You need a 1000 nm shortpass filter. You find the Edmund Optics product page for the #68-576. The specs look right. You click "Add to Cart." Done. Simple.
That's what I thought, too. In September 2022, I ordered 25 units of the #68-576 1000 nm shortpass filter for a new laser cutter integration project. The result? $1,200 worth of optical glass that was, for our application, useless. It looked perfect on the spec sheet. It was the wrong part for the job.
The Surface Problem: A Filter That Didn't Filter
The project was straightforward: integrate a safety filter into a custom laser cutting setup for thin metal and jewelry prototyping. The laser source emitted around 1064 nm. We needed a filter that would block everything above 1000 nm to protect a sensor. The Edmund Optics #68-576 filter is literally named "1000 nm Shortpass Filter." The product description said it transmits shorter wavelengths and reflects longer ones. Match made in heaven. Or so I believed.
The filters arrived. They were beautiful—precision optics, clean edges, Edmund Optics quality. We installed one. The sensor readings were off. Way off. After a frustrating day of troubleshooting, we realized the filter wasn't blocking the 1064 nm line with the efficiency we needed. The laser energy we were trying to stop was bleeding through. The core function failed.
My first reaction was confusion, then frustration at the vendor. Did they send the wrong part? I checked the box, the label, the glass etching: #68-576. Exactly what I ordered. So the problem must be with the product itself, right?
The Deep, Hidden Reason: "Shortpass" Is a Story, Not a Switch
This is where I learned the hard way that ordering precision optics isn't like ordering office supplies. The surface problem was a failed filter. The real problem was my fundamental misunderstanding of what a spec sheet actually tells you—and, more critically, what it doesn't.
I called Edmund Optics tech support (who, to their credit, were patient and helpful). The engineer asked me one question I hadn't considered: "What was your required optical density (OD) at 1064 nm?"
I went silent. I didn't know. I hadn't specified it. I assumed a "1000 nm shortpass filter" meant it blocked everything above 1000 nm. Full stop. But that's not how it works. Blocking isn't binary; it's a gradient. The spec sheet for the #68-576 showed its cut-on wavelength and transmission curves. It started blocking above 1000 nm, but to reach high attenuation (like the OD 4+ we needed), the wavelength had to be significantly higher. At 1064 nm, just 64 nm past the cutoff, its blocking power was marginal for our high-power application.
My mistake wasn't picking the wrong product number. It was thinking in binary terms—pass or block—when optics deal in curves, slopes, and decibels of attenuation. I bought a filter based on its name, not its performance graph. The spec sheet had all the data; I just didn't know how to read it. The vendor sold me exactly what they advertised. I was the one who misapplied it.
The Cost of a Shallow Understanding
The immediate cost was clear: $1,200 for 25 paperweights. But the real costs piled up quickly.
Project Delay: The laser cutter integration stalled for a week. We had to source a correct filter (a different, more specialized Edmund Optics part, the 47-822, as it happened) on a rush order. That added a 65% premium.
Credibility Erosion: I had to explain to the engineering team why their timeline was blown. "The filters didn't work" sounds like a vendor problem. "I didn't understand the specifications" is a me problem. That shift in blame is a career currency I spent.
The Ripple of Rework: This was a custom mount design. The replacement filter had slightly different thickness and diameter specs. The mechanical team had to modify the CAD files and re-machine three prototype mounts. More time, more money.
Honestly, the financial loss stung, but the professional embarrassment lasted longer. I was handling optical orders for a laser equipment company. I should have known better. I felt like a beginner again.
The Checklist That Came From the Crash
That $1,200 mistake now lives at the top of our team's "Optical Component Procurement" checklist. It's not complicated. Its only purpose is to force a conversation with the spec sheet that I failed to have.
1. Never Trust the Name. Interrogate the Graph.
What is the exact wavelength of interest? Find it on the transmission/blocking curve. What is the specified attenuation (in OD or dB) at that exact point? If the curve isn't in the datasheet, request it.
2. Define "Block" With a Number.
“Block” is meaningless. “OD 4 at 1064 nm” is a specification. Before searching, know your required Optical Density or attenuation factor for the wavelength you need to stop. This is the single most important question for any filter, shortpass or longpass.
3. Context Is King: Angle and Power.
Are you using this at a 0-degree angle of incidence (AOI) or 45 degrees? Transmission curves shift with angle. What is the power density (W/cm²) of your source? Some filters can handle low-power lab beams but will degrade or damage under industrial laser cutter intensities.
4. The Edmund Optics (or Any Vendor) Lifeline.
Before finalizing any cart with a component like the #68-576 or 47-822, use the vendor's application engineering support. Send them your exact requirements: "We have a 50W source at 1064 nm, need OD >4 at that line, 0-degree AOI, beam diameter 5mm." Their job is to say, "The #68-576 is wrong for that; you need this other part." Let them be the expert. I pay for that service in the product price; I might as well use it.
Simple. Three questions and a phone call. It would have saved $1,200, a week, and a chunk of my pride.
I still have one of those #68-576 filters on my desk. It's a reminder that the most expensive component isn't the one with the highest price tag—it's the one you buy without understanding why you need it. Now, when someone on my team asks about ordering a "laser cutter filter" or wonders "what can I do with a laser cutter," I tell them this story. The machine's capabilities are vast (cutting, welding, jewelry, marking), but its success starts long before the beam turns on. It starts with a humble, thorough read of a spec sheet.
