd/m2

---------------------------------------------------------------------------------------
Durachko:
Samsung SyncMaster 710n ,powered off
4.2%
--------------------------------------------------------------------------------------
Meyer2:
Sun Test: BenQ 567 6.49%, Samsung 740B 6.04%
SupraTest: ie. measuring on-screen ansi lumens with LCD both out and in projector but with power to LCD off.
BenQ 567 5.90%, Samsung 740B 4.40%
the samsung puzzle deepens? theoretically the 740b is the same as the 740N exxcept it has a dvi connection...
-----------------------------------------------------------------------------------------
ChuckL:
BenQ 567SV2, AG removed 6.7% sun test
...
Just removed rear tac and redid the transmission readings in a controled lighting enviroment. 6.8 % before 7.5% after tac removal. 500 w halogen as lighting source 2 feet from panel. 5 measurments each averaged in it's frame and painted.Ambiant blocked by cardboard box.

-----------------------------------------------------------------------------------------
phutton: CMV 520D 15" AG on, halogen bulb
6-6.5%
--------------------------------------------------------------------------------------------
gadgetsmith:
Sharp LQ154M1LW12 panel completely unmolested...

AT SCREEN TEST
LCD in PJ 'on' (white screen using PGEN)
ave lux = 15.7 (-33% lower compared to 15" CMV 520D w/ replacement polarizer, which averaged 23.6 lux on last reading before switching LCD's)
LCD out of PJ
ave lux = 490
LCD transmissivity = 15.7/490 = 3.2% ! YIKES... (this is 37% lower than the measured 5.1% transmissivity of the CMV panel done in the same manner)
SUNLIGHT TEST
Light meter tripod mounted.
direct sun reading = 107 klux
reading from a/g side of panel = 6.1 klux (transmissivity = 6.1/107 = 5.7%)
reading from shiny polarizer side = 6.8 klux (transmissivity = 6.8/107 = 6.4%)
----(mikyd-) after ag removal and other film removal - same results..... I think this is the same panel as jonjandran? boy this gets frustrating and confusing...especially the way different results with jonjandran AND the different results with the suntest... but could that be the elken effect?(different spectrums transmitting different percnetages?)
----------------------------------------------------------------------------------------------
miedosracing:

Balance CM2015 (walmart) NOTE:Not the CMV2015....
8.3%
--------------------------------------------------------------------------------------------------

My thoughts right now are that the most meaningful measurement would be:
lcd powered on, white screen displayed on LCD and using collimated light, the latter if for no other reason than that the thickness of the sensor will not matter then since the light won't be diverging/converging...

I measured 5.5% on my hami 8".

I did the test with the sun on a suny day, just put the lcd straight to the rays and meassured after and before. luxmeter in contact to the lcd.

If you are doing it with a closer light source... the difusivenes (light point source is difusive) will change the meassurement... I think the sun is best way.

by the way, I found once on google search the generic LCD are 6.6% trasmisive.

true, its probably more accurate, but I think what is most relevant for our purposes is measuring it as it will be used (collimated light, powered on, white screen displayed on lcd)... but it may not make that big of a difference, we'll have to do some testing.... and that 6.6% is why I think we need to actually measure some lcds..... mine being a good example... since the "generic" lcd will be 43% brighter than mine, a not inconsiderable amount ..so I'd tell people, assuming my measurements hold out, NOT to buy the same LCD as mine....

I agree with Rox on this one. The Sun provides an infinite source of perfectly (for our needs) collimated light. I think his test was the best test a person could come up with unless of course it is a cloudy day.

again, I agree in principle, but I will see if there is indeed no difference between transmission with the lcd powered on with a white screen displayed and lcd powered off .... well, unless you want to take the whole thing outside and test it, and seeing as its about 20 degrees F here... I still think I will test inside
it may be that there is no appreciable difference....

lucky agrees with the "sun" thing, about the lcd turned on with white or off is not going to change anything IF the setup is at hig contrast and hi brightness values...

the average until know is [8,8% (supra) + 4.6% (mikyd) + 5.5% (Rox)]/3= 6.3%

new data is welcome

Oh i hope someone with a WUXGA panel takes you up on that Rox!!

Well actually i might be able to add to this data myself... I just bought a Lux meter.

sounds excellent and I know a lot of people are anxious to hear about that monitor..hope its good news I'll add the measurements to the first post when ya got 'em....
just wired up my new bulb(moving to the big 400 watts and ran the same test(can't do the collimated light ..I mis measured my wood for the frames..well...actually I bought 1x2 and I knew the 1 was 3/4 but I thought the 2 would be 2 not 1.5..ah well) at least I will be able to tell if collimated makes a difference tomorrow... anyway, lets see if I get the same result:
5400 lux at 1 meter before LCD, unplugged
270 after = 5%

well ... the meter is +- 5% and the result is higher(good thing) but still within 10% of my previous... I'll box in the light path this week and get the fresnel up and get a little more controlled conditions and see how it goes, but I would say I have a definitley below average lcd...
I'll add supras measurements to the first post when I find his specific info... he is still the leader!

