Many DLP printer makers experience a brightness fall-off of their projected images towards the edges of the vat floor. This will lead to inhomogeneous curing of the resin and therefore failed prints. In this post, I will explain the reason for this effect and provide a simple solution using Monkeyprint.
What is the reason for this brightness fall-off and how do we counteract? Read on to find out…
The inhomogeneous DLP printer brightness distribution results from two effects: The given brightness distribution of your projector, reflection and absorption by the vat floor and natural vignetting at the vat floor.
Projector brightness distribution
Every projector will have a brightness fall-off towards the edges of the image. This can be experienced with almost any projector and unfortunately there’s nothing we can do about it.
Absorption, reflection and natural vignetting
All of theses three effects depend on the incident angle at which the light from your projector hits the vat floor. In the center of the vat, the incident angle will ideally be α = 0° (this means normal to the vat floor) and towards the edges the angle will become more oblique.
Absorption happens because your vat floor is not 100% transparent and will therefore absorb some of the light. The amount of absorption depends on the distance d that the light has to travel through the glass. This distance will be the shortest if the light hits the vat floor at an incident angle of α = 0°. The more oblique the angle gets, the longer the distance and the amount of absorption will be. Therefore, absorption will increase from the center (α = 0°) towards the edges (α > 0°).
Natural vignetting results from the same phenomenon that makes winters cold and summers warm. A light ray of a given diameter will illuminate a specific area when falling onto a surface. Compared to an incident angle of α = 0° this area will increase if the ray hits the surface in a more oblique angle (α > 0°). If the same amount of light illuminates a larger area, this will result in a lower light intensity. As with absorption, the brightness will decrease from center to the edges of the vat floor.
Reflection of light will happen at all positions of the vat floor. The amount of light that is reflected depends on the incident angle. For glass, it will be the smallest (about 3.5%) at incident angles of α = 0°. As the angle grows towards the edges of the vat, the amount of reflection will gradually increase. At α = 30° it will be about 7% according to this article on reflection.
The following diagram shows the combination of all the effects taking place at your vat floor.
All these effects can lead to a strong fall-off of the vat floor light intensity towards the edges. In the Groningen printer we experienced an intensity decrease of 24% in the corners of the vat.
So how do we handle this to achieve an homogeneous DLP printer brightness distribution and better print results? Find out in the next paragraphs…
The common sense approach
First, we can deduce a simple rule from what we’ve learned before. All the negative effects described above increased with the projector light’s incident angle. So the first step when aiming for a homogeneous light distribution should be to decrease this angle.
The software approach
Additionally, I have implemented a calibration image setting in my Monkeyprint DLP print software to resolve an uneven brightness distribution. It has two steps. The first is taking a calibration image of your illuminated vat floor which will show the brightness distribution. The second is to use this calibration image to adjust your slice images for a homogeneous brightness distribution.
1: Take your calibration image
- a digital camera
- preferably a tripod
- a large white surface
- a piece of paper that snuggly fits your vat floor
- your DLP printer
In this process, you will use a digital camera to take an image of your illuminated vat floor. As a camera is an optical instrument, it will show vignetting itself which needs to be taken into account as well.
First, we will extract the vignetting of your camera. To do so, take a photograph of the white surface in a bright, evenly lit surrounding. Important: Adjust your camera’s settings so that the image is as bright as possible without being over-exposed. The result will not be white completely but will tend to get darker towards the corners. This is the vignetting of your camera’s lens which we need to take into account when creating the vat floor calibration image. To finish off your camera vignetting image, load it into a graphics software like Gimp, convert it to gray scale (Image > Mode > Grayscale) and adjust the brightness so that the brightest spot (usually in the center) has a value of 100%. To do this, sample the brightness at the brightest spot using the Color Picker Tool. Create a new layer and fill it with a gray value of 100 minus your sampled value. Set the layer mode to Addition. Now the brightest spot will have a brightness of 100%.
Then, fire up your DLP printer’s projector and throw a white image onto your vat floor. Place the piece of white paper in your vat. Mount the camera onto your tripod and point it down vertically onto the vat floor. Try to get the vat leveled in the viewfinder. The vat should appear rectangular with as little distortion as possible.
