A Low Cost Instrumented FPTV Gray Scale Tracking

Posted by Guy Kuo [IP: 207.207.88.144] on May 30, 1999 at 20:36:49
Using Mozilla/4.5 (Macintosh; U; PPC):

A Low Cost Instrumented Gray Scale Tracking Method for the Do it Yourselfer

Guy Kuo

Before I begin, let me STRONGLY emphasize that a professional with a colorimeter can do gray scale setup more efficiently and accurately than this method. If you can at all have it professionally done, let a pro set it up. This isn't something for most people to try. You can really mess up your picture.

For those do it yourselfers on this board who are setting their own gray scales, I submit the following, low cost instrumentation method of refining a gray scale on a FPTV. This could be done on a RPTV but you'd need direct access to the lens assemblies. I've tried doing this by measuring light at the screen but the light is too dim to produce sufficient digits of precision using this crude but effective equipment.

Materials:

1. Silicon Solar Cell (just a single un-encased cell with leads attached to it from Radio Shack will work)
2. Lens cap which fits your FPTV lenses.
3. Digital Multimeter
4. 1 uf capacitor
5. Some tape

Premise:

The current output of a silicon solar cell is related to the amount of light striking its surface. It is possible to measure the current output from a cell to measure how light output tracks input signal. If gray scale is consistent throughout the range from black to white, then the ratio of current produced light from the three color guns should remain constant. Imbalances in the ratio indicate gray scale tracking errors which are easier to measure than to see visually. This yields a low cost, crude but effective means of checking and adjusting gray scale tracking. It takes MUCH more work than a colorimeter but should yield results which are more consistent than simple visual methods. At the low end of the IRE scale (darker) it is difficult to visually judge color differences between an optical comparator and the screen. Measurement of light output from each gun using a simple apparatus yields superior results.

Caveat:

This method does NOT measure out and produce an instrumented D65 color of gray. The reference white is still optically set. However, this does yield a gray scale which accurately tracks the selected color of white.

Method:

Build your optical sensor by placing the silicon solar cell inside the lens cap. Route the output leads out a hole in the cap. Seal that hole to prevent light from leaking into the cap. Attach the capacitor in parallel with the solar cell to act as a filter. Attach sufficient lead wire to reach from your lens cap sensor assembly to your digital multimeter.

You must now set your color of white and measure the red, green, and blue CRT current outputs. Use an optical comparator, and a 90 IRE window pattern to guide your adjustment of the projector's gain controls to match the reference white of your comparator. It is absolutely important to get the reference white as neutral as possible. If you don't have a comparator, you can try using a Kodak gray card and a Daylite fluorescent bulb as a white reference. Hint: you may find it easier to set the reference gray by viewing a black and white scene and adjusting the bias controls until the overall scene gray matches your reference. (You'll adjust bias again later but this will get you to the first measurement step).

Place your measuring sensor on each lens and then measure and record the red, green, and blue current outputs while the projector is displaying a 90 IRE window pattern. We'll refer to these numbers as the RedRef, GreenRef, and BlueRef. Use the current range on the meter which gives you the most digits of precision. This measurement will serve as the guide to the light output ratios that are needed to reproduce that particular color of white.

Set your projector's brightness control to its center position. Use the "Black Bars" pattern in AVIA to verify that green gun cutoff is correctly set to make the black raster just black. If needed adjust green CRT bias to make it correct. It's easiest if you look directly into the lens while making this adjustment.

You should probably leave the green gain alone. (You could also set things such that the maximal contrast setting of the projector control just makes the green gun bloom, but that entails more work).

From here on do NOT adjust your brightness, contrast, green bias, or green gain controls until after the entire procedure is complete.

Display a 20 IRE window pattern and measure output of the green gun. Again use the current range on the DMM which yields the most digits of precision. Record this as Green20. Switch to a 100 IRE window pattern and measure the green gun again. Don't take too long measuring at 100 IRE because the phosphors temporarily quench (fade) slightly over time. Record as Green100.

(BTW: Ever notice that the AVIA DVD places the 100 IRE window, and a copy of the 20 IRE window and Black bars patterns in sequences? This was so a calibrator could easily switch between these three using just the prev & next buttons of the DVD remote while doing gray scale.)

Now you get to do some calculations. You already have the RedRef, GreenRef, and BlueRef reference currents measured in the first step. You also have Green20 and Green100 measurements. It's now simple ratios to calculate the desired Red20, Red100, Blue20 & Blue100 measurements to achieve.

Predicted Red20 = Green20 * RedRef/GreenRef
Predicted Red100 = Green100 * RedRef/GreenRef

Predicted Blue20 = Green20 * BlueRef/GreenRef
Predicted Blue100 = Geeen100 * BlueRef/GreenRef

First the red gun. Move the sensor to the red gun.

Repeat the following a & b until both 20 and 100 IRE window measurements match the predicted values as close as possible

a. Display the 20 IRE window pattern and adjust red BIAS until the measurement equals predicted Red20.

b. Display the 100 IRE window pattern and adjust red GAIN until the measurement equals predicted Red100.

You have to do this several times in order to get both ends correct.

Next the blue gun. Move the sensor to the blue gun.

Repeat the following a & b until both 20 and 100 IRE window measurements match the predicted values as close as possible

a. Display the 20 IRE window pattern and adjust blue BIAS until the measurement equals predicted Blue20.

b. Display the 100 IRE window pattern and adjust blue GAIN until the measurement equals predicted Blue100.

Again, you have to do this several times in order to get both ends correct.

Now you are done setting a gray scale which is consistent at 20 IRE and 100 IRE. Depending on how well your projector tracks the values in between, you may or may not be done.

Measure the output of each gun as for IRE windows 20 through 100.

Use the green gun measurements as the guide and calculate the predicted red and green values for each IRE window. Also calculate and plot the percentage error for red and blue at each IRE. (It's actually instructive to do this plotting with your old projector settings so you have a point of reference).

Adjust the red and blue guns as needed to get the curves to track best. You will probably not get an exact match at all IRE levels, but you should get better results than doing it visually.

Once, more. A professional with a colorimeter will probably laugh at this method, but it does produce more consistent gray scale tracking than visual methods and gives you a way of gathering data about your projector's tracking characteristics. It also could serve as a long term measurement of how your phosphors are degrading over time if you measure the outputs using the same settings and instrumentation.

Anyway. Have fun.

Guy Kuo
www.ovationsw.com

Follow Ups:

Return to the new SMR Forums Menu

Design & HTML © SMR Home Theatre, Images © SMR Home Theatre cannot be reproduced without permission.  The images on this page are digitally watermarked.  New forum messages should be posted into SMR Forums v2 - http://www.smr-forums.com/

	The Web   SMR Archive