Improve Your Projector

Now the real fun begins…..

You have done you calibration run, and everything looks good. Its time to start optimizing your projector. This will take several steps. First we will set the black level for each color by adjusting the bias level for that color. Then we will proceed to getting the most contrast and light out of the projector while maintaining the same quality of color balance we achieved in the calibration run. Lastly we can use our calibrated detector to really improve things using CC filters.

Setting Bias Levels
Bias Levels should be set so they represent the projector's lowest possible output for a given color. To understand more about bias and what it does, please read the introduction.

To set bias, you first must put your projector into a mode that allows you to adjust the Red, Green and Blue Bias Values. On the SONY this is accomplished by putting the 10HT into service mode.(Please Click here to find out about the menu system in the service mode).

Since we do NOT want to change the original low setting, we will instead start to use custom 1. Use the 10HT menu to select the Custom 1 color temperature for the input you are using. (S-video, Component A etc) Then return to the RGB Gain/Bias menu and select Custom 1. Now copy the Gain and Bias settings for the low color temperature into Custom 1, so that it is an exact copy. When you are finished, the custom 1 color temperature should produce the same color display as low when selected for the input you are using.

Now go to the IRE Windows menu in AVIA. Choose IRE 0. In order to keep the 0 IRE window displayed during this procedure, you will have to periodically "skip back" to the beginning of the test pattern using your DVD player controls.

Next, place your green filter in your detector. Make sure it covers the entire surface of the photocell and that no light leaks around the sides.
Now reduce green bias to zero and wait for DMM display to settle. Now start to raise green bias until there is a noticeable decrease in the measured resistance. Make note of the Bias setting where the first decrease in resistance occurred and repeat the procedure a few times, each time making note of the point where the DMM display decreases. Since most DMMs take a while to track resistance at high values, you may have to change the bias values very slowly to obtain an accurate reading. By repeating the procedure a few times, you can make up for the DMMs lack of speed and get a very good idea what value of bias begins to decrease the measured resistance. That value is the proper bias point for that color on your projector.

Once you have set the proper bias point for Green, place the red filter in the detector and repeat the procedure, this time modifying the red bias value.

Finally, place the blue filter in the detector and repeat the procedure, this time modifying the blue bias value.
Do not be alarmed if you get significantly different resistance readings for each of the 3 colors. This is normal. The point of this exercise is to set bias to the point where the measurements just begin to CHANGE, NOT to get the same reading for each color.
Write down your Optimal Bias settings from the 10HT RGB Bias menu here

Gain Adjustment
Okay, now that we have set or bias points, we can now adjust the gain settings to maximize the total brightness of the image and make sure that our color balance tracks properly. .

First we have to set a baseline by maximizing the gain on the projector’s "weakest" color.
This can be determined by looking at your gain setting for low. Which ever color has the highest gain setting for the low color temperature is your weakest one. On the 10HT this will always be red.

First, make sure that the custom 1 color temperature is set for the input you are using.

Ok, now go back to your projectors Gain/Bias Menu. Set Red to its highest level. On the 10HT this is a value of 255. If you wrote down the gains for Green and Blue that SMART calculated as part of the "Calibrate" step, enter them as well.
We will now begin an iterative process that will allow us to improve our settings.

So lets do it.

The Data Set (Improve run)
Before starting any data run, you should always verify your White and Black levels using the temperature you are modifying. Make sure the input you are using is set to "custom 1" color temperature. Then, use the AVIA test disk to set black level (brightness) and white level (contrast) as described in a previous section. Be very careful to not set contrast too high.

Next, write down the Gain values you will be using for this run in the spreadsheet in the spaces provided (For data set 2 these are cells D58, D59 and D60) Now go to the AVIA IRE window menu. Put up the 0 IRE screen.

Put the red filter in the detector and take a measurement of the photocell resistance as measured by the DMM. Remember that the value may take a significant time to "settle" when measuring low light values. Now record this resistance value in K½, into the spaces for Set #2 (starts with F58)

Once you have taken the measurement for red at 0 IRE and input it into the appropriate cell in set 2, advance the DVD one chapter. This should now display a 10 IRE window.

Take a new resistance measurement using the 10 IRE window and input it (in K½) in cell directly to the right, just as you did for the calibration run. . Continue advancing to higher IRE levels (20,30,40,50…) and taking measurements, each time putting the new measurement in the cell to the immediate right of the last cell until you write down the last value for IRE 100.
Then remove the Red filter from the detector and replace it with the green one. Use the AVIA menu to put the 0 IRE field back up again. Then, in the same fashion as above, record the resistance value (in K½) in the appropriate space from Green. Advance through the IRE levels to IRE 100, taking a resistance measurement on each IRE window and recording it to the immediate right of the last cell.

Finally, replace the green filter with the blue one, and take a resistance measurement for each IRE field just as you did for the other two colors and record them in the cells marked for blue.

Okay, now go up and use the Previous/Next button so that set 2 is displayed (Cell D$ has a 2 in it)
Line 45 of the spreadsheet will tell you what to do next with the various gain settings. It could tell you to lower contrast and/or to change gain values for the next run. . It will also tell you when you are done!

For instance if the advice is to "change green gain by –39% to 173" then enter the projectors gain/bias menu, and change the green gain on the color temperature you are working with (custom1) to 173. (Pretty simple huh?)

The spreadsheet may also recommend a CC filter. Ignore this recommendation for now, we will come back to it in the next section.
Once you have modified the gain values in the (Custom 1) color temp setting to match those recommended by the spreadsheet, write down the new gain values in the spaces for data set 3. Check White level and black level again and then start collecting your next set of data, recording your results in the spaces for data set 3. When you complete the set, use the Precious/Next button to select data set 3 (cell D4 should have a 3 in it)

The spreadsheet will make new recommendations based on the data for set 3 in line 45.

