Calibrating Plasma - Waste of Time?

[dingwall]

I bought a spyder, calman and getgrey to experiment with my displays.

I first tested my small Sony CRT - and found the gamma and primary points to be nearly spot on. The RGB levels were 95% R, 95% G, and 110% B. Delta E for D65 was obviously out by a bit. Picture is and always has been subjectively very good, however.

I then tested my small Sony LCD - the gamma was spot on, and the primary points were undersaturated but not deviated to one another. The RGB levels were surprisingly all within a couple of % of each other. Delta E for D65 was under 4 for most IRE values. It took me under 10 minutes to get at close to zero for all values, and for RGB levels to be reported as within 0.02%. Colour has always been subjectively very good. Putting it next to the CRT with the same material showing was an eye-opener: the CRT was revealed to have too much blue in skys and the such like, with the LCD being perfect. The CRT's colours were very much like watching a TV, the LCD like being at the cinema...which I guess makes sense. This calibration lark was looking to be amazing.

To keep the next bit simple: I'll talk about just one of my plasmas which I had in the same room as these two displays. This was a Panasonic - the 37PD60, in normal mode and warm preset. The gamma curve was a bit off, as were the primary colour points, especially green which was ridiculously oversaturated, as is the case with all plasmas. The RGB levels were all surpisingly within a few percent of each other, and delta E was under 4 for all levels of 20IRE and above. I therefore did not calibrate any further. But here's the thing: subjectively the display's colours look terrible. Everything looks ridiculously green and unpleasant. Looking next to the LCD highlighted it obviously. The difference between the LCD and CRT is miniscule in comparison. I really can't stress how bad it looks.

My conclusion is therefore that the oversaturated green on plasma means calibrating the greyscale leads to subjectively very poor (I'd say unwatchable) results. What do those who calibrate plasmas think?

When I can borrow a digital camera, I'll glady put up some photos to compare.

[Neil Davidson]

If the end result of the greyscale calibration to D65 is displeasing (ie large DE in primary and secondary) then you can recalibrate to a different white point which minimises the DE of the primary and secondary colours.

Identify the alternative white point by drawing a line from each primary in to the opposing secondary (ie Green to Magenta, Red to Cyan, Blue to Yellow). The point where these lines intersect can be used as the new target whitepoint.

This situation doesn't arrise too often but the display manufacturers are starting to understand the need for better control. The new Pioneer displays not only have a colour management system, they also have selectable colour spaces which allow the famously oversaturated green of earlier models to be easliy tamed.

HTH

Neil

[dingwall]

Thanks for the reply - I'll give it a try, but have a couple of Qs.

What are the target secondary points, given they have to be along the line connecting the actual primaries - is it to get the reference Y, Y and X values for cyan, yellow and magenta respectively?

The three lines to determine the white point target are not going to intersect at one point, but three (each with two secondaries spot-on and one significantly off). Should the target point be the average of these three, or do certain secondaries have priority given the problem?

[Neil Davidson]

Hi Dingwall,

If you are using CalMAN then you should measure the primary and secondary points which should result in a plot of the actual gamut of the display. Print this out.

On the print out draw a line from red to cyan, green to magenta and blue to yellow as previously described.

Depending on the display you may (as you suggest) find that some or all of the secondaries lie outside of the gamut triangle defined by the primaries. On some displays you can use the colour management system to correct these with respect to D65 (line from primary to complementary secondary passes through D65) and to the gamut triangle (secondary lies on triangle). On your display I am 90% sure without checking that this is not possible and therefore the steps taken so far are pretty much just for information.

We now turn to the D65 point. As you correctly suggest it is possible that the lines from the primary to complementary secondary do not all intersect at the same point resulting in a sort of mini triangle being defined. Your task is to find the white point within that minimises your measured DE of white whilst trying to balance out the error in the primary->secondary lines. My advice is to start in the middle of the small triangle and take it from there.

Hope the above makes sense - it's a lot easier to draw than explain!

Neil

[dingwall]

Don't worry, I'm perfectly with you except for one point...where the target secondary points should be.

I don't have the actual values to hand, but on the attached image, imagining red and blue are measured on-reference and green as marked, then is the target yellow the upper or lower point? Must appear a silly Q when you know the answer, sorry!

The display has colour management on/off in the user settings...what does this do, is it worth measuring with it on/off? Subjectively, it does more harm as it does good.

Thanks again.

[Neil Davidson]

Unless you have a display with a full colour management system then the target for yellow is largely a moot point. You just need to measure what the display can do and try and compensate for it with the whitepoint adjustment. If you do find that the measured primaries are altered during the goal is always to minimise dE with respect to the colourspace in use. Secondaries should lie on the gamut triangle on the intersection of a line drawn from the opposing primary through the whitepoint.

You may as well do some measurements with the colour management on and off if only to understand what it does.

Neil

[Mr.D]

I caled my 42PHD8 and have similarly oversaturated red and green ( I'll dig out the spreadsheet and post it). However the picture itself is very acceptable I certainly wouldn't describe it as terrible ( incidentally I couldn't get the color axis controls in the service menu to affect any inputs ( I suspected as much from vga , hdmi but was a bit puzzled when they did zip for component)

My HS20 projector (pdf is on here somewhere) was also less than steller with the primaries but again I don't have any real problems with the picture. And as I colour grade motion picture film for a living...etc etc.

http://www.avforums.com/forums/attac...1&d=1164317457

[dingwall]

After much experimenting, I found that the slight tweak of extending the white point of the 'normal' colour preset to below 20IRE (it went all over the place below this at default settings) gives the best subjective picture. Colours didn't look bad, and things don't look ridiculously green, maybe even verging onto the good side, but they don't look great. If my eye is noticing problems, then I'm not happy. The best PQ for me personally is when I don't notice the PQ.

Off topic, but I took radical, if not heretical, action and paid a little bit to swap the screen for a Hitachi 37" LCD. Now, I was all ready to send this back too, 'cos LCDs are surely rubbish - but all I can say is wow! Out-of-the-box colours look truly excellent, and the picture is genuinely superb. Black level isn't as good as the 37" Panasonic, but it's about the same as my 50" Panasonic plasma. I think the most amazing thing, is that more than 4 feet away, I would swear this was a plasma. The viewing angle, motion, everything - rocksteady and not like any other LCD I've seen or own. I'm gobsmacked. Football was slightly laggy and smeary on the Panny, but is crisp and clean on the Hitachi! I've seen those Sony X 40"+ screens and they are terrible...if Hitachi made bigger LCDs, it would be very interesting versus plasma. I should maybe post this in the appropriate forums, I'm not sure many plasma-philes would believe it until they saw it. I wouldn't have either tbh.

Next step is to swap the 50" Panny for a Pioneer!