Before you say, “6500K is the standard, and the way broadcast monitors are set, and movies are mastered, etc…” let me explain.

So I calibrated both my projector and plasma for 6500K, and can achieve a very flat grayscale around 6500K (+/- 50K). The problem is, the grays look too red when compared to a neutral whites and grays in indirect sunlight.

My plasma for instance is in a room that can get a large amount of ambient light from outdoors. If I look at neutral white or grey objects which are being lit by indirect sunlight(even on cloudy days), and then look at my plasma with gray patterns, the plasma looks tinted towards red even though the grayscale is 6500K, dead on. I even took measurements of a neutral screen material in indirect sunlight, and measured closer to 7500-7700K.

So I did an experiment and readjusted my TV to 7500K and now whites look very neutral, and so do the rest of the grays.

I did this on my projector as well (Epson 1080UB), and found the greys to look more natural. The only downside on the Epson is that it shifted my entire color gamut away from red, so I need to recalibrate my color saturations and hue for each color.

I like what I’m seeing, and it’s not “too blue”, or unnatural. It looks more like what I see in real life.

Anyone else experience this?

Dan

My guess is that it is caused by inaccuracies in the colorimeter you use. Our vision is very sensitive to even relatively small errors in the white point and affordable colorimeters can easily have a xy0.01-0.015 margin of error with white.

BTW, instead of 7500k, what RGB balance are we talking about? 6500K is not really the standard--D65 is, and the two are not the same.

Before you say, “6500K is the standard, and the way broadcast monitors are set, and movies are mastered, etc…” let me explain.

So I calibrated both my projector and plasma for 6500K, and can achieve a very flat grayscale around 6500K (+/- 50K). The problem is, the grays look too red when compared to a neutral whites and grays in indirect sunlight.

My plasma for instance is in a room that can get a large amount of ambient light from outdoors. If I look at neutral white or grey objects which are being lit by indirect sunlight(even on cloudy days), and then look at my plasma with gray patterns, the plasma looks tinted towards red even though the grayscale is 6500K, dead on. I even took measurements of a neutral screen material in indirect sunlight, and measured closer to 7500-7700K.

So I did an experiment and readjusted my TV to 7500K and now whites look very neutral, and so do the rest of the grays.

I did this on my projector as well (Epson 1080UB), and found the greys to look more natural. The only downside on the Epson is that it shifted my entire color gamut away from red, so I need to recalibrate my color saturations and hue for each color.

I like what I’m seeing, and it’s not “too blue”, or unnatural. It looks more like what I see in real life.

Anyone else experience this?

Dan

Maybe it is the colorometer. That was my first thought too, but wasn't sure it could be off that much. If that is the case, my second thought was, if the meter is reading too much red, it will impact ALL of my saturation measurements as well for the primaries and secondaries.

...and Yes, I'm referring to D65, sorry. x=0.313 and y=0.329 REC.709

A question:
1) Assuming my colorometer is off. Let's say that to get a neutral gray, I need a white point that measures more like x=0.294 and y=0.329. (R=30, G=40, B=40), rather than x=0.313, y=0.313 (R=30, G=30, B=30) Is there a way to compensate for this error for when I'm measuring color saturation, hue, and brightness for all the primaries and secondaries? If the meter is reading too much red for example, my red saturation will be too strong as well, even though it looks correct at 0.640, 0.330. (x,y REC.709)

Thanks Tom.

Dan

Maybe it is the colorometer. That was my first thought too, but wasn't sure it could be off that much. If that is the case, my second thought was, if the meter is reading too much red, it will impact ALL of my saturation measurements as well for the primaries and secondaries.

...and Yes, I'm referring to D65, sorry. x=0.313 and y=0.329 REC.709

A question:
1) Assuming my colorometer is off. Let's say that to get a neutral gray, I need a white point that measures more like x=0.294 and y=0.329. (R=30, G=40, B=40), rather than x=0.313, y=0.313 (R=30, G=30, B=30) Is there a way to compensate for this error for when I'm measuring color saturation, hue, and brightness for all the primaries and secondaries? If the meter is reading too much red for example, my red saturation will be too strong as well, even though it looks correct at 0.640, 0.330. (x,y REC.709)

Thanks Tom.

Dan

What meter are we talking about? If you have to assume the meter is off and make adjustments using your own eyes, then you're right back to where you started from IMO. Try renting a good quality meter. The peace of mind is worth it alone.

A question:
1) Assuming my colorometer is off. Let's say that to get a neutral gray, I need a white point that measures more like x=0.294 and y=0.329. (R=30, G=40, B=40), rather than x=0.313, y=0.313 (R=30, G=30, B=30) Is there a way to compensate for this error for when I'm measuring color saturation, hue, and brightness for all the primaries and secondaries? If the meter is reading too much red for example, my red saturation will be too strong as well, even though it looks correct at 0.640, 0.330. (x,y REC.709).Dan:

First, let me be clear that my diagnosis was just an informed guess. Without actually seeing your display I can't know for sure.

Second, assuming that this is indeed the problem, what you would need is some way to create an offset that applies to your readings to improve their accuracy. Most software packages allow you to do this.

Third, the only way to know how big an offset and in what direction is required is to compare your readings against the readings of a reference instrument. If you don't have one, then you'll have to borrow or rent one, or pay a pro who uses a reference instrument to calibrate your display. Then you can measure using your colorimeter and you have your offset.

Fourth, I wouldn't assume that an error in the white point automatically entails a similar error in the gamut. I have seen many displays in which the white point error was large in the red channel with an affordable colorimeter but the red primary measured quite accurately. It is really a case-by-case thing. You need independent measurements of WRGBCYM.

