Hmm. I always used ~0 and ~200 black and white targets for every devices with higher native black and lower peak white luminances. It never caused a problem like this (nor any other error I noticed). I actually thought it's better to set up the parameters like this, so dispcal won't lift up the black level in case the random measurement error caused a slightly too high black level reading when dispcal establish the "native black level" (and the same for th other end of the scale in inverse with similar small dynamic range clipping). Good to know it's bad idea. It's CCFL but I don't think the brightness drifts that much. It was properly warmed up. And I left the i1d3 on it for ~10 minutes measuring the white point before I started the calibration (it's not as hot as the desktop monitors, but I had the time, so let it acclimatize...) and the white luminance didn't change too much (otherwise I would have gotten scared about the hardware and/or tried to adjust the luminance back to the desired point). This is an old laptop and it was calibrated a few times before. I just wanted to re-calibrate it (the last calibration was about 6 months ago). The contrast wasn't clipped like this before. But now, I repeated the calibration twice (with slightly different settings, like gamma 2.2 instead of 2.35) and I always got a clipped LUT like this. I will give it a try with "native" black and white luminance settings tomorrow.