AVS Forum banner
1 - 9 of 9 Posts

·
Registered
Joined
·
283 Posts
Discussion Starter · #1 ·
There have been quite a few topics regarding the accuracy of colorimeters. The aforementioned colorimeter (eye one) usually seems to suffer from under reading of 'Red' giving a pinkish tint to the greyscale. However, mine seems to over read 'red'. This can thus affect the overall colour temperature settings.


My idea is to use a source, which has a known colour temperature, for example a 100 watt incandescent light bulb, which has a temperature of 2700k.


This form of colorimeter calibration may be very basic and primitive; however, it will give you a basic idea of how the colorimeter is performing


Any other ideas on basic checks that can be provisioned?


Thanks


Delphi
 

·
Registered
Joined
·
6,074 Posts
Sadly, that won't really tell you anything. Unless you are using a bulb that has been measured with a reference instrument (manufacturers' specs for consumer-grade bulbs should be taken with a large grain of salt), you won't truly know what's being measured and thus won't have an accurate idea of the meter's performance.


Also, one should know the exact CIE x,y coordinates of what is being measured to compare with what the probe is saying. CCT by itself won't do. For instance, D65 (0.313, 0.329) is widely held to have a CCT of ~6504K. Yet years ago, Michael Chen demonstrated that 6504K can range from greenish to magenta-ish (is that a word?) and still be reported as 6504K. Your assumed 2700K could also have a reasonably wide variation in its actual color and still be read as 2700K. So how could you make a valid assumption about the meter's performance in such a case?
 

·
Registered
Joined
·
283 Posts
Discussion Starter · #3 ·
I was more thinking about a basic test giving an idea whether the meter is over reading/under reading red/blue. It may be said how would you know whether the problem is red or blue. That can be decided by looking at the greyscale ramp, and then looking at the discrepancy in colour temperature taken from the source reference and applying that to the display measurements.


The problem would be finding a source which can provide a reasonable accurate colour temperature.
 

·
Registered
Joined
·
5,916 Posts

Quote:
Originally Posted by delphiplasma /forum/post/19527219


I was more thinking about a basic test giving an idea whether the meter is over reading/under reading red/blue. It may be said how would you know whether the problem is red or blue. That can be decided by looking at the greyscale ramp, and then looking at the discrepancy in colour temperature taken from the source reference and applying that to the display measurements.


The problem would be finding a source which can provide a reasonable accurate colour temperature.

As usual, calibrating by eye doesn't work. If you want to find out where your meter stands against reasonable reference you have to do something like this .
 

·
Registered
Joined
·
6,074 Posts

Quote:
Originally Posted by delphiplasma /forum/post/19527219


I was more thinking about a basic test giving an idea whether the meter is over reading/under reading red/blue. It may be said how would you know whether the problem is red or blue. That can be decided by looking at the greyscale ramp, and then looking at the discrepancy in colour temperature taken from the source reference and applying that to the display measurements.


The problem would be finding a source which can provide a reasonable accurate colour temperature.

Short of the meter profiling that Buzz mentioned, the best way to validate what the meter is telling you is with an optical comparator. These have a neutral grey surface(s) and a D65 light source. You would calibrate the display's greyscale with the meter as usual then view a ramp or step pattern through the comparator. This would show you if the scale is off, and in which direction, but not allow you to develop a correction factor unless you compared the x and y readings before and after using the comparator. And, the correction factor would likely not be portable (usable on other displays), since displays-and other light sources-have different spectral power distributions that affect what the meter "sees". Meter profiling also has the same limitation.
 

·
Registered
Joined
·
1,465 Posts
Quote:
Originally Posted by Rolls-Royce
Short of the meter profiling that Buzz mentioned, the best way to validate what the meter is telling you is with an optical comparator. These have a neutral grey surface(s) and a D65 light source. You would calibrate the display's greyscale with the meter as usual then view a ramp or step pattern through the comparator. This would show you if the scale is off, and in which direction, but not allow you to develop a correction factor unless you compared the x and y readings before and after using the comparator. And, the correction factor would likely not be portable (usable on other displays), since displays-and other light sources-have different spectral power distributions that affect what the meter "sees". Meter profiling also has the same limitation.
Optical comparator - now thats something I havent heard since I had my old mits crt 55807 - with lockout color decoder that was hacked thru a i2c device back years ago on the ***************...scary times for lots of us owners..nice tv though when calibrated - in fact- its the best tv I ever had when calibrated..CRT is gorgeous...i miss the beast..
 

·
Registered
Joined
·
12,276 Posts

Quote:
Originally Posted by delphiplasma /forum/post/19525057


There have been quite a few topics regarding the accuracy of colorimeters. The aforementioned colorimeter (eye one) usually seems to suffer from under reading of 'Red' giving a pinkish tint to the greyscale. However, mine seems to over read 'red'. This can thus affect the overall colour temperature settings.


My idea is to use a source, which has a known colour temperature, for example a 100 watt incandescent light bulb, which has a temperature of 2700k.


This form of colorimeter calibration may be very basic and primitive; however, it will give you a basic idea of how the colorimeter is performing


Any other ideas on basic checks that can be provisioned?


Thanks


Delphi

I'll bet you take 100 different 2700k incadescent lamps and measure them...you'll have 100 different results. Lamps of any kind are one of the most inconsistent things made.
 

·
Registered
Joined
·
283 Posts
Discussion Starter · #9 ·
Thanks for the responses. I do understand there are too many factors taken into account when assessing d65 and that a light bulb is far too basic.


However, just one more query. How would a monochrome monitor/tv compare? Is a monochrome CRT phosphor set to d65?
 
1 - 9 of 9 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top