Introduction to Light Interference Color

If you are interested in pursuing projects in laser color marking, I encourage you to start with at least a few of the existing YouTube videos about color via light interference. There are many ways to treat steel to obtain colors, but they all take advantage of the same underlying principle: a transparent chromium oxide layer on the surface of the steel creates an interference effect when light reflects from the top and bottom surface of the oxide layer. In our case, we are using the laser to heat or melt the surface of the steel plate in order to change the thickness of that chromium oxide layer. The heated steel reacts with the oxygen in the air and forms an essentially transparent metal oxide (like glass). The color we see is due to light bouncing off both the top oxide surface and bottom metal/oxide interface. These two reflections add constructively or destructively to result in one wavelength of light being enhanced in reflection. As the oxide thickness increases, the enhanced wavelength also increases, resulting in colors that start out blue/violet, and shift towards red. This is the same thing that happens with the colors you may have seen next to steel welds or tempered steel. It is also possible to accomplish the same oxide growth using an electrolyte bath and a voltage / current source. Many of the existing videos are presented from the perspective of Fiber laser owners, which have some additional control “knobs” to what we have on diode lasers, but the general concepts are broadly applicable to any laser system with the capability of heating / melting the surface of stainless steel. That’s how I started these investigations, and it was a great way to get a quick intro to the field. A good example is linked below:

Light interference Colors on Stainless Steel MOPA Fiber Laser Color

Watching the videos above will give you a good idea of the science underlying laser color marking (thin film optical interference) and reasonable expectations about what kinds of colors can be obtained. That’s pretty much the starting point I had when I began my own experiments.

Materials and Useful Accessories

A quick note about stainless steel. There are a vast number of different kinds of steel with differing chemical compositions, and the opinions about what the best steel to use for laser color marking seem to pretty much zero in on 304 Stainless (with 201 Stainless being the 2(nd) choice). I have yet to find an explanation of why 304 Stainless seems to work best, so I can’t comment on the truth of this opinion. I’ll come back and edit this later if I find out. Since I’m not trying to get another PhD, I decided to start with 304 stainless and not worry about it for the time being.

My best results to date have come from using “frosted” steel plates of at least 1mm thickness, sourced from Amazon. The heat generated by the laser for some of the colors is easily enough to warp the steel due to local heating, so plates thinner than 1mm were harder for me to work with. Ultimately, I purchased an aluminum hold-down plate with many tapped holes and then used Flanged Button Head screws to hold down the edges of the steel plates during laser marking. Otherwise, the plates could bend so much that they could defocus the laser. I used wide washers and more screws to hold down the middles and corners of the steel plate on the “free” sides shown in this picture. The washers provide hold-down points for pieces whose sizes aren’t an exact multiple of the tapped hole spacing.

Setting the Laser Parameters

My next step was to take a screenshot out of one of the videos that Snapmaker released to show the capabilities of the new RAY system where they BRIEFLY showed a test array for laser color marking one of their engineers had done. That gave me a place to start for laser power, line interval spacing, and laser work speed. From there I began a multi-week dive into the various parameters used to produce different laser colors. Indispensable to that effort was the LightBurn program (https://lightburnsoftware.com/), where I was able to make excellent use of their 30 day free trial period. LightBurn has a built-in option to create test arrays where you can vary the different laser parameters in a consistent way. One example is shown below:

In an array like the one above you can see how different settings, such as Line Interval Spacing and Travel Speed, impact the resulting color. You can also leave one of these parameters fixed, and instead vary Laser Power. As a word to the wise, when you start getting close to colors that you like, make sure to try arrays of larger area than the small squares above. Even if a small square looks uniform, a large square at those same setting will often show visual defects that weren’t obvious in a smaller area. I learned (and re-learned!) that lesson many times!

Filling an Area with Color (Rastered Lines vs. Dot Filling)

One thing to keep in mind here is that, while the color is primarily dependent on the thickness of the metal oxide you create, there are multiple ways to influence that thickness. Some changes have large effects, while others can be more subtle. Depending on what effect you may be trying to achieve, different ways the laser power is delivered to the steel surface can result in two areas that get the same “integrated” power looking different from each other. Maybe the most obvious example of this comes from (I assume) the “roughness” of the buried metal surface. Here I’m referring to the interface between the transparent metal oxide and the steel. A very smooth metal surface acts much like a mirror (what is referred to as “specular” reflection). Light bouncing off a mirror-like surface reflects off at the same angle as the incoming light. A rough surface, on the other hand, produces what is called a “diffuse” reflection, bouncing light off much more uniformly, over a wider range of angles. In the case of laser color marking, these effects have a strong impact on what colors we see, especially as we look from different angles. If a surface is highly reflective, we may see very different colors in any given area as we change the angle from which we are looking. If a surface is more diffuse, the color will be much more uniform over multiple angles. The “frosted” steel plate surface is an excellent example of this effect. I found that difference in angular color appearance to be particularly pronounced when comparing color markings done using the standard “rastered lines” for filling areas as opposed to Luban’s option for “dot filling”. I suspect this is primarily from the difference in the roughness of the buried metal surface between these two methods of filling space. Whatever the cause, I found the colors obtained from “dot filling” to be much more consistent when viewed from different angles. As always, I encourage you to try out both “dot filling” and “rastered line filling” to see which style appeals to you more. The dot filling approach is generally slower (you are stopping and starting the laser motion MANY more times than with rastering), but it does look better (in my opinion). One significant downside to dot-filling is that it is not available as an option to fill an area in LightBurn. Below I show some images of two butterflies with similar colors (at least in some viewing directions) made using line-rastering (on the right) and dot filling (on the left). As you can see, the colors of the dot-filled areas are much more consistent when viewed from different angles, while the line-filled areas look dramatically different from some angles where the “specular” effect is quite strong. For the butterfly images, I found the dot-filled color areas to look much better from multiple angles, although a more specular area can really “pop” at the right angle.

