[jbelljbell]

 http://www.lagom.nl/lcd-test/contrast_ratio.php


I'm assuming this would be an acceptable way to measure contrast ratio of a projector/screen/ambient light combination.


Is there an equally simple method of comparing screen gain, from a known to an unknown screen using a digital camera? (sounds reasonable,right?) Scroll to the bottom of the page where it describes luminance accuracy. (maybe even just comparing luminance of the white photo's of a known, vs unknown screen would give a %gain)


It would be nice for any DIY'er to be able compare their effort with what the 'standard' is for the application they chose. In particular, it would be nice to know what to expect compared either to TWH, or a standard 1.0 gain white tripod screen that many may be able to borrow from a local school, church or other institution.


It would also be nice to have a sticky thread for those who choose to measure, to post their results. Not as a 'mine is better than yours' but so that there is a pool of measured screens, and a new effort can be measured and see 'if it's in the ballpark' or if something is terribly wrong. As long as all measurements are referenced back to a known standard (like TWH) then I think the results would be meaningful.


BTW, it returned very believable results on my LCD monitor rated at 250cd/m2 and 700:1 contrast considering I don't have it turned up full bright, and there was some ambient light in the room....

 

[K-Spaz]

I tried using this to check my DLP Projector and I think I've gotten some erroneous numbers. I got two values of 34xx and 96xx using two different camera apertures. First, I tried the recommended f2.8 and then after that yielded some error messages I tried F8. Either way, I got warnings about under/over exposure and so I'm guessing that while it did give me a number for what it found from the photos I gave it, it admits those pics were no good.


I tried using the exact numbers it suggested but since I have a DLP, I was getting a big blob of an incorrect color (red/blue...) The white pics were being shot at way too fast a shutter speed for a DLP. If this were a LCD, i think it would work out as white. When it suggested a 26.x sec exposure on the black screen, I got a white pic cause it burned it. For all I know, that image may well have been the correct one to use but it sure looked whiter than the white image was with the suggested settings for white.


I'll give a couple more tries.

 

[K-Spaz]

Well, that took a few tries but I finally got rid of most of the warnings. It still is complaining about my exposure time on the black image, so perhaps that would lower these numbers. But at least it didn't just outright say the exposure is too low. The 1/3 additional stops probably would not make a hill of beans on the measurement.


My Pj is rated for 2500:1 contrast, and I shot these in low power mode, with a gamma correction option enabled, and with a slight ambient light. Still the numbers seem pretty close to what I should be expecting using a latex painted black out cloth screen. Not perhaps what they could be if I was using a higher performance screen.


I finally had to switch lenses for this project cause the site just wasn't willing to accept pics from my 50mm F1.4. The only lens it wanted a pic from was the kitlens that came wtih my body when I got it and it appears that the lens has all sorts of edge issues cause when shooting with the 50, the pictures had no hot spot in the middle of the photo. They looked very nice actually. Course they should, that's a lot nicer lens.


I could easily buy these numbers. I'd be happier if it could take values from that checkerboard image and just work out the details from there. Someday maybe.

 

[jbelljbell]

I'm glad to see someone else tried this on a screen, with believable results. Yes it takes a few tries to get your camera set up, for me, it was figuring out how long of exposure to use for the black picture. Once you get the black picture taken, it's easy to do the white.


The real benefit to DIY'ers in my opinion, is that once you get your camera set, and you get believable results, you can then do comparisons. What happens to contrast when that lamp in the back corner is on vs. off, how does my screen compare to a typical tripod commercial white screen that I can borrow from my local (wherever) ? Then you can post your % differences in your DIY screen vs a known screen (or even compare to a $9 Thrifty White Hardboard)


As long as everything gets referenced back to a known... then you can figure out what you've gained (pun intended) or lost in the gain/contrast departments. And after all... measuring your efforts seems like just part of the fun.

 

[K-Spaz]

I just ordered a DaLite HC White screen to replace/augment the one I've got now. That's a 1.1 gain pull down version vs the tensioned one I've got on the frame now. I also got some samples of materials in the mail and I would say that the screen I am projecting on now is a .9 gain, give a tenth or take very little to nothing (-0 to +.1 I'd estimate)


The gray I'm using now is somewhat whiter than the gray samples I received which are .8 and .9 but my surface is a little rougher than the matte surfaces on these samples. Clearly, I'm not in the 1.2 or 1.3 gain area cause my screen is not even close to that bright.


Hopefully by the weekend I'll have my new screen here and if I get a chance I'll redo these numbers and post results of what I came up with.


As for getting it right, I think the key is just to set the camera manually for a proper exposure. On my EOS, probably getting out of Full Manual and Just going AV would have made my life simpler, but I don't seem to like simple. Oh, and I had more trouble on the white than the black, due to the very short exposure time making the colors go haywire on my DLP. For all I know those red/blue/green pics may well have worked. I didn't try em.

 

[MississippiMan]

Your "DaLite results" will belong over on "Screens", as they have no bearing with DIY Screens.


No offense, of course.

 

[K-Spaz]

^

MM,


None taken but...


Just exactly what will the numbers mean if they are not compared to a known?


