Originally Posted by strutter
i noticed a little more detail in shades of black, and a little more depth in the picture. blacks are a little blacker now,
I'd imagine your calibration guy diddled with the iris min/max/gain in the service menu, just as owen and others here have done to their own sets. I'm playing with those settings right now myself. You can get a small improvement over factory settings, which are generally going to be pretty conservative and oriented towards looking good under bright lights on a factory floor.
I'll throw my iris adjustments into my post, but not many people are going to be willing to mess with the service menu, and that's a good thing.
my set pushes blue in the blacks. after calibration its only noticeable when i turn the set off and the bulb is still on for a short time.
Yeah, these sets actually seem to push blue in the blacks all the time, but since the whole range is too blue from the factory, you don't notice it so much. When all but the low end is brought into line, it's harder to ignore. You'll probably see that on, err, the left end of one of the graphs your calibrator gave you. Conversely, I think red is lacking at that point. There's only so much you can do.
I think I personally would probably prefer that blue and red gently curve away from green toward the left, instead of taking sharp turns up and down at the last step. That, or allow the whole range to be off-hue, but consistently. It'd be less noticeable, even though it's less correct. If I ever get a calibrator out here, I suspect I'll ask him to do that, or at least talk over the options.
the most noticeable change was in shades of red and green.
Yeah, that's the hard part to get right without diddling with the actual color matrices in the service menu. The color space is rotated, but not around its center, so you need a full matrix adjustment to correct it. It also might be skewed a bit.
My own hue and bias/gain adjustments are gross hacks that attempt to approximate a proper color space rotation. However, the hue adjustment doesn't rotate the space around the right point, so "G 2" simply puts two primaries in marginally-wrong places instead of one primary in a really wrong place. Something like that. I try to compensate by changing the ranges (bias/gains) of the two wrong primaries a bit, but that's not really the appropriate math and so it's only an approximation and it has flaws. Really, my method just trades a couple of glaring flaws for a handful of minor flaws.
A proper calibration is going to do its best to reduce or eliminate all flaws, instead of trying to equalize them. The only problem with calibration is that there may be certain flaws (e.g. blue blacks) that cannot be fixed, and, in contrast, those stand out next to the rest of what is now near-perfection. Calibration is still better, though.
also, although i had checked convergence and it looked right to me. UMR said it was off. he adjusted it and said mine was now spot on, something he had only seen a few capable of doing. a side effect of this is that i have slighty more over scan but i think its a good trade.
If you have one panel a little misaligned, you might benefit from a 1-pixel shift, but not many people are going to be that far out of alignment. One of my panels is about a third of a pixel off on one side of the screen, but hey, that's no worse than LCD/plasma direct-view.
By the way, you shouldn't have gotten any more or less overscan from a convergence adjustment. It's just a wholesale pixel shift left or right (or up or down). What you lose off of one side, you gain off of the other.
i wasn't sure i liked the post calibration picture. i had become so accustomed to the way it was. however friends said it looked better to them so i gave it some time to grow on me.
We all fear change.
I'd be worried, but at least I'd know red would finally look right. Red is the one main thing I've never been able to get looking good with user controls. It's always in need of a little push towards orange that I just can't give it without messing everything else up even worse.
later, I'll see if i can scan the pre and post calibration sheets and post them.
That would be awesome. Are the "pre" sheets actually based on the settings you had on the set when he showed up, or did he reset to the factory defaults before taking those readings? I'd love to see what my settings actually look like on graphs, to see if I actually made anything better, or if it's all in my head.