Using the CA813 light meter to measure your front projector

I've had a lot of questions recently about using a light meter to measure lumens, ftL, on/off CR and ANSI CR so I'm putting together this basic tutorial. I hope others with experience will chime in and add their tips and techniques that have worked great for them as well.

Note that the following is a depiction of how I use the meter and what works for me. Others techniques I'm sure can be used as well, which may even be better. So don't take this information as definitive, but rather just to give you some ideas. I'm rather new to this meter so it would be great to hear the feedback from others.

Also I'll talk about specific measurements I got with my Ruby at 525 hours on the bulb and using near max throw. For reference I have a 106" diag Firehawk with 1.3 gain.

OK, let's get started...

What is the AEMC CA813?
The AEMC CA813 is a light meter that can be used to measure light output from your pj in foot candles (fc) and lux. It is a popular light meter among AVSers because it is accurate, inexpensive (about $165), and convenient to use.

Using the meter to measure lumens
Set the meter to measure in lux with one decimal point of precision. Display a 100 IRE full field pattern. Ideally your pj should be first calibrated to D65. However this is not necessary if you want to know what the max light output is that your pj is capable of regardless of its calibration.

Stand AT THE SCREEN and rest the back of the sensor gently on your screen (or hold it just in front of it), facing the pj lens. You could use a tripod but doing so doesn't allow you to measure exactly at the screen which I prefer.

Because it is a bright reading the number should dial in within a few seconds. Hold the meter still for best results. If it shows 0 and you have a very bright pj try removing the decimal point percision, or you can also measure in fc instead.

In my case I measured 62.3 lux. However to determine lumens you must now factor in the size of the screen. At first I was confused by this. I thought to myself "What does the screen size have to do with anything since I am measuring the light directly from the pj and not off the screen?".

Well the answer of course is that the lens zoom and focus changes the intensity of the light. So as you zoom out and throw a larger picture, the intensity of the light goes down, and vice-versa. The lumen measurement is designed in such a way that, although the screen size factors into the measurement, it provides a result which can be compared apples-to-apples with anyone else's lumens measurements regardless of their screen size (assuming they measured correctly).

OK, next we convert our lux or fc measurement to lumens. The formulas for this are:

lumens = (lux / 10.76) * square feet of screen in FEET (not inches)

or

lumens = fc * square feet of screen

So in my case my 106" diag FH is 33.22 sq. feet. So I had (62.3 / 10.76) * 33.22 which comes out to 192.4 lumens.

Converting lumens to ftL

To convert your lumens to ftL use this formula:
ftL = lumens * screen gain / (area of screen in sq. feet not inches)

So in my case I was getting 192.4 * 1.3 gain / 33.22 = 7.53ftL

Just a side note: I considered the picture I was getting from my Ruby to be quite bright and punchy, so I was SHOCKED to determine I was only getting 7.5ftL. This taught me an important lesson, because previously I was under the impression that I "had to have" at least 12 ftL to be satisfied. Your preference may be different so you really need to get a sense for yourself what is acceptable. It may also have been the super high 20,200:1 CR I was getting that helped the perceived brightness to be much higher than I would have thought.

How to measure ON/OFF CR
Unlike the previous measurement for lumens, the measurement for on/off CR should be done very close to the pj lens - say about 12-18 inches. This is because the AEMC will not give you sufficient accuracy at the screen with a 0 IRE measurement, especially if your pj has excellent black levels.

The key here is that you do NOT move the zoom or focus of the pjs lens when taking this measurement. Just move your meter up to the front of the lens. You will also need a tripod for this to get an accurate black level reading. Although it can be done without one, it grows tiring trying to keep the sensor stable enough to get an accurate reading otherwise.

So with your meter now just in front of the pj lens using a tripod, set the meter to measure lux with two decimal points of precision. Display a 0 IRE full field pattern. It will take a while for the sensor to adjust. You'll likely see an initial high reading, and then the meter will start dropping off. Give it some time - it can take up to a minute or two. Eventually it will stop dropping and at that point you have your reading.