Take an image of the illuminated vat floor. Again, adjust your camera’s settings so that the the image is as bright as possible without being over-exposed.
Next, we need to correct the vat floor image for the amount of vignetting that was introduced by your camera. To do so, load the vat floor image into Gimp. Then load the camera vignetting image as a new layer. Invert the camera vignetting image layer and set the layer mode to Addition. This will brighten the vat floor image by the amount of vignetting that was introduced by your camera.
Now, you need to crop your vat floor image. In Gimp, activate the Rectangle Select Tool. In the tool options, set the Aspect Ratio to Fixed and enter your projectors resolution (1920:1080 for a full HD projector). Now, drag a rectangle over the image of the vat floor, that lies just inside of the bright area. In the menu select Image > Crop to Selection.
In a last step, get rid of the paper texture that might be apparent in your image. Select Filters > Blur > Gaussian Blur and set the Radius to something like 100 px. The image should not look patchy anymore.
Congratulations, your vat floor calibration image is now ready! To use it during prints with Monkeyprint, commence with step 2.
2: Load the calibration image into Monkeyprint
Monkeyprint now has a method to adjust the slice images according to your vat floor calibration image. You can supply a calibration image in the settings dialog. Simply go to Options > Settings, click on the Projector tab and load your calibration image.
During slice preview and printing Monkeyprint will darken the slice images in regions where the calibration image is brightest and retain the brightness where the calibration image is dark. The result will be a slice image that equalizes the brightness distribution created by your projector and the vat floor, resulting in a homogeneous exposure during prints.
If you like this or if you have questions or remarks, please use the comment function below.
8 thoughts on “DLP printer brightness distribution correction in Monkeyprint”
That’s what I want to have for my DLP. Thanks for this great implementation. Taking photo and manipulating it with a software is a good solution. I want to ask if we can measure the UV illumination with an uv sensor and generate the image. What kind of sensor can we use? Measuring the uv wavelength is more meaningful that taking the photo of the visible spectrum. What do you think?
good to hear you like it!
I’ve just done some reading on UV photography. Seems like it takes some modifications to your camera (removing UV filter, having UV passing filter on the lens etc).
I think it might be an alternative to put some fluorescent sheet material onto the vat floor and record the visible fluorescent light only. Then, of course, you would have to be sure your projector emits UV only…
Your suggestion might also be good and is easy to implement.
You could take a UV sensor like this one and measure the intensity in all four corners and in the center of the vat floor.
Then, average the four corner values and divide the result by the center value.
You could then create a calibration image with a radial gradient ranging from a value of 1.0 in the center to the result of the above calculation in the corners.
If you try it, please let us know about the results 🙂
What I have in my mind is a grid print out. I mean a paper which has 10mmx10mm squares on it. You will put this print out in your vat and start the projector with a white image. People will use the sensor on the corners of the squares and press the button to record the uv to analog voltage at that points. The software will generate the filter image. If you can implement it inside the Monkeyprint I can work on the hardware side.
The sensor can be added to my board I believe and the in the software a calibration procedure could be implemented that automatically projects spots of white on the vat floor at different locations successively and samples the UV value on mouse click.
This will take some time, however, as it’s not on the top of my list right now.
In the meantime, you could do some testing with the sensor and see if it works at all…
Thank you for this very helpful tutorial.
I have a LittleRP which uses a mirror to reflect the image. In your diagram above, it looks like the projector is facing straight at a vat. Now, in terms of homogeneous light how far should the projector be away from the mirror? As far back as possible?
Once I’ve established this, I’ll start on the procedure.
thanks for your comment 🙂
Regarding the projector distance: well, there is no definite rule about this I’d say. There’s always a trade-off between homogeneity and practicability. You might get a very homogeneous intensity distribution if you position your projector 2 meters away from your printer — but that takes a ridiculous amount of space of course…
So I’d suggest you try setting your projector to the nearest and farthest zoom distance, position it so that the projection fits your vat each time and take a calibration image with these two settings. This way you can assess the effect of the distance on brightness homogeneity.
Regarding the mirror: I don’t think there’s much of a difference between direct projection or via the mirror.
Once you’ve got your tests running, it would be nice if you posted some pictures here 🙂
Thanks and kind regards,