My current light path is:
bulb|hot-mirror|coll fres|replacement 3dlens polar|LCD|field fres|standard LL triplet


Lamp: Osram HQIBT400WD 32000Lumens, 5200K, CRI group 1a - Ra 90-100, 25mm arc
(Note: I have a hot-mirror, but no dividing glass/polycarb)
Magview 17" M17 antiglare/polar stripped

the following readings were taken along the lightpath centreline (a couple of weeks ago), no averages.

light side of coll fres 100500 lux
after coll fres/polar 34000 lux
after lcd 5000 lux

which works out to about 4.7% transmission through the coll fres/polar/lcd, or 14.7% just through the stripped LCD

(I pretty much finsihed the PJ enclosure over the weekend, but when I put it all back together i only had 65 Lux in the centre (I promptly stopped taking readings..) gotta go and find out who stole my lumens.... )

14,7%.... mmm

I wonder if the sun test will make such a huge diference...

am I asking to much when I ask; could you repeat it at sun test?

No wonder Supra's hits 200 lumens easy!! Just checked my measurements and I guess I must be the lowest
5 readings before and after LCD (unplugged) using the lamp as the light source and removing all fresnels.

Before: 1789.8 x10 (precondenser usage + S400dd vertical to panel)
After: 549.2 x10
3% transmission

15" Cornea CT1503
500:1 contrast
Antiglare from front and rear TAC removed

Someone help me do the math here if it's possible. If I'm currently projecting 185.69 ansi right now averaging 59.9 on 3.10 sq/mtr screen with this panel, then how much more would I be putting out if I had percentages like Supra's panel? My real rough calculation came up to 450 ansi.

Well that 14.7% is for an LCD with one polar missing........

I wonder what the percentage is for

Replacement polar + LCD ?

This would be a better comparison, since everyone else is accounting for losses through two polars!!

Just did a google on tranparency of LCD's. Check the second page, continuing paragraph. They state LCD panels are between 3 and 8%.
http://www.cs.yorku.ca/~wolfgang/papers/hdrdisplays.pdf

I'll have it apart again this weekend for final housing finishing so I can probably arrange that, although the weather for this weekend isn't supposed to be favourable...

not to depress you further, but if supra has 8.5% then you'd have 524 lumens at the same... but the good news is I feel better .... I'm going to make my lcd sled tonight and try the test powered on with a white screen and see if that makes a difference, I don't believe it will theoretically but you never know... however , on the bright side sim, then you have really achieved something with your precondenser.... if your readings bear out, then someone with an average trans. lcd can get well over 350 lumens duplicating your setup!! I don't have a ladle, didn't you have like 150 without the ladle, just the pro reflector? means I may get 225 or so(4x what I had with my 250 watt system) ...should know by the end of the weekend...
But this is exactly why I thought this would be an important thread.... you had your sensor touching the back and front of the lcd? can you check it with the fresnels in place? , you are way ahead of where you should be according to Irongeckos calculator! its a tribute to your methods this makes the precondensor even more important than we thought... imagine if, indeed, you really have a 3% lcd(lets see how the testing goes, we may discover some oddities in the measurements) then you have shown a way to the 500 lumen DIY pj!

well, so far thats almost exactly the range we've shown so far! armed with some more testing you may have made a huge advance in DIY pjs with the precondenser/vertical burn etc...

Depress? Yes, but gratified that we now know why we aren't getting near average amount of lumens compared to each other. I'm squeezing every bit of light out of this precondenser tests and still couldn't figure out why I haven't come even close to Supra's measurements when he was using the standard triplet before he went pro.

The sensor was touching both behind and in front of the panel, 5 readings on center each side. Here are the measurements you requested.

LCD on but dettached from my computer w/fresnels in place both 317mm's, average from 5 readings.
Front of lamp: 1797.4
After collector fresnel: 580.2

LCD off with fresnels in place, average with new 5 readings.
Front of lamp: 1805.08
After collector fresnel: 577.2

Looks like I'm in the market for a better panel
sim

Edit: Rox just corrected me on my math. So I'll hold back on the final percentages until I do the sun test. 30% is just not conceivable.

this is a verygood idea to average 5 readings. I know the luxmeter is very angle dependent....

the 3-8% typical trasmisivity (average 5.5%) agrees with my hami and the generuc 6.6% trasmisivity i found once as well. The 14%... and such values are something sounds very strange to me yet...

more trasmisivity at expense of contrast is something that is still not proved too... don´t know. We will end buying the same lcs supra has finally

Added Supras numbers and panels to the first post.... try not to drive the prices up on them too much

Ooooh! That might explain why my projection is good enough to watch with the lights on and others are constantly complaining about theirs. For the longest time I thought that all of the complainers were simply making bad builds or had their expectations too high, or had bad eyesight. I am using a CMV 529A panel also. According to Supra's numbers I am transmitting close to 2 or 3 times as much light through the lcd.