I believe the Op suggested just such a comparison in two different posts, as if asking for exactly that. I tend to agree. I don't think it'll make much sense to give out numbers here that refer to a display device, without also then CHANGING THE SCREEN and showing what the SCREEN does to the numbers. How in gawds name am I supposed to compare my screen to someone elses via the internet? Certainly these numbers without reference are not going to help me do it.

 

[MississippiMan]

Hi,


Well on this Forum it's always a bit of a quandary....and in this case jbelljbell's finding of the program bodes well for DIY'er and Mfg Screen owners alike. It was simply presented a bit too generically, and being non-specific in intent, attracted an interested party.


If there was a comparison being made directly between two applications, one DIY and one Mfg, for the reason of determining the former's viability or potential/failings...then that would be perfectly ok.


In your case, there is no such comparison going on. That makes it a effort whose findings have no bearing on the doings in this Forum. However it would seem that now, (...in Gawd's name of course....

) that you do have the ways and means to make a "Internet Comparison' between two Screens. If they are Mfg Screens, on "Screens', I think everyone over there will certainly sit up and take notice of this new method.


It's all good, really.

 

[Harpmaker]

I don't mean to be a "wet blanket" here, but I think there are a number of problems using this test with front projectors. Even the results from LCD monitors are suspect.


From the site itself:
"Luminance accuracy


The form also shows the absolute luminance as calculated from the data entered. All numbers should be known and not guessed or this number will not be meaningful. And even then, the real ISO numbers of cameras may vary considerably from one camera model to the next, so don't take this number as an absolute truth.


The calculation assumes that the sensor or RGB image saturates at a luminance

Ls = 85.5 × (f#)2 / (ISO × t),


where f# is the aperture number, and t is the exposure time. This equation uses the "standard output sensitivity (SOS)" definition of the digital ISO value. [see Wikipedia]. From some experiments with a calibrated luminance meter and various cameras, it appears that the actual sensitivies can differ considerably from this equation, and that it is not always consistent on the same camera at different "ISO" settings."


While not directly related to the goal of this thread, you might want to take a look at this link about using a camera as a Lux meter. http://www.padfield.org/tim/cfys/lightmtr/luxmtr1.php


One of the problems using a camera or a photographic light meter to measure Lux is that most of them are designed to measure much brighter scenes than our PJ screens. The other problem is that cameras and photographic light meters don't measure in small enough increments. Most consumer cameras only measure in 1/3 stop increments. A full stop is a 100% change in light intensity so 1/3 stop would be 33.333%. While this is accurate enough for taking photos, it leaves a lot to be desired when measuring PJ output or screen gain.

 

[jbelljbell]

You are correct -- this method is not accurate in absolute terms. I would never rely on the luminance values presented as accurate.


However -- the original thought was always a comparison. It doesn't matter how far off a camera is from 100% accurate vs a calibrated light meter. As long as the settings are the same from measuring one screen to the other -- the error is a non issue. AKA, it is as far wrong on one screen as the other. Therefore, as long as you are always comparing to a known screen -- you can come up with a percent value difference for the unknown screen in terms of contrast and gain.


The section you quoted was in reference manually typing in values, not in letting the application read the submitted picture info. There is also a section at the bottom of that page talking about the luminance error vs calibrated light meter that has been observed on several standard cameras. There are suggestions on using lowest possible camera resolution, and taking an out of focus picture, to force the camera to average, and reduce the error that is obviously going to be part of this kind of measurement.


As a side note, I've measured a couple projector/screen combination's, and with my cheap-o kodak, I got shockingly believable measurements.


I really think this has value, and uses a common measuring device that everyone has access too. Just have to keep in perspective the potential for error, and make sure that all measurements are referenced back to a known.

 

[xflapx]

Don't you have to assume that deviation in every aspect is linear in order to compare based on a referenced known? What if a difference in settings on a camera altered results in an exponential manner? Wouldn't you also have to run an initial test of the room in it's ambient light state before testing? If none of these questions are applicable and this worked, it would be a great help for the forum. However, even if unknown to everyone, it seemed to work but was completely wrong, it would be relatively harmful. I think it's a great idea if more equipped and qualified people than myself were to test and confirm the method.

 

[jbelljbell]



Quote:
Originally Posted by Harpmaker /forum/post/16908724


One of the problems using a camera or a photographic light meter to measure Lux is that most of them are designed to measure much brighter scenes than our PJ screens. The other problem is that cameras and photographic light meters don't measure in small enough increments. Most consumer cameras only measure in 1/3 stop increments. A full stop is a 100% change in light intensity so 1/3 stop would be 33.333%. While this is accurate enough for taking photos, it leaves a lot to be desired when measuring PJ output or screen gain.

Quote:
Originally Posted by xflapx /forum/post/16910540


Don't you have to assume that deviation in every aspect is linear in order to compare based on a referenced known? What if a difference in settings on a camera altered results in an exponential manner? Wouldn't you also have to run an initial test of the room in it's ambient light state before testing? If none of these questions are applicable and this worked, it would be a great help for the forum. However, even if unknown to everyone, it seemed to work but was completely wrong, it would be relatively harmful. I think it's a great idea if more equipped and qualified people than myself were to test and confirm the method.

Well, since Harp is the 'objective' guy around here, and has 1degree light meter handy... how about a test of a White 1.0 gain, vs a D65 grey, with both light meter, and camera... and post the % difference (luminance and contrast) between the 2 screens measured both ways.