Calibrating the Projector

Initially, I was trying another calibration print with the teardrop cylinders, but this time I tried using an object support base to avoid my cylinders from hitting the resin relief holes in the lift. After seeing the first layer float, I decided to keep trying to make a good first layer that stuck to the lift. I kept my attempts in a spreadsheet and took some pictures.

A couple of the trials did stick, so I allowed the next layers to be printed on top. These next layers further exposed the bottom layer, causing it to shrink and peel off the lift. It seems the translucency of each layer allows the bottom layers to get more exposed than the top. I believe this is the reason for the top 1 mm of layers in my second die print being a little bigger than the others. I haven't cured that die yet, and it may shrink fully after putting it in the sun for a day.

The biggest lesson from these trials was just how uneven my projector illuminates each layer. Half way through my experiment trials, I started modifying the first layer with gradients to darken parts that would cure first. This caused curing much more evenly. I decided to try to figure out how to even the illumination.

I tried the lazy way first - took a picture of a white square on the lift. I hoped my camera would pick up the subtle illumination differences, but even a white sheet of paper wouldn't show unevenness.

Using a photoresistor to measure the projector luminance

After some internet research, I found that a photoresistor being measured with a digital multimeter can help to measure the incident light for cheap. I took the projector off the printer and projected a white screen at my wall, ensuring the image was rectangular. Using the photoresistor, I got a feel for the resistance highs and lows. To even out the projector, I needed the image to measure at the resistance high independent the location of the photoresistor in the white square. Then I took a picture of the screen using my camera, being careful to have the image only use the middle 1/9 of the screen (1/3 vertically, and 1/3 horizontally) to prevent the vignetting of the camera from affecting the measured luminance. Using GIMP, I fixed my perspective in the camera image (was slightly off center), heavily blurred to remove the wall texture, inverted the colors and then used the "levels" tool so that the brightest spot, corresponding to the darkest projected spot, was maxed out. The darkest spot in the lower left I decided to be a no-print zone to keep exposure times lower, and worked with the second darkest spot. After a guess at how dark the brightest projected spots needed to be, I put the image on the projector and measured with the photoresistor. After a few iterations to get the levels tool to the right settings, I had a great starting point.

My camera shot of the projector's white screen. Notice the dark bottom left.

With the leveling image being projected, I started to manually "GIMP airbrush" the brighter spots to bring them within 10% of the resistance of my darkest spot. My projector has the most light bottom center and on the right side. This unfortunately means my first 2 calibration prints did not absolutely find the best exposure time, though 6.0 seconds seems to have worked ok on the second die. With the whole image now darker, I'll bump my exposure to 10 seconds.