I can only imagine if my brightness were cut in half or even by two-thirds. I would be incensed, too.

It all is starting to make sense, now.

This will probably be one of the most important threads in this community since the lcd transmitivity seems to be becoming the largest factor in getting a good projection.

Boy, did I get lucky with that panel, considering I purchased it before we were aware that there was such a large difference in lcd transmission.

So the panels to buy so far are:

CMV-529A
Samsung 740N

Someone might want to try measuring a CMV-520D. They seem very similar to the 529A panels.

really, I've wondered ever since the issue came up in the AG thread if this isn;t the reason some people are so happy with their pj and others are so dissatisfied, but it was easy to get distracted by all the other developments, now with a AG free panel,a simul8r style lightbox/precondenser (gee, he IS a troublemaker isn't he) and the right panel, I think we will see the day soon that a 500 lumen DIY pj is a standard and thems REAL lumens not some commercial "marketing"... what I;m interested in is one of the old standbys like the,,oh heck, the LCD that was so popular for so long..chei mei ax150 or something? and wondering if it is one of the high percentage lcds....

This may be a stupid question but would the brightness spec. of an LCD give a hint as to it's transmissivity? I know it's probably only a measurement of the backlight but I took a quick look at the brightness measurements for the LCDs in this thread.

CMV - 8.5% - 450 cd/m2
Samsung - 8.5% - 300 cd/m2
Daewoo - 4.6% - 250 cd/m2
LG - 5.5% - 250 cd/m2
Hami - 5.5% - 250 cd/m2

Maybe there is a slight correlation?

hard to say, it could be afactor, but I would think unless they all use essentially the same backlight I would say its a vague correlation... but its something to keep an eye on anyway...

Since I still have the backlights for the LG and the Samsung, I can compare them. They both seem to be the same basic deisgn, so there are some things that I could try to compare them.

Oh, and for the Canadians, Future shop has a 15" LCD, branded "Metro" that is now residing in my 15" build, and seems pretty bright. I will try to get some luxmeter readings to see what it's hitting, but my personal guess is around the 180 lumen mark. (I have no idea what the nominal specs are for that LCD, though.)

Good question. Rather than waiting on others to measure their panels and getting a list, is there some way we can just distinguish by looking at the specs of the monitor? Figure this out and we are gold!

15.4" Wuxga.
Antiglare stripped.
Rear Tac and reflective layers stripped.
T15 pulse strike Ushio.
Pro Reflector.
Reflector/Light/220mm/650mm/Lcd/triplet = My setup

This is with my meter at x100

Middle of Lcd.
Lamp side with back of meter touching Lcd : 913
Triplet side with front of meter touching Lcd : 98

So I'm getting 10.7 %

Here is my attempt at explaining the factors affecting panel transmittance:

http://www.lumenlab.com/forums/index.php?s...indpost&p=98562

I think I would look to the dot pitch of these monitors and see if there is a notable relationship.

Mark.

well, in general, there are basically only 2 dot pitches that I know fo for most(non-laptop) lcds, .264 for 17" and .297 for 15" aren't there?

Tested a Benq FP-567S
Antiglare and rear Tac was stripped off.


Sunlight test. Meter at x100

Before Lcd / against glass : 320
After Lcd / against glass : 25
Conclusion : 7.8%

I was thinking that also. We would have to assume (big assumption there) that the backlight technology is basically standard. In that case the brightness spec woul probably be a reflection of lcd transmission. Take a look at this. The average brightness spec for the CMV and Samsung is 375 cd/m2. the average brightness spec for the LG and Hami is 250 cd/m2. Now let's take the ratio of those and compare it with the ratio of the measured transmissivity. Brightness ratio: 375:250 = 1.5, Transmissivity ratio: 8.5:5.5=1.5.

Coincidence? Maybe CMV was a little liberal with their numbers and Samsung was conservative with theirs (maybe Samsung was more conscientious of reputation than CMV), but they seem to indicate that the lcd specs may provide a clue as to its transmissivity.

Rox, what if I removed the LCD and take measurements on screen with both fresnels in place? Would this still be feasible determining transmission with and without?

I have to agree.

That's what I'm shooting for with my next build. I gotta get at least a cheap luxmeter.

Your making sense here Mark. I can see pitch might be also a factor in this.

That will work perfect in my opinion. The additional glass between the lcd and screen will affect the readings somewhat, but by the same proportion.