Yes, I do think my meter is a little off. I notice this mainly in the color of greys, and the hue of blue (which has always seemed a little tinted with purple). I noticed green being slightly lime green at 100% saturation, but not toooo bad.

I did adjust the XYZ matrix in HCFR with an educated guess for how off I think my meter is. It adjusted my calibration locations for colors and greys. When I set my meter to those points now, it does look more accurate, like it should. Thanks for the help. When I get a chance, I'll see if I can rent or borrow a reference meter for a short time to get a true idea of the offsets involved.

Cheers, and thanks for the suggestions.

BTW, my meter is the Eye-One Display LT (I1LT)

Dan

Another issue you have is that human vision is completely imperfect when it comes to comparing things in the conditions you described. Sunlight is not necessarily d65... it can be WAY off (usually in the direction of blue). When your eyes see something that is more blue than what you are comparing it to, your eye-brain locks on the thing that is more blue as being white and everything else you look at is compared to that standard. So the sunlight could EASILY be making the objects you view APPEAR to be neutral/white which throws your basis of comparison way off for the panel.

Furthermore... any color in your room will affect how you perceive the panel also.

The ONLY time you can compare ANYTHING to ANYTHING is in a dark room using an accurate d65 light source with accurate gray/white comparison targets from a reliable sourse like Munsel. (and it would help if the dark room is painted neutral gray with a neutral gray ceiling and and neutral gray carpet (not what you THINK is neutral gray in a store, but something that is TRULY neutral gray)

Anyway... the point is... you are on a real slippery slope with your comparisons. Personally, under the same conditions, I would not trust what I was seeing. I would ONLY rely on my meter - but I have a meter I know is correct (it had better be considering what it cost and how much re-calibration costs).

You are in a situation where you probably can't trust your meter -- but, seriously, you can't trust your eyes either. Keep in mind that your eyes are easily fooled by what is present on the screen. I've had plenty of times when I thought the whiter steps in a grayscale ramp pattern were really white and grayer steps had a color tint, but it turned out the white(r) steps were wrong and the grayer (darker) steps were closer to d65.

This "sticky" thread (http://www.avsforum.com/avs-vb/showthread.php?t=849430) discusses some of the issues Doug mentioned and links to other demonstrations that vividly reveal how easily fooled our human visual system is.

Best regards and beautiful pictures,
G. Alan Brown, President
CinemaQuest, Inc.
A Lion AV Consultants Affiliate

"Advancing the art and science of electronic imaging"

Thanks for the input guys. The illusions in the sticky thread are great, especially how grey can look so different depending on environment.

I think for me, it might be a combination of my meter being slightly off, and using daylight as a reference. In a completetly darkened room, with a low IRE grey pattern on my screen (so as not to send much light into the room reflecting off wall surfaces), when I calibrate to D65 there is still a slight red tint to the grey pattern. It's not a gross red tint, but I tend to see a scew towards red. I think I'm just more sensitive now that I've been adjusting greyscale on my displays for 13 years now. For a long time, it was by eye using gains and cuts, but recenly using my colorometer. I notice this slight red tint wether it's on my projector in the basement, or the plasma on my main floor. Friends, or my wife, would never notice the stuff I'm mentioning here.

When you guys, Tom, Doug, George... look at a D65 pattern on a properly calibrated screen, do you notice any color tint or does it appear completely neutral grey?

I'll probably stop chasing this one soon and use a reference meter to verify or discount what I've been seeing, or just completely rely on my meter, which has done a fantastic job overall to be honest. Overall, it's readings verify what I see on the screen.

Dan

If you are displaying 1 window pattern at a time and there are no other gray levels on the screen at the same time, and you feel like some steps are favoring red, chances are there is a bit too much red in those steps. If you are evaluating this using a ramp pattern (o%-100% all on on the screen at the same time), you cannot trust what you are seeing.

You are in a situation with NO reference you can trust so there is no way to know for sure what is the right thing to do. You either need a meter you can trust or a pro calibrator so equipped.

So to conclude this post...

So it seems that my LT meter is off. Fellow member Knd created an offset of his LT to an i1pro on an Epson ****UB projector. This is the same line of projector that I have, in the same color mode and lamp setting. Granted, no two meters are identical, but when I look at his offset, it lines up nearly perfectly with what I was doing by eye to get my LT calibrations to look correct. Basically, I was removing about 12% red to get my greys to look correct. I didn't need to touch blue and green. This is the same as what his offset showed.

I calibrated with his offset, including the related shifts to the color gamut, and my calibration is hands down the best I have seen to date, and extremely natural looking. Just wanted to comment that while it's not complete proof that my meter is off, I'm very happy with my calibration now. Even more than what I was getting before, which wasn't bad, but reds just seemed pumped, and my green and blue hue always seemed to skewed towards red. Not anymore!

So for me, the only reason 7000-7500K looked more natural was because my meter was off, when in reality, I was close to D65. To prove this 100%, I need purchase or borrow an i1pro meter, but I'm so happy with my results now that I probably won't bother.

Dan

So for me, the only reason 7000-7500K looked more natural was because my meter was off, when in reality, I was close to D65. To prove this 100%, I need purchase or borrow an i1pro meter, but I'm so happy with my results now that I probably won't bother.

Dan

Please notice that an i1 Pro is not a reference-grade meter. It is a great value-for-money allround tool, that I happen to use myself, and if it is calibrated recently it will most likely be much more accurate than your LT. However, if it hasn't been recently calibrated, you will not be able to use it to accurately evaluate the performance of your own meter. I've seen differences in the same range as you're noticing, between two different i1's. My own had just been serviced, the other one was visibly off, much like your experience. Sadly, not everyone takes the servicing interval seriously.

Thanks Otto. Very good points.

Dan