If you run a small laser shop like I do, you have probably seen this happen: the design file is the same, the layers are untouched, the dimensions are unchanged, but you switch machines or materials and the result looks different. The first reaction is often “the machine is the problem.” Then someone starts adjusting power, and eventually the job gets close enough by trial and error.
The problem is usually not the file, and it is not necessarily the operator’s skill. The problem is that the file only records what shape to make. It does not record the conditions under which the job was made. Laser results come from the file, the machine, the material, the settings and the assist conditions working together. A file name can only carry so much information. The conditions that actually determine repeatability live outside the file name.
Using Caverjoy’s own Z12 fiber series and C1 CO₂ machine as examples, this article is about one question: why a project record has to capture process conditions, and what those conditions actually are.
[Image 1: Shop handoff scene—one design file moving across different machines and materials; use a photo or diagram.]
1. What a file name can tell you, and what it cannot
A project usually goes through several rounds from first draft to delivery. Names like “final,” “final2” and “final-confirmed” can tell you the order of files, but they cannot tell you three critical things:
- Which machine the file is for;
- Which material, thickness and surface condition it applies to;
- Which settings produced the result, and whether that result was actually verified.
A project record exists to fill those gaps. Its purpose is not to “manage files” for its own sake. Its purpose is to tell the next person why the file was created, what scenario it applies to, which conditions have been verified, and which are still only reference information.
Take Caverjoy Z12. Z12 is a series. The reference configuration includes 20 W and 30 W standard fiber (MAX source), 50 W 3D fiber (MAX source), and 60 W MOPA (JPT source). Under the same series name, the source type and adjustable parameters are not the same. If the record only says “Z12 settings,” the next user cannot tell which configuration those settings belong to.
Take Caverjoy C1. C1 is a 45 W CO₂ machine with a 640 × 350 mm work area, a built-in positioning camera, water cooling and an air pump. Reference single-pass cutting capability is up to 15 mm wood and 10 mm acrylic under suitable conditions. “Wood” and “acrylic” are material categories, not process conditions. If the record only says “10 mm acrylic,” a different batch or a different air-assist setting can change the edge from clean to amber.
So the first rule of a project record is simple: record conditions, not conclusions. “It can cut 10 mm acrylic” is a conclusion. “45 W, this speed, this power, air assist on, this focal setting, 10 mm acrylic, acceptable edge” is a condition.
2. Inside the Z12 series: why 20 W and 60 W MOPA settings do not transfer
This is the most underestimated point. It is easy to assume that because both machines are called Z12, the settings can be scaled up or down. In the reference configuration, however, 20 W/30 W are standard fiber, while 60 W is MOPA. The adjustable parameter space itself is different.
On a standard fiber laser, pulse width is essentially fