More Paterns & Info / Gray Scale stuff

[Alaric]

Hi Guys,

In my ever trawl for a better picture i've come across this site :

http://www.walvisions.com/

The convergance pattern looks curious and worth a try.

I've also been playing with some monitor calibration bits and xyY data with gray (grey?) scales....may be getting tempted with a Spyder2, but my old Spectral Images monitor calibrator works with OptiCal to give the data, how acurate is another issue....I also have a Barco CAST item that was an Ebay special, which under the hood is an Ocean Optics USB2000, which i guess is super acurate, but is more lab based kit....Light Wavelengths to anything actualy useful anyone

Cya,
Lee

[Gary Lightfoot]

Hi Lee,

You could try Optical and see how the xyY data looks when used with the Excel spreadsheet. If you can get one of the lower IRE fields to the D65 coordinates and it looks grey with no colour tints, then it might work OK to do the whole greyscale. I did start a thread on the SpyderTV both here and on the DLP forum so you'll find either the spreadsheet itself or links to some threads (on avs IIRC) which will take you through basic calibration. I think you have to tweak the gamma on a CRT to help with getting the RGBs to D65 so it's a bit more involved than what I'm used to with a FPD, but maybe Gordon or one of the other CRT gurus will poke their nose in.

Gary

I just recently built my own convergence only DVD. It has big squares with dots inside and very little squares with dots inside... and diagonal squares with dots inside... Worked great.

That walvision grid is interesing tho, I guess because you can see the colours seperately. It might be good for blue/green.

[Alaric]

Hi There,

My current fave on my BG808 via my htpc for convergance is the philips software patern that is a load of magenta and cyan crosses spaced one after each other and alternating colour. It means you align the edges of each cross and hence can completely see each colour unlike an overlaping grid that once your within the overlap you can't see what is going on.

The added advantage with this patern is that should there be an error (ie in a corner) then you can even the error out.

There is also a full white grid which is nice for checking.
cya,
Lee

[Tom.W]

There are some good HD test pattern links if you follow the service manual link I posted a couple days ago.I also posted the link I found here over there and wanted to thank Alaric !

[Alaric]

Hi There,

Tom, I take it your talking about the .ts files....a nice and chunky 150meg download for the full lot....What do you view them with, as i only briefly had a look on my PC and didn't have a chance to look at them on my HTPC, but WM9 didn't seem to have the codec ???

The prog with them seems to be a file splitter, doing image and sound ???

I had a quick flick through the EvilTwin website and it looks quite techie....I take it that its focus (pun!) is a dual stack and heavy on the modifying side ???

I wish my soldering and electrical engineering was better.

Gary - The device is a Sequel Imaging one, not spectral, it seems to be a for runner of the Spyder, called a Chroma 3. It was rebadged as everything BUT a Vauhal Astra from searching. Playing with it and some test paterns there seems to be some variation between readings, particularly at low levels, however doing some AVS research (beast of a site that) it also apears that the Spyder is unresponsive at low levels (IRE30 & less) though that is what the Spyder2 brings to the party along with LCD/DLP tags.

I have also found that my initial testing was flawed in excess light from other areas of the screen - my test area was relativly small in relation to the screen and there were other programs/icons showing, so light polution is an issue (probably not helped by the fact that my HTPC source is also the test machine so i need to switch between apps). There seems to be some argument on which way to point the sensor, ie at the screen (low light,shadow problems but calibrates for screen push) or direct in the light stream (brighter, not at the focal plane, ignores screen colour).

The bonus news is after speaking to Barco they have sent me the lost Calibration file for the CAST, which means i can get colour data from it, though on a quick look it gives U' and V' as axis data - I presume, equates to XY, though need to read the PDF manual to check interpretation and which colour space it equates to. Both the P series and D series software which i have come with pdf manuals which include some colour theory if anyone is interested. Shame i don't have a suitable projector wall/controler as it looks like it can do some funcky auto cal stuff.

Cya,
Lee

[Mad Mr H]

Originally Posted by Tom.W

I wanted to thank Alaric !

PRESS THE "THANK YOU" BUTTON THEN TOM!! , its a great idea.

And it shows in your profile too.......Both being thanked and thanking people (glad I spelt that right )

[Alaric]

Hi There,

Not quite sure of the reasoning on that button...I guess its a nicety.

Can we save them up and get cookies or something ???

Is there a prize for the most helpful, or most polite person ???

cya,
Lee

[cosaw]

Is there a prize for the most helpful, or most polite person ???

Now that sounds like a good idea!

kiss my *ss

[Tom.W]

I was wondering what that button was for...

[Gary Lightfoot]

Hi Lee,

This site might be useful for converting between colour models:

http://www.brucelindbloom.com/

Top one has the converters.

Most colorimiters are less accurate below 30IRE as the light level dimishes, but one way to get round that is to move the sensor closer - with a CRT that can mean reading more green than red and blue so trying to get a good reading off the screen might be a problem. I wonder if using a large piece of paper closer to the pj will help - focus isn't important so you might get a better reading provided you're getting a good example of the overall colour balance rather than one of the primaries.

I would think you'll need to have the contrast modulation cct on or it will be even trickier to get a good reading for any of the grey scales. You could PM Gordon and see if he has any tips.

Gary

Colorfacts