CurtPalme.com Home Theater Forum

[perisoft]

I decided to check the G2 on the ToastyBarco last night, and lo - all three tubes had incorrect G2 according to the LEDs on the neck boards. The R and B tubes wanted to be bumped up a tiny bit, but having gone through the G tube's entire range a few times - first, slowly, and then VERY slowly, and then, my-god-I-never-want-to-see-a-tiny-screwdriver-again slowly, I was unable to make the G G2 LED turn off. God knows the green tube's output is bright enough, and it visually appeared to adjust smoothly all the way through the range, so I'm not sure what's wrong.

In the end, I set the G G2 at the point right before 'black' rasters became visible on the onscreen text (which I think is 50b/50c). It wasn't noticeably different than before.

Any advice? The projector seems rock solid aside from this, haven't seen any fishy behavior in probably close to 100 hours of setup since I got it.


The really interesting bit is that I finally got the thing to project *white* after this. I don't think it was because of the G2 adjustment though; I just started from scratch and ignored the default gain settings. What's odd is that, with a burned blue tube and G2 set correctly on blue and red at least, you'd think that you'd need to pump up the blue or cut back the red - particularly with a ROSCO filter on the blue tube to cut out the burn yellowing (actually greening).

And through the last few setups I'd done, I'd had the gains like:

R: 40
G: 55
B: 65

...and C/B at 50/50.

But I went back to a gamma test image I made, where I interleaved R/G/B steps with on/off dithered R/G/B patterns of equivalent average brightness (multiple brightness level equivalents of the 50/50 hash commonly used to set gamma on CRT monitors).

In the middle of the image, I've got two gray ramps done that way, and I noticed that the dithered parts of the gray ramps (full on/off) pixels were actually pink, even though the solid bits weren't. Obviously, that ain't right. So I thought, what the hell, and turned down the R gain until the pink went away.

A bit more tweaking, and I had gains like:

R: 19
G: 60
B: 65

...and C/B at 65/35.

There was still a yellow cast to the whites, though - but it was VASTLY closer! So close I could *taste* it! But this damned yellow haze wouldn't go away, and despite my blue filters I thought maybe draganm actually was right and I'd never get white... hmmm...

Then I remembered that I hung up this hunk of white fabric for my test screen when color balance was far away in my mind. I took a couple of sheets of paper up to the screen, and - cue the cathedral choir - WHITE. White that stayed white in comparison to the CRT monitor across the room. White that was even whiter than my white cat or my attempts at rapping.

Hot damn!

I fired up 'Stranger than Fiction' which is nicely shot, and relies heavily on bright whites and grays, and it looked fantastic; suddenly, Will Ferrell's teeth didn't make him look British.

So, today, I'm off to JoAnn fabric to get the whitest damn fabric in the store to replace my temporary screen.

But I find it odd that with a burned blue tube, and the G2 set correctly (aside from the caveat above about the G) that the main issue with white balance was that the red was completely blown out, not that the blue was too dark. And with the setup as-is (which might even have the red a bit low because I tried to compensate for the freaking yellow screen) even just the red in the nokia test red/green/blue color bars looks far more natural in terms of brightness relative to g/b.

Any comments on what's going on? Did I miss something basic earlier on that's making this look like the problem? Or did somebody put ephedrine in the red tube's glycol before I got the PJ?

[Ile]

Those led adjustments tend to drift, then led won't go off.

Forgot those leds and adjust G2 using video black method.

Of course you can check/recalibrate amp using scope and then adjust led to go off, but it won't chance anything...

[perisoft]

Grabbed some blackout cloth, some blazing-white something-or-other cloth (all the middle-aged women in the fabric store must have been wondering what this 20-something dude was doing carrying a bat of white fabric around, holding it up to all the other kinds of white fabric... as if I don't get enough sideways glances when I go in there as it is) at JoAnne's today, along with a few yards of black cloth (not velvet, but it'll do for the moment) to put on the ceiling above and wall behind the screen for my test setup. Should help me evaluate black level and white performance drastically better than before. I'm jazzed after seeing what stuff looked like with the balance adjustments, ROSCO filter on the blue, and an actual white screen surface.

Hopefully I can get a chance to set it up tomorrow evening.