New Posts  All Forums:Forum Nav:

Circle hatches??

post #1 of 6
Thread Starter 
Hi,

I'm trying to understand whether, when utilizing the AVIA circle hatch pattern, I am expected to tweak my Hitachi 51G500's vertical size (compression?) somehow so that, when measured, 'each' circle is precisely that; a circle.

Pie are squared .. me? I'm just confused.



Joe
post #2 of 6

Re: Circle hatches??

As far as I know with most calibration disc visual patterns you're supposed to evaluate what they look like, not measure them.

Do the circles look like circles? Then you're good to go. If they don't - if they look "stretched" from top to bottom or "squashed" so that they're wider than they are tall, then an adjustment needs to be made.

You don't say what kind of TV you have. Some test patterns aren't applicable to some kinds of sets because the technology is either immune to a given problem or very rarely afllicted by it. (LCD, DLP and LCoS TVs, for instance, won't "bend" the veritical white line in one common test pattern because their pictures aren't created by using beams of electrons, and therefore won't distort an image based on voltage.)

Regards,

Joe
post #3 of 6
Thread Starter 

Re: Circle hatches??

Makes sense Joe.

Let me quote from the ARIA manual:

Quote:
Circles: Perfect circles are a simple way to verify picture geometry, especially if you take a ruler or
tape measure to a Circle Hatch pattern
and confirm that every circle’s height is precisely equal to its
width. This is often not the case, even with fixed-pixel displays like LCD and plasma that otherwise
pass alignment and geometry tests that are difficult for projection displays. That’s because many
widescreen displays are not precisely 16:9 in their pixel count, and to completely fill such a display
with a 16:9 image (like all the patterns in Avia II) the picture is slightly stretched, usually in the vertical
direction. The circles heights will then be very slightly greater than their widths.

This left me unsure whether completely perfect circles were a necessity. It is clear to me though that measurement is something to attempt if you value accurate geometry.

I have an Hitachi 51G500 ... a CRT HDTV projection. Maybe I can refer to it by "CRT" from now on, if that tag helps more than others. (Note to self: write a sig line.. )


Joe
post #4 of 6

Re: Circle hatches??

CRT sets usually have a whole slew of controls for adjusting picture geometry. The circle hatch pattern is provided so that you can get _all_ of the circles as circular as possible. And yes, measuring them is a good idea. Your eyes can be deceived. You may find that you can't get all of them circular, but you can do an overall good job. You may need a manual for your set to find out which adjustments will do the trick, and how to access them.
post #5 of 6
Thread Starter 

Re: Circle hatches??

I agree with looks can be deceiving. Each circle I measured revealed the opposite to how my eyes perceived them. IOW's they each looked larger in the y axis, but in fact were wider than they were tall. Wierd.

So, another Q:

In the end I ignored pure geometry (tweaking circles) and utilized my sets hori/vert size adjustment pots to tweak overscan.

Is it actually possible to set the overscan perfectly, and 'then' somehow tweak circle geometry independantly?

Thanks,

Joe
post #6 of 6

Re: Circle hatches??

Due to the vagaries of CRT electronics, it is usually not possible to get a variety of circles in different places on the screen all perfectly shaped.

In order to get the geometry correct all over the screen, it is usually necessary to have individual controls for each of numerous small zones, perhaps 64 zones side by side in an 8x8 grid covering the entire image area.

For most viewing, such precise calibration is overkill.

Many upscale CRT projectors, have such numerous small zones for convergence.

Video hints: Video Technicalia Made Easy
New Posts  All Forums:Forum Nav:
  Return Home