Issues arising from my latest calibration

[Anonymous]

I calibrated an RS55 yesterday for a client who runs a colour grading business. He uses a MacPro and DaVinci software and has a 3 month old Flanders Scientific monitor. Prior to booking me had purchased an Eye-One Display Pro and Chromapure in order to try and calibrate the RS55 himself. He could only get so far with the RS55 without the knowledge and experience so he called me. In the meantime he took readings from his Flanders Scientific monitor and it showed a significant lack of red in the greyscale. He thought this was odd because the monitor was supposed to be factory calibrated with a SpectraScan PR-655. So he got another Eye-One Display Pro, thinking that it must be at fault. However, this too showed the same readings resulting in several DEs of error. When I calibrated the RS55 (using my Eye-One Pro/Eye-One Display combo) we compared the two pictures. There was a visible difference. The client couldn’t understand why, if they had both been calibrated to the same standard that they should look different. I took the opportunity to take a greyscale reading from his monitor and low and behold I got the same results as he did. Thinking that there must be something wrong with his monitor he contacted Flanders Scientific. They immediately became defensive, saying that they had calibrated with a PR-655 and that the Eye-One Pro was a piece of you-know-what and that’s why there was at least 4 DEs of error on the readings I had taken. They sent their original calibration file over and it looked fine (with perhaps a slight oversaturation in the colour gamut on the Red, Magenta, Blue axis). When asked why 4 separate meters gave the same readings they then started saying that even if the projector and monitor were calibrated to the same standards there would still be a perceptual difference. Does anyone have any light to shed on all this? I can accept that the PR-655 is a fantastic piece of kit, but really, 4-5 DEs better across the whole greyscale? Why would 4 separate meters give practically the same results? I’m scratching my head.

Topic

[Anonymous]

So your i1 pro agrees with the d3 units? The d3 units by themselves are using calibration tables and always have a potential for errors there.

If everything is calibrated the same, then the weaknesses of specific displays show up easier …

So if one TV has crappy black levels, you can’t fix that and it might stand out.

[Gregg Loewen]

Seems to me like 1 of the 2 isnt calibrated (despite what the company says).

The flanders does have a WB adjustment on it. This would help to get it closer.

Even when calibrated, the 2 displays will look perceptually different due to differences in contrast ratio etc.

[Anonymous]

Is the Flanders Broken in before it is calibrated at the factory? How many hours has he put on the unit over those three months? Could it not just have simply drifted 4 De?

[Gregg Loewen]

hi Charles. Good points and questions.

[Anonymous]

My company owns 2 Flanders LM2461W monitors and I had a similar problem. These panels do not work well with the standard lcd profiles for colorimeters. I first used my Display 3 Pro to balance the gray scale. Chromapure showed the gray scale to be flat, but the picture was screaming red. When I got my JETI 1211 I attempted the calibration again and it was perfect. Since then I haven’t had any other issues with the Flanders. The gamma is extremely accurate and chromaticity falls right in the boxes without any cms system.

[Anonymous]

in conclusion what do you say now?

[Anonymous]

I’ve calibrated three, maybe 4 displays over the years where the owner had an i1Pro and was either using it for calibration or was using it to profile a colorimeter. In each of those cases, the owner wasn’t happy with their calibration and wanted to see if what they were doing was procedural or software or meter related. When I calibrated with a Konica-Minolta CS-200 (a hybrid meter… some characteristics of a colorimeter, some characteristics of a spectro meter), the result was perfect, visually. My measurements of their calibrations easily showed the problems they were seeing, while their own re-measurement of the displays looked good with their software. So the problem was clearly meter-related since they weren’t doing anything but taking measurements and comparing those to my measurements. Two of them were pretty pissed-off that they had spent somewhere over $1100 for meters that would not calibrate their displays correctly. At the time, I didn’t think to run a spectral distribution of the light emitted by those displays to see if there was anything unusual, but my guess would be that the pro monitors mentioned in this thread have a backlight source that has something different/unusual about the spectral distribution of the light being measured… for example there could be too many long wavelengths of light (towards infrared) or too few… if the meter measures those wavelengths, it affects the readings but the long wavelengths are invisible so you would have a condition where the display measures right but seems to be lacking red when the meter is too sensitive to infrared (or if the display emits too much infrared and the meter is “normally” sensitive to infrared. Another possibility is that the meter does not see the longer red wavelengths, even though those wavelengths haven’t drifted into the “invisible zone” quite yet (as an example, CD players use infrared lasers, but the wavelength of light produced by the laser is short enough that it is still visible to us so it’s “visible infrared” rather than infrared being invisible as we expect. Same sort of thing can happen at the blue end of the spectrum where, when wavelengths get short enough, they go invisible and become ultraviolet… there’s a transition zone there also where the wavelengths are short enough to be considered ultraviolet, but the are still visible. My guess is that if you use a high-end meter to measure that pro display’s light spectrum, you’ll find it has either much more or much less light in the longest and/or shortest visible wavelengths. ANd when you measure consumer displays of the same type to see their spectral distribution (how the wavelengths are distrubuted), you’d get a very different result. And I’d also guess that somewhere in the specs of the i1Pro, there’s some sort of spec that define’s the meter’s accuracy in terms of the wavelengths included and excluded from that accuracy statement. Somewhere around 450 nm and lower, you are in ultraviolet, and somewhere around 940 nm you’re into infrared. The better meters used for video display calibration will mimic human visual sensitivity and not respond to the shorter wavelengths nor the longer ones while less expensive meters… who knows?

[Author]

Replies