I just received a Minolta LS-100 luminance meter. This is a very expensive and sensitive instrument, measuring down to 0.001 fL.

I ran some tests. First I measured the peak output of a Kuro plasma using the Minolta and then using a Chroma 5 colorimeter. The results were virtually identical. So far, so good.

Next, I attempted to read the black level of the Kuro. The Chroma 5 couldn't read anything. The Minolta read 0.001 fL. Since the peak output was 34.7 fL, this means that the Minolta measured the Kuro's on/off contrast as 34,700:1. Again, so far so good. The Minolta simply reads lower than the Chroma 5, which is what I expected.

Next, I measured the peak output and black level of the Sharp 20K projector with the Minolta. The bulb is nearing the end of its life cycle, so I got 8.6/0.003 fL for a contrast ratio of 2,867:1 (in the Medium Iris setting).

Next, I measured the same thing, except this time I used a AEMC CA813 illluminance meter pointed towards the lens placed about half way between the lens and the screen. I got 175/0.06 lux for a CR of 2,917:1, nearly the same as what I got with the MUCH more expensive Minolta. Contrary to what some have suggested, the AEMC device seems suitable for such readings.

Here's the weird part. When I tried to measure the checkboard CR of the Kuro, the Chroma 5 and Minolta again measured the same for the white squares, but they returned radically different values for the black squares. The Minolta read 0.07 fL but the Chroma 5 was, again, unable to return any value. 0.07 fL is well above the Chroma 5's range of operation, so it was trying to read a much darker value.

Anyone have a theory to explain this discrepancy? I have always estimated the Kuro's checkerboard contrast to be a stratospheric value in the many thousands to one. The Minolta tells me that it is a mediocre 415:1.

This is one case where I am inclined to trust the much less expensive contact probe over the fancy non-contact meter.

The Minolta read 0.001 fL. Since the peak output was 34.7 fL, this means that the Minolta measured the Kuro's on/off contrast as 34,700:1.One thing to keep in mind is that you were down at the limits of the meter for black, so the margin of error there is very large. Even if the meter is perfect you would have a range of something like 0.00050 to 0.00149 fL for a white level reading. I suggest sliding the battery compartment cover off and changing to cd/m2 to increase the bottom end accuracy a little bit.
Here's the weird part. When I tried to measure the checkboard CR of the Kuro, the Chroma 5 and Minolta again measured the same for the white squares, but they returned radically different values for the black squares. The Minolta read 0.07 fL but the Chroma 5 was, again, unable to return any value. 0.07 fL is well above the Chroma 5's range of operation, so it was trying to read a much darker value.

Anyone have a theory to explain this discrepancy? I have always estimated the Kuro's checkerboard contrast to be a stratospheric value in the many thousands to one. The Minolta tells me that it is a mediocre 415:1.Did you have the meter up close to the Kuro? I use a Minolta LS-110 (basically the same thing with the range all moved up 10x higher, so reads down to .01) and what I found with front projectors is that I couldn't measure over about 200:1 for ANSI CR without using a light pipe I made that is lined with black velvet and maybe 18" long. I put it on the end of the meter and shoot through it so that not much light would get into the meter from the white rectangles. It wouldn't take much light getting into the front of the meter from the side and getting into the sensor to reduce thousands to one down to 415:1. Like you, from what I have heard the Kuros have a lot of ANSI CR.

I have changed how I do it most of the time to avoid the light pipe. I now use a black board covered with black velvet and a hole in it. I let the light from one rectangle at a time fall on the screen and measure it. But of course that is for front projectors.

--Darin

Here's the weird part. When I tried to measure the checkboard CR of the Kuro, the Chroma 5 and Minolta again measured the same for the white squares, but they returned radically different values for the black squares. The Minolta read 0.07 fL but the Chroma 5 was, again, unable to return any value. 0.07 fL is well above the Chroma 5's range of operation, so it was trying to read a much darker value.

Anyone have a theory to explain this discrepancy? I have always estimated the Kuro's checkerboard contrast to be a stratospheric value in the many thousands to one. The Minolta tells me that it is a mediocre 415:1.

Does the Chroma 5 "seal" to the display better than the Minolta? My guess would be that the Minolta gets a lot of "noise" from the white squares since it wouldn't seal to the screen well, while the Chroma 5, sealing to the screen would read better.

The Minolta doesn't seal at all. It is a non-contact meter that you take readings with from several feet away.

Thanks, guys. I think that you are right. Even though the reading area is much smaller than the square itself, the readings of the dark squares are obviously being contaminated by the light from the light squares.

I'll have to come up with something like Darin's light tube to get an accurate reading for checkerboard contrast. I think that the lower limit of the Chroma 5 is 0.01 cd/m2, which means that the ANSI contrast of the Kuro is at least 9,500:1.

I spent about 2 minutes attaching a cardboard tube for paper towels to end of the Minolta with masking tape and the black level reading on the checkerboard fell from 0.23 to 0.06, which is still "only" 1,600:1.

BTW, when I tested on/off again in the cd/m2 mode I got 119.3/0.003 or 39,767:1, about 15% higher than the first reading.

When you have one significant figure in your measurement, 15% doesn't mean much.

I used Darin's suggestion about using a mask and this worked quite well. I had a large piece of black velvet cloth that covered the entire face of the Kuro panel. I cut a hole in it approximately the size of one of the checkerboard squares.

Using this method, the Minolta now reads 102/0.012 cd/m2 or 8,500:1. Interestingly, by placing the head of the AEMC CA813 directly against the panel I get 131.6/0.015 lux (it fluctuated between 0.02 and 0.01). This is 8,773:1, further evidence that this relatively inexpensive device is reasonably accurate and a great value.

Definitely agree that some method of blocking stray light is necessary for accurate ANSI measurements.

I'm still not sure why the C5 isn't measuring the black level properly though. The luminance spec is 0.05 cd/m^2 although I think that this is the spec for accurate color measurements. When used for contrast only measurements I think you're right that it should be accurate lower and probably down to around 0.01. I've never had issues measuring low black levels with the C5 so long as the instrument is mounted close enough to the display. When measuring around 40k:1+ it can be tricky to measure this high of dynamic range without running into problems at the ends (either hitting a rail at the top with too much luminance or not measuring the bottom due to too little luminance). Placing the sensor at the right distance from the display is critical.

I'm still not sure why the C5 isn't measuring the black level properly though. The luminance spec is 0.05 cd/m^2 although I think that this is the spec for accurate color measurements. When used for contrast only measurements I think you're right that it should be accurate lower and probably down to around 0.01.Since my reading with the Minolta was 0.012 cd/m2, the lower limit for the C5 must be higher than that. I am waiting for some official specs from X-Rite about this.