HOW TO: Laser Engrave a Photograph

By Michael Reilly
Let’s say you have this great photograph and want to engrave it on a piece of wood. The image on the left is what we sent to the laser. But what we got was something closer to the image on the right. What went wrong?To find out, open Corel, create a 6″ long by 0.5″ tall rectangle and fill it with a gradient. Set the gradient to go from RGB white to black in 255 steps. Engrave that on your material at the recommended speed/power settings for your material. 375595650_5a6a65950d_orig 375595650_5a6a65950d_engraved
This isn’t a limitation of the laser, it’s a limitation of the material. Most materials have a threshold point (white point) at which they go from no visible mark, to visible mark. And another point (black point) where a visible mark doesn’t get any darker/bolder than it already is. This greatly limits the range of grays that can be reproduced in your material.
On the left is the levels panel in Photoshop CS5. The main feature of this panel is a histogram, a bar graph showing the number of pixels in your image at each of 256 shades of gray. We can see that this image is composed of a wide range of gray values.Because of the threshold behavior of the material, we lose all the pixels in the yellow shaded area. As you can see in our image comparison above, those pixels represent a lot of the detail in this image. Screen Shot 2012-06-21 at 8.06.08 PM Screen Shot 2012-06-21 at 9.04.12 PM
So we’ve discovered that although there are 256 possible levels of gray, in this material, the laser is only able to reproduce the middle 85 shades of gray. Therefore, if we want as much of the original detail in the engraving as possible, we need to force all the pixels into that range. We do this by adjusting Output Levels in the Levels control panel to the minimum and maximum points we found when we engraved the gradient.Although it doesn’t show this normally, I used a different method to modify the image which allows us to see what is really happening when we change those controls. You can see that all the pixels have been pushed to the middle of the range. Screen Shot 2012-06-21 at 8.58.07 PM Screen Shot 2012-06-21 at 9.02.09 PM
375595650_5a6a65950d_adjusted Although it doesn’t look good on the screen, this image will engrave much better than the original. It won’t be perfect however. As we can see in the compressed version of the histogram above, while the curve is similar to the uncompressed version, it’s not the same. This is because we’re asking it to reproduce 256 shades of gray using only 85, so it is forced to pick one of acceptable 85 shades for each of the 256 original ones. This decreases image contrast but retains detail.The software package Photograv is a tool that optimizes images for laser engraving on a variety of materials. While I’m sure it’s more complex than what I’ve shown here, it is effectively doing the same thing. They have analyzed how each material behaves and built a database of adjustments that will make an image engrave in the best possible way. I haven’t used it, but I’ve seen the results and they are impressive. If you engrave images once in a while as we do, it might not be a big deal to experiment and tweak the image files manually. But if you do it all the time, it would be worth having a purpose-built tool to do it for you. Either way, now you know what it’s doing for you!
A few other tips:

  • If the material being engraved is dark in color and the engraving result is light, you will get better results if you apply the adjustments above and then invert the image.
  • If you start with a color image, the limitations of laser engraving are compounded. Instead of 256 shades of gray, a color photo consists of a combination of 256 shades of red, plus 256 shades of green, plus 256 shades of blue as shown in the histogram to the right. In our example here, that would mean compressing as many as 1.67 MILLION colors into the 85 we can reproduce. As we might expect, that will require compromise. Old color scanners and digital cameras used to simply use the green channel of an RGB image. But software has gotten better, so I would recommend starting with the “convert to grayscale” option, and work from there.

The image used in my example is by Linda Cronin and was used under Creative-Commons Attribution license.