Common Misconceptions with Scratch Dig, Flatness and Homogeneity

In this article, I’m sharing a primer on some of the most common questions I get in optics manufacturing. Throughout, you'll see links to past articles that dive deeper into specific topics. 

From the perspective of an optical manufacturer, there's often a disconnect between what a customer specifies and what they actually require. A common default is, “Just give me the best you offer.” In reality, you may end up paying for a component that far exceeds the requirements of your application. 

Scratch/Dig
Let’s start with an easy one: scratch/dig specifications. Whether specified under ISO 10110 or MIL standards this requirement can quickly increase processing complexity and cost. 

Say you need a 1" UV fused silica window, 0.25" thick, and you specify 20/10 scratch/dig. For this example, I'll reference MIL specs, which define physical defect size, whereas ISO uses a formulaic average. Achieving a 20/10 spec means extra polishing to remove fine scratches and pits, but for a small 1" part, it’s usually attainable without significant added cost. Now, imagine the same customer returns and says, “Hey Mike, I have another project—same specs and material—but I need 25 pieces of an 8” diameter window.” Seems simple, right? However, at 8", the surface area is vastly larger, and even a single scratch outside the 20/10 spec can fail the part. To prevent this issue, we invest significantly more time in grinding and polishing to ensure the surface meets the higher standard. 

The quoted price difference between the 1"part and 8" part is often drastic. For many new to optics, this seems confusing; they assume similar machining steps and marginal material cost increase. That’s why I always prefer a quick conversation before we dive into quoting. Understanding the final application often reveals that a less stringent spec is acceptable. Sometimes, the customer just assumes “higher specs = better,” when they could actually save money without sacrificing performance. 

Flatness
That said, for laser-grade finishes, tight specs are necessary, and in those cases, the price usually makes sense.  

Optical flatness and transmission are related, but distinct, concepts. 

While flatness affects the quality of transmitted light, it is not the same as the amount of light that gets through the optic. 

Flatness is especially critical for mirrors or other reflective surfaces. The manufacturing spec here is called Reflected Wavefront Error. This defines how well light reflects off a surface and how much it is distorted in the process, based on how precisely “flat” the surface is. 

For transmissive optics, the relevant spec is Transmitted Wavefront Error (TWE), also known as Transmitted Wavefront Distortion (TWD). This measures how well an image or laser beam passes through a window without distortion. 

These distinctions are key because they directly impact manufacturing methods and costs. At specs of 1–2 waves flatness, the difference isn't huge, but when you're talking about ¼ wave or tighter (down to 1/20 wave), the manufacturing becomes significantly more challenging and expensive. 

Here is a quick way to think about it: a perfectly flat window transmits light cleanly. But a wavy surface distorts the light, degrading image quality and reducing effective transmission. So yes, flatness affects transmission, but it is not the same thing. Flatness is a physical surface property, while transmission measures light interaction with material. 

Homogeneity
Now let’s shift to material specs rather than manufacturing specs. One important concept is homogeneity: the uniformity of the refractive index throughout a piece of optical material. In high-performance optics, even small variations can lead to optical aberrations or reduced system performance. 

Glass manufacturers like Schott, Heraeus, OHARA, and Corning offer different homogeneity grades. Let’s take Corning Fused Silica, a popular material, as an example. A typical spec is “0A”: 

Here’s the key point: customers often include "0A" for small optics, but for parts under 10 mm, homogeneity usually isn't a critical issue. You won’t see a meaningful difference between a 0A and a 0G in this size range. However, as you scale to 10 mm and above, especially in semiconductors or high-precision applications, then yes, homogeneity becomes important and should be specified. 

In conclusion, the team at Esco and I are always available to answer questions, solve problems, and talk shop. Our capabilities cover the gamut from spherical parts to plano-optics to custom CNC-machined configurations. Let’s work together to make sure you are specifying exactly what you need—nothing more, nothing less. Please reach out to me at Mike@EscoOptics.com 

Read more of my blog series:
Optical Design for Manufacturability
Optical Coatings and Cost
7 Tips for Pressure Window Design
Prototypes Optics vs. Production Optics
Commercial Grade vs. Laser Grade Optics