As an example, if 10,000s lumen leaves the collimation fresnal then without the lcd the condensing fresnal and triplet reduced that by 20% you should read 800 lumens at the screen. Then if you place the lcd in there and it has 10% transmissivity, the amount of light leaving the lcd should be 100 lumens. After that passes the condensing fresnal and triplet it would drop to 80 lumens.

the ratio would be the same:

10,000/100=10%
800/80=10%

So, if you want to just measure lcd transmissivity then it probably would work to just measure the brightness at the screen with and without the lcd in the projector.

Sounds great to me. Accept that depending on how collimated your light engine's output is, you could get different results. So someone who has managed to collimate a bit better or worse could get a different reading.This page: http://www.epinions.com/Monitors--reviews-..._pitch__search_ gives a pretty good idea of the spread for 17" monitors, so you are right- the monitors we use are probably typically the same in dot pitch. There is a spread there though, so it would still be good to know for sure what people's dot pitches are.Unfortunately there are numerous brightness enhancement methods that can be employed that do not have a useful effect in our light engines. A reflective polarizer, for instance, can boost brightness in a monitor assembly by up to around 60%, while it will most likely reduce brightness in our setups. Prismatic filters will cause a similar boost (around 40%) but these will also have no positive effect (will eliminate transmittance completely) in a collimated setup.

It is impossible to tell if these sorts of films are in use in a display without disassembly.

I was reading more at this site I previously posted when I came across section 3.1 Local Contrast Perception. From what I understand 300:1 contrast is a optimately working area for LCD projectors compared to DLP or HDR which they can work with higher contrast

http://www.cs.yorku.ca/~wolfgang/papers/hdrdisplays.pdf

Do any of you think that transmissiveness is highly dependent on the ratio of our polars. It might be interesting to find out what each of our panels are at along with Lumen measurements.

Can someone read the section I mention and decipher if what I read is correct?

Ok, here's the full test using my projector.

With LCD on and projecting 3.25 sq/mtr (80"x63")
50 58 48
60 75 59
52 60 50
Average 56.89 = 184.89 ANSI

With LCD off and unplugged with same area projection
50 58 49
61 75 60
53 60 50
Average 57.33 = 186.32 ANSI

With LCD panel removed and just 317mm fresnels in place with same area projection
889.6 / 970.8 / 809
1103.6/ 1302.4 / 983.6
1009.8/ 1095.4 / 902.8
Average 1007.44 = 3274.18 ANSI

Divide 186.32 by 3274.18 = 5.7%

Not bad, but still I wouldn't mind having 8.5%.

veri nice test.

mm, about fresnell+lcd test... I think it is very dependent to where the luxmeter is at. As well as the lamp placement... mmm, the sun test, or LCD removing from the projector is very welcome in my opinion.

edit;
I wouyld say that if you had tried to toogle the lamp positioning on the no LCD setup... you could have achieved higer lumens on screen... this doesn´t mean a less trasmisive lcd though.

nlike rox said, nice test! but I still wish (for my sake) you had 3%, means the numbers would be better with a more transmissive panel .... though you're still about 40% behind supra!

was this with the panel in the pj? could you do the sun test? I wonder if your numbers are so high because of the smaller dot pitch? this is the best by far! is that really 9800 lux after the lcd IN the projector? whats your lumens on the screen? I may have to cough up the big bucks for this panel(though I've been considering it anyway)

Here’s something I found at diyaudio about an LCD and how they measure it.
http://www.bonsaijogi.de/Community/Downloa...6.4-LQ64SP1.pdf

DJ

I could try a sun test in a few days. It's a pain taking down my projestor and removing stuff.

Yes 9800 lux in the projector on the screen, with the average around 8800 at different spots on the Lcd.

Lumens on the screen is 250 at 175" Diagonal / 22 feet from Projector to screen. I am going to rebuild with larger fresnels to get rid of dim corners. Right now my fresnels are smaller than my Lcd which is hurting my corner and side lumens.

how bright is your picture? (on a personal level I mean...thats one huge projection!, guess I'll check out your plog) can you post 9 point readings at the lux/lumens measurement thread(you can use either supras test pic or mine in the lux measurement thread)? I know a lot of people are interested in this particular setup since its as state of the art as you can get..thats the mionten group buy lcd/controller?

I don't want to clutter up this thread, check out my Plog for most answers and post there if you want to.

Hey jonjandran thanks for doing that! Now i can estimate the brightness of my next build.

Actually, considering that for a laptop, the backlight is a signifigant use of power, it makes sense for them to design a laptop panel to be as transmissive as possible. This allows them to save battery power. This is likely why the WUXGA panel shows as being more transmissive.

Never thought about the laptop LCD and baterry power... makes good sense... and sounds like music to my ears . Hopefully it's something I can look forward to.