AVS Forum banner
Status
Not open for further replies.
1 - 14 of 14 Posts

·
Registered
Joined
·
779 Posts
Discussion Starter · #1 ·
Recently I have become interested in understanding how my projector is performing, but not so much that I was willing to spend lots of money on a light meter as I would rather buy DVDs or other consumer electronics. So I was spinning my wheels online the other day and did a search looking for a cheap LUX meter. I came across this fx-101 , it claims to measure 0-50000 Lux, and it was only $35+$7 shipping, so I ordered it. Since I haven't received it I don't know if it is any good or not, but for under $50 I figure it was worth it just so I can track if my bulb is dimming or it is my imagination.


So here is the purpose of my post:

1. Does anyone know if this unit is any good?


2. I have looked through the archive for posts on measuring contrast ratio and came across this but I was hoping that someone who has some experience in measuring projectors might be able to provide a synopsis of interesting measurements, techniques and processes for measuring things such as contrast.


3. At >$50 (if this meter works) I though others might be interested in measuring their projectors so they can provide measured quantitative results using an agreed upon process and standards on their projector performance. I thought this might be more interesting that feedback like "This projector blows the doors off any other projector". From some of the archive post I read it might help people understand the affect of things like room reflections on contrast and black levels.


Any thoughts?


Regards,


Brian
 

·
Registered
Joined
·
3,667 Posts
Quote:
Originally posted by Steve Smallcombe:
Make that the square of the reciprocal of the distance.
Not necessarily, the inverse square law only holds true for point sources.


Also Steve can you post a link to SMART.


------------------

Ken Elliott


[This message has been edited by kelliot (edited 09-01-2001).]
 

·
Registered
Joined
·
343 Posts
Brian,


I have a similar Lux meter made by Sper Scientific for which I paid about $100. I used it to get started in the measurement game. What I found was that the resolution of 0.1 Lux was similar to my black level and hence did not allow accurate measurement or tweaking. This, of course led me to the photocells and Digital Multimeter approach, which surprisingly was much more sensitive at low light levels, and did allow tweaking of the bias levels of my SONY 10HT. I did find the Lux meter useful in selecting and confirming the linearity of my photocell. The first one I tried was not quite linear.


As you know for SMART, perfect linearity is not required, as the main object is to get consistent tracking of the RGB color balance while maximizing the reported contrast and lowering the black levels. Most people find that the contrast ratio with the 10HT doubles in the process of SMART tweaking and the addition of CC filters.


If however you want to know the absolute contrast ratio, you will need to check your detector's linearity. The cheapest way is to repeat the measurement series at a different distance. Ideally, you should get exactly the same traces and contrast ratio, but with the measured light level scaled by the reciprocal of the distance of the detector from the meter.


The other way of confirming your detector linearity and also measuring absolute Lux levels is obviously to buy a meter, or since we are kind of neighbors, borrow mine.


Steve



------------------

Steve Smallcombe
 

·
Registered
Joined
·
3,667 Posts
Actually the prior discussion illustrates some issues on light control and brightness. Typical ambient levels in a room without light control can easily run at 10 lux. Under these circumstances, ambient completely controls contrast.


Its tough to get light levels down to 0.1 lux. The reflected light off of most screens raises the light level.


------------------

Ken Elliott
 

·
Registered
Joined
·
343 Posts
Make that the square of the reciprocal of the distance.


Steve


------------------

Steve Smallcombe
 

·
Registered
Joined
·
343 Posts
 http://home.pacbell.net/steve367 should get you there.


The inverse square law seems pretty close in practice, from what I have seen. The most important thing is that the measured contrast ratio shouldn't change with distance if the detector is "linear".


------------------

Steve Smallcombe
 

·
Registered
Joined
·
111 Posts
I too tried an expensive light meter ($60) from Extech with a sensitivity of .1 lux. It was returned because it wasn't sensitive enough - 'black' registered .1.


Extech also has a 'light probe' with a sensitivity of .01 lux for a little over $100. Seems this should work though I haven't picked one up yet. In the interests of economy I'm building the photo resistor light sensor that Steve describes on his site.


Extech's product page url is http://www.extech.com/Products/EXTECHProdset.html


I'm unfamiliar with light measurement units. The Extech product docs seem to imply 1 lux = 1 foot-candle. Anyone know if this is correct?


Frank L
 

·
Registered
Joined
·
343 Posts
Another way of checking detector linearity is via use of ND filters. I have a selection of them, as I was once a strong proponent of using ND to improve the black level of LCD projectors, such as my 10HT.


The point of this test is that an ND filter should attenuate the light by the same factor, e.g.2, at all IRE levels, and that is essentially what I found with my photocell detector.


What I also found, was that my expensive multicoated ND filter was not completely neutral, i.e., it attenuated blue at bit more than red, which was attenuated a bit more than green. While this was easily measurable and correctable using SMART, it seemed to me to be going in the wrong direction, as it meant turning down the green gain to re-establish color balance, and this hurt the measured and perceived contrast ratios. As I thought about the implication of this, I came up with the idea of using CC filters, after the lens, to proactively improve contrast and black levels. With the 10HT, this makes a fairly dramatic improvement in picture quality, as many have now experienced as well. Needless to say, the ND filter is now in a drawer.


BTW, my light meter also reads in FC or Foot Candles, and it's resolution is 0.01 in the FC mode. The ratio of Lux to FC is roughly 11:1 so a 0.01 FC meter is no more sensitive than a 0.1 Lux meter.



------------------

Steve Smallcombe
 

·
Registered
Joined
·
835 Posts
What we've all got to remember is that these instruments are not going to be exactly accurate, irregardless of how sensitive they are (0.1 lux or lower). Most inexpensive sensors will not follow the response curve of the eye. For example, our eyes are less sensitive the farther into the red or blue we go, the maximum response is in the green range. If your sensor weights red or blue as equal to green you will get unreliable results. Of course, meters calibrated to the eye's response curve are *expensive*. http://www.avsforum.com/ubb/frown.gif


That being said, it should be pretty good for comparative measurements. That is you can tweak some settings and compare the before and after readings, but for absolute measurements like screen brightness the results are likely not accurate.


Regards,


Kam Fung
 

·
Registered
Joined
·
343 Posts
As part of the SMART process the photocell is calibrated to correct for the quantum efficiencies of the detector at the various, R, G, and B wavelengths. Certainly using the raw detector response can be misleadingm and SMART does not use it. Instead, the 3 colors are isolated and then calibrated.


The SMART tweaking/CC filter process is also heavily based on the perceptual phenomena you mentioned, as I place much more emphasis on getting the lowest possible green and blue contributions to the black level and less on color balance at the lowest light levels. Who can see red at IRE 0 anyway? That is why the "green contrast ratio", I report, may be the most relevant. Fortunately, this can all be done with a $2 photocell, DMM, the SMART spreadsheet, and for best results with the 10HT, a $20 CC filter.





------------------

Steve Smallcombe
 

·
Registered
Joined
·
3,667 Posts
Some of them have been engineered for correct photometric response for job related lighting.


If they are accurate to a few precent that may be good enough.


------------------

Ken Elliott
 

·
Registered
Joined
·
3,667 Posts
Check Steve's post above, its on his web site.


------------------

Ken Elliott
 
1 - 14 of 14 Posts
Status
Not open for further replies.
Top