Two additional settings/parameters I will discuss are Overscanning and Constant Power Mode / Adaptive Power Mode. They are somewhat interrelated, so I will talk about them together. Overscanning is a parameter in LightBurn that is typically represented as a percentage (%) of your travel speed. This percentage refers only to the speed of laser travel and will extend the length of the laser travel beyond the boundaries of your fill area. For any given Overscanning percentage, a higher speed will result in a larger amount of extra travel. The purpose here is to give the laser time to start moving and get up to the target speed before getting to the area where the laser is to turn on, and then to wait to start slowing down until after the laser has been turned off. A larger travel speed setting will need more space/time for the laser to get up to speed, so making the Overscanning distance a percentage dependent on speed makes sense. Since the laser power is a critical determinant of the oxide thickness, any variations as you move across the fill area will result in color shifts, especially near the edges of the area. If you leave “Constant Power Mode” off, the Snapmaker firmware will attempt to compensate for variations in travel speed by adjusting the laser power. Thus, if the laser is traveling more slowly than the target speed, the firmware will reduce the laser power in an attempt to provide the same integrated power to a given area. Unfortunately, the accuracy of this compensation method isn’t always where it needs to be for laser color marking. I typically found my best color results with “Constant Power Mode” switched ON in LightBurn and the Overscanning parameter set between 10% and 50%. Of the two parameters, Overscanning is much more important than the power mode. If you set the Overscanning parameter high enough, you shouldn’t need to invoke Constant Power Mode at all. I will note here that Luban also includes an Overscanning parameter, but in my limited attempts I could not make it work. If you are sticking entirely with Luban, using Constant Power Mode will be critical, as it is the only way you will have to compensate for speed variations.

LightBurn’s Cut Settings Editor Window

Below, I show an image of the LightBurn Cut Settings Editor popup window, where I set the line rastering parameters for the color I have chosen to call “Blue”. We can see that I have chosen a Line Speed of 2500mm/min, a Power of 13.5%, and a Line Interval spacing of 0.05mm. The Constant Power Mode is turned OFF, and the Overscanning is ON and set to 10%. This overscan results in an “extra” 4.17mm being added to each end of the line to allow the laser to get to the target speed before it turns on, and then to slow down after it turns off. The higher the percentage you set, the more extra travel and thus the better chance of hitting the right speed before the laser turns on. On the other hand, this also increases the length of time needed for each line, and thus your entire color marking project. Since the extra distance is only a function of your speed and not the size of the object you’re filling, the time “multiplier” will be worse for smaller areas than for larger ones. In other words, if the object you are filling is only 4mm wide, this 10% Overscanning (at a speed of 2500mm/min) will roughly triple the time needed to fill it, since it adds ~4mm onto both ends of each line. If your fill object is instead 40mm wide, an extra 4mm travel on each side increases the time needed by only 20%. Just find a value that works for you.

For the sake of convenience, I have tried to find a set of parameters that would give me colors similar to the presets in LightBurn (on the left of the above image), so that any images I created there could look at least something like what I could expect to produce from my laser setup. In the end, I got close with some, less close with others, and way off with a few. Good red and green colors still elude my attempts, for example, while you can have pretty much any kind of blue you want. What follows are my best efforts thus far, for both line-filled and dot-filled colors. They seem to give somewhat different colors each time I switch to a new batch of steel plates, so there is always some fine tuning to be done if I’m trying to hit a specific color. Anything that affects the amount of laser power absorbed at the surface of the metal will change the amount of oxide you get, so differences in surface chemistry and surface roughness will both come into play here. Another factor I haven’t mentioned here is the thermal conductivity of the metal. How much oxide you get will also depend on how quickly heat dissipates from the laser spot. Thus, metals with high thermal conductivity (like copper and aluminum) don’t work for laser color marking. Titanium, on the other hand, has a very low thermal conductivity and absorbs light well at the wavelength of our blue diode lasers, and thus works well for color marking. If I can find some stainless steel with a significantly different thermal conductivity than 301, I’ll test it out to see how this affects the color marking.

Example Laser Parameter Sets

Below I show examples of my current “best” set of parameters for both rastered line color and dot filled color. I attempted to create the same set of colors using both techniques, so the missing squares in the dot-filled color example mean that I wasn’t able to find a good set of parameters for those colors.

Rastered Line Color

Dot Filled Color