In my case for 0 IRE I measured 0.16 lux. [Side note: You cannot compare this black level by itself to that of your own, as it is dependent on many factors such as the distance to the lens and zoom/focus position. ]

Next change to a 100 IRE full field pattern. For this you must switch the meter to measure in fc because it will be very high when measured at the pj (much more so then when measured at the screen previously).

Also important is that you measure the 100 IRE pattern at the same exact position as the 0 IRE pattern. This is why using a tripod is so important.

For my 100 IRE pattern I measured 300 fc. To calculate my contrast ratio I just need to convert the fc to lux. To convert fc to lux just multiply the fc by 10.76391. So in my case 300 fc = 3229.173 lux.

To calculate the on/off CR, simply divide the two. In my case:
3229.173 from 100 IRE measurement / 0.16 from 0 IRE measurement = 20182:1

(In case you are wondering, I am using Darin's iris tweaks for the Ruby which yields substantially greater on/off CR than otherwise out of the box).

How to measure ANSI CR
First just a note - I think technically to measure true "ANSI" CR one must make a number of measurements from various parts of the screen and do averages or something along those lines. I do not do this as I think others do not as well. So its important to note that I do not believe the technique I am explaining here measures real ANSI CR but gives you a comparable measurement.

ANSI CR can be measured both at the pj and at the screen. This is because the 0 IRE block in the checkerboard will be a lot higher than a full 0 IRE pattern which enables you to get a good reading at the screen. Reading at the screen is also important as it takes room reflections and any ambient light into account for what you are actually getting at the screen.

Now onto our measurements. Put up the first checkerboard pattern. Put the tripod in front of the pj, similarly to how you did for the on/off measurement. Except this time you want to move the tripod so the sensor is off center of the screen, and instead it should be centered with a black box on the checkerboard pattern. I use the first black box off center. So look back at the screen and position the sensor so that the shadow from the sensor thrown on the screen is centered in that black box.

With the sensor position, take a reading in lux and note it. Now move to the next checkerboard pattern so the sensor is then centered in a 100 IRE white box. You will likely need to read this in fc. Convert this to lux and note this value. Divide the two. The result is your ANSI CR measurement at the pj. In my case I had:
3864.2 from the 100 IRE checkerbox / 37.5 from the 0 IRE checkerbox = 103 ANSI CR at the pj.

Now, repeat this same process, except measure this AT THE SCREEN. Your ANSI will go down, unless you have a perfect bat cave. In my case my ANSI CR at the screen was 70.9 lux / 0.84 lux = 84.4 ANSI CR at the screen.

An interesting expiriment, which I will soon be doing, it to cover up reflective areas of the room and repeat the ANSI CR measurements at the screen to see how things improve.

As a side note - the AEMC does not come with a tripod mount and I don't think one is available. I made a makeshift one very easily which makes taking the 0 IRE measurements much easier. For this I cut a block of wood so that it fits tightly within the top of the tripod mount (insert in its tray). I then drilled two holes in the top that were just big enough to have the two prongs on the sensor fit in snugly. I then covered the whole thing in black felt.

I then also made a black square of felt and cut a hole in the middle just big enough to fit over the black ring which contains the sensor itself. This blocks the light that would otherwise be reflected back at the pj lens and then back into the meter because otherwise that area is a bright yellow plastic that is highly reflective (surprising they wouldn't have at least made this black). Total time about a half hour.

I know I mentioned it here quite a bit, but again I want to stress the importance of using a tripod mount for all these measurements.

Please feel free to ask any questions and post your own experiences and techniques you use in performing these measurements.

For related reading about determining whether a certain screen type and size will work with a projector you are considering, see Bob's excellent thread "Will this projector/screen work in my room? - A tutorial" here: http://www.avsforum.com/avs-vb/showthread.php?t=753497

Agree 100%. I think many people will be surprised at how bright 7-8ftL looks with a super high CR pj in a light controlled room.


In my case I think most of the light is coming from the light reflected off the screen, to the back wall, and then back onto the screen. So I will try covering up the back wall with some material and remeasure. My walls are already a medium-dark green and due to WAF I'm not going to get much better in that regard. However if I can come up with even a temporary solution like hanging material across the back wall just when sitting down to do some serious viewing and then removing it, so be it!

Also my carpet is quite light and I suspect that is casting quite a bit of light back into the room. So I'll try covering that up separately and see how that impacts the measurements.

That's whats so great about the CA813 is that its makes it so easy to experiment with your room and get direct feedback on what type of impact the changes are making.

As a side note, I thought some of you may be interested in this comparision between my Ruby and old Sharp 10K. With both projectors calibrated @ D5:

Ruby: ANSI CR at pj: 103:1, ANSI CR at screen: 84:1, on/off: 20200:1

Sharp 10: ANSI CR at pj: 279:1, ANSI CR at screen: 100:1, on/off: 1320:1

Although the Ruby of course blows away the Sharp 10K in many respects, it is interesting to see DLP's distinct advantage in ANSI CR even when comparing LCoS to DLP technology that's 3 years old. If the specs on the JVC turn out to indeed be 300:1 or more on ANSI CR that would be a nice improvement. Of course as we've been talking about here, much work is needed to control room reflections to get the maximum benefits of such improvements.

Hi Tom - thanks for the great input. How did you secure the black velvet to the ceiling? Do you think black felt would be equally effective? Wow, 200:1 ANSI CR at the screen is indeed excellent by all accounts - nice work!


Yes indeed. I was quite surprised myself as I've said. Having the CA813 has completely redefined my standards for what is acceptable ftL because I KNOW what brightness actually looks like at these levels, vs. just having to go by recommended standards.

Based on my research I have concluded that indeed the Sharp 20K would work well for me. Based on your report of 344 lumens in high lamp / high contrast mode (which is the only mode I'd use) I'd start out at about 13.5ftL which would be great. Then assuming I lost half the brightness at about 600-800 hours I'd be where I am now which is acceptable and just change the bulb once a year.

It it wasn't for the 20K's several thousand dollars more street price, it would be hanging on my ceiling already. I still haven't completely ruled it out. However if the JVC specs how out to be true, it'll likely deliver similar (or better) on/off CR while providing twice the brightness and for much less $. I love the sharpness of DLP however... trade offs, trade offs...

That's a great question. I don't want to get too off topic since this is specific for front pjs and posted in the front pj forums. But I suppose you do the same except you put the sensor right up to the screen or close to it.

Before you even go that route however, I can tell you this is not a great idea... Your RPTV will be throwing a ftL measurement which is "off the charts" higher compared to what you'll get from a front pj and its not really even a fair comparison. What I'd recommend instead is going to a HT store and looking at those pjs and measure their lumens and ftL. That will give you a much better idea.

As a follow up to my previous posts - I had a chance last night to try various experiments in the room to improve ANSI CR. See http://www.avsforum.com/avs-vb/showthread.php?t=757920 for details on what I found. Thanks.

BTW I should also add that its a good idea to wear dark clothing, particularly a black shirt. After taking readings in consecutive nights I found that my black level readings with the ANSI CR pattern were 0.04 lux lower with the black shirt since I was holding the meter at the screen. At a glance that may not seem like a lot but it actually affects the ANSI CR reading pretty significantly.

I can't recall where I got mine but a Google search will turn up options (this is what I did).

My pleasure guys.

One thing I would add is that when taking the on/off measurements at the lens - care should be taken to minimize any reflected light. This can be done one of two ways, and preferably both.

One - if your pj is ceiling mounted, temporarily put some black material on the ceiling especially if it is just a few inches away from the lens. With my RS1 the lens is no more than 6 inches from the ceiling, so some light at 0 IRE still goes up to the ceiling and makes it back to the probe. Just the slightest amount of stray light can skew the on/off results considerably.

Second - create a "light trap" behind the tripod mounted meter to prevent light from researching the screen, thereby minimizing the stray light that could make it back to the meter. I do this by holding a large black piece of velvet behind the tripod and checking the screen to make sure its blocking all light from reaching the screen.