CurtPalme.com Home Theater Forum

[Satanier]

I'm having some serious geometry issues, bad enough its bothering me when watching video.
I tried EVERYTHING in alignment, I can't fix it. Source is 1920x1080 from HTPC, windows desktop properties.


Here is a picture::: Look at the circles. There's almost no distortion from the camera here, its just as bad as it looks. The center is an egg and all 4 outer circles are different.









Also, two other things:

1. Is it normal for the corners to be dimmer than the center?
2. I still cant get the left //// right side of my green, and to a somewhat lesser extent blue, to focus right, and its really bad in the upper left corner. In the upper left corner theres some kind of distortion that pulls things out of square, like when dragging a computer window up to that corner, and the focus is just awful there. I followed instructions to neutralize all electronic focus controls, then adjusted the focus pots for perfect center focus, then I checked beam astigmatism, then I adjusted electronic focus, it just wont dial in on the left side especially, and the right as well. Upper left corner is worst of all.



These issues are really bothering me, and preventing me from fully enjoying my new toy. Any insight would be awesome!

[kschmit2]

did you start by resetting Geometry, Convergence and Point Convergence?
And did you set phase properly?

[Elaine Benes]

What is your screen size and shape ? How far from the screen is your projector ie. screen center to green lens center ? Have you measured to determine it is EXACTLY perpendicular to the screen surface ? What are your lens flapping tabs set to for each color ?

When you did your initial Signal Entry for this input, what information did you choose, ie. Default, Temporary, etc ?

[dropzone7]

Of course others here know better than me but at first glance I would say that your linearity is off. Look at the height and width of the blocks on your top row compared to your bottom row. If you have converged to that geometry then of course all three colors are off. You need to turn off red and blue and get your linearity correct for green. Then go back and work on converging red and blue to your corrected green.

Yes, it is normal to see a somewhat dimmer image near the edges of the screen compared to the middle. This is especially true if your screen surface is "hotspotting" at all. It might not be that your edges are dimmer but that your center is actually brighter due to this hotspotting. I use a piece of Designer White Laminate for my screen and have observed this in my setup. As for focus, it takes a lot of practice and careful setup to get corner and edge focus looking as good as center focus. This is especially true if you are maximizing your rasters. Corner focus will suffer a bit so it's a trade off of a larger raster and image size, utilizing more phosphur or a smaller raster and image size, achieving better overall focus. The choice is yours but personally I would rather have a big screen with good center focus than a smaller one with perfect focus. After all, most of what you look at on screen is in the center and not at the far edges so chances are that you won't notice less than perfect focus at the edges during most movies.

[Curt Palme]

I agree, if the pix above is not distorted, then the V lin is off. Also check the lin balance, done by pressing 'ctrl' and 'lin.

[Curt Palme]

I agree, if the pix above is not distorted, then the V lin is off. Also check the lin balance, done by pressing 'ctrl' and 'lin.

[Satanier]

[Satanier]

[Satanier]

[Curt Palme]

Have you seen this?

http://www.curtpalme.com/CRTSetupGuide.shtm

Good work BTW on that image with no point. Let that picture be a benchmark on how accurate you can get convergence without a bit of point. Kal, take note, might be an idea to post that pix in the NEC PG/XG section regarding excessive point use. I can write something up...

Satanier, would you mind posting a simular pix of the internal crosshatch pattern?

[Curt Palme]

Have you seen this?

http://www.curtpalme.com/CRTSetupGuide.shtm

Good work BTW on that image with no point. Let that picture be a benchmark on how accurate you can get convergence without a bit of point. Kal, take note, might be an idea to post that pix in the NEC PG/XG section regarding excessive point use. I can write something up...

Satanier, would you mind posting a simular pix of the internal crosshatch pattern?

[Satanier]

[Curt Palme]

From the pix it looks like it's your main V lin that's off. Shrink the top, stretch the bottom.

Generally the lin balance controls need to be set to 0 unless you're mounting on some screwy angle.

You can also use the point on green to move the raster a bit in various areas, then adjust the R and B on top of the green image, but try the lin and lin balance controls first.

[Curt Palme]

From the pix it looks like it's your main V lin that's off. Shrink the top, stretch the bottom.

Generally the lin balance controls need to be set to 0 unless you're mounting on some screwy angle.

You can also use the point on green to move the raster a bit in various areas, then adjust the R and B on top of the green image, but try the lin and lin balance controls first.

[Satanier]

[Curt Palme]

You're being very anal about this setup.... that's a good thing..

Yes, V lin can sometimes be close to maxed out, ditto for H width.

I would now use the point controls to tweak the green geometry to perfection, there's no control short of point that will allow the width of the outside edges to be adjusted. Like I said, massively good job on the convergence without using point so far!

[Curt Palme]

You're being very anal about this setup.... that's a good thing..

Yes, V lin can sometimes be close to maxed out, ditto for H width.

I would now use the point controls to tweak the green geometry to perfection, there's no control short of point that will allow the width of the outside edges to be adjusted. Like I said, massively good job on the convergence without using point so far!

[Satanier]

[perisoft]

That's some damned good work, Satanier. I think you get full membership in the CRT club for this effort - not bad at all so quickly. How much time do you think you've spent on it?

I'd be interested to see some closeup shots of the resolution hashes in ntest. It looks like you've got stuff pretty sharp, but of course it's hard to evaluate sharpness of a 1920-wide image with a 1500-wide photo...

[Satanier]

[Elaine Benes]

I'll bet the majority of your focus issues are caused by : "The projector is around 9 feet from the screen or so, I measured only roughly by eye. The green wasn't measured to be exactly perpendicular to the screen, but it's close enough that I don't think it could be causing this."

Pretty much exactly the opposite way you are supposed to physically set an NEC. They MUST be EXACTLY perpendicular to their screen, and exactly centered too...NEC's don't have infinitely variable lens flapping, they DEPEND on exact physical placement in order that the lenses can focus properly over the entire screen.

When you have to re-place your projector after the carpets are cleaned. Measure and place it so it is EXACTLY in the correct orientation to the screen and see if your focus is still an issue...

[Bert Randolph]

[kschmit2]

Can you try the NEC Test Pattern Generator, please, and tell us how the circles look using that tool.

[Satanier]