In the process of adjusting the various gain settings you may get a message to lower the contrast. If you want to understand more about what this message means and some of the factors that can cause the message to appear, click here.

You should continue this process, each time using the next data set down (4,5,6) SMART says you are done. At this point, your Red, Green and Blue curves should be overlaid pretty closely on top of one another. (If your wife really wants you to come to bed, and you're not done, it's OK. You can start the next run on another day. No problem.)

The set of values in Custom 1 will now represent your optimum settings without a CC filter. Write them down.
Now, as a final double-check, Look now at the contrast ratio displayed in cells K17 and P17. These are your calculated contrast ratios for your projector with the new settings.

Use the previous/next keys to look at Data set 1. (your calibration set) and look at K17 and P17 for that set. With any luck, they are much lower numbers than the ones calculated for your optimum set and your Contrast has improved

In some cases, you may find that your traces don't follow the desired smooth curve very well.

In those cases you might want to visit, Advanced topics - 3D gamma adjustment.

Gain Adjustment with a CC Filter
What Kind of filter do I need?
Okay, up to this point you have
1) Made your detector and learned how to measure resistance with a DMM
2) Measured minimum output levels and set bias
3) Adjusted gain and maximized the contrast of your display using the above process
Good Job! But it's still not enough huh?

Okay, let's look at what a Color Correcting filter might do for you.
First, look at the Optimum values that you figured out for your projector without the filter. Is one of the gains much higher than the others? In all likelihood, the gain for your weakest color is much larger than the other two.

If so, you may want to consider a Color Correction filter. The Color Correction filter should match the color of your weakest color. On a 10HT, that color is red, or actually pink. A CC Filter will not make your image any brighter. You have already used up all of your GCP color that you are going to get. But A CC filter will help you get darker blacks WITHOUT making the brightest image any darker.
Okay, now go back to the last data set on the spreadsheet that shows your "optimized" gain settings and look at the recommendation on line 45. It should suggest the use of a CC filter, based on how much correction you need. If you only have a little bit of headroom in your colors, it will suggest a CC10, if you have a lot it may suggest a CC40. The numbers after the CC designation indicate how much light will be removed by the filter.

This is the type of filter you should go and purchase.

Keep in mind that not all filter materials are made equal. Most Optical (camera) quality filters will filter colors equally. For example, an optical quality Optiflex Red CC resin filter will remove equal amounts of Blue and Green. Optical CC filters can run anything from $20-200 US depending on material, style and size, with Glass being the most expensive.. We have found for the 10HT, a 4X4 inch resin filter works very nicely.

Other substances like Color Correcting gels for Arc lights, will still perform the same function, but may reduce blue more than green, etc. They are, however, MUCH cheaper ($6 for a 2ft X 2ft sheet)

Okay, My filter arrived, now what?

Now that you have your filter, you are going to have to figure out a way to mount it. In some cases, especially with glass filters, your mount should allow the filter to be placed at an angle to the plane of the lens. If you don’t do this, it will result in reflections in your image. We have not found this necessary with the plastic CC filters such as the Optiflex 100 filters. In any case, don't mount the filter so it touches the glass elements of the lens. The picture above shows Steve's projector with an Optiflex CC30 R filter in front of the lens, held on with bits of black tape.

For various ways of mounting CC filters, click here.

Once you have the filter in place, you need to adjust the Projector’s color temperature so that you end up with a low color temperature on the other side of the filter. Again, the Smart Spreadsheet will come to our aid and do all of the heavy thinking for us.

First, enter the gain/bias menu and copy your optimum settings into custom 2. Then select custom 2 as the color temperature for the input you are using. CC Filter Data SetsMeasurements with CC filters are basically the same as those you took without the CC filter in place.
One difference is that you should write the letters "CC" in the comment field of the next open measurement set. These letters will indicate to the spreadsheet that the data in this set corresponds to the use of a CC filter. Make sure that you enter these letters in the first comment field for every set of data you collect with the CC filter attached.

Form that point on, you simply collect measurements as you did before:
Verify that your input is using the correct color temperature (custom 2) and your White and Black levels are set correctly
Next, write down the Gain values you will be using for this run in the spreadsheet in the spaces provided. As a starting point, you can either use the gain settings you optimized without the CC filter, or "go for it" and simply set all the gains to 255. Since 255 will probably be closer that the optimized settings, starting with these values will probably eliminate an iteration or two.

Now go to the AVIA IRE window menu. Put up the 0 IRE –100 IRE screens and take measurements using the red, green and blue filters and record the results measured in K½, in the appropriate spaces on the spreadsheet.
Again, Line 45 of the spreadsheet will tell you what to do next.

If the CC filter chosen over-corrects slightly, you might find your self being asked to reduce the contrast setting , this time because Blue or Green have maxed out. This is normal.
Continue following the recommendations, and repeating the process until the SMART says you are done. At this point, your Red, Green and Blue curves should once again be overlaid pretty closely on top of one another.

The set of values in Custom 2 will now represent your optimum settings with your CC filter. Write them down .

Congratulations, you have just completed one of the best performance enhancing tweaks available for the 10HT.
The spreadsheet may suggest additional filtering. Additional filtering may improve the picture or it may lead to over-correction. The decision to overcorrect is a personal one and left up to the user. If you decide to follow the recommendation and try a larger filter, simply repeat the process using the new filter.

Again, as an exercise, you may want to compare the contrast ratios computed for your optimized CC filter vs. the projector without the filter vs. the contrast setting computed for low. You should see a very satisfying improvement in contrast and perceived picture quality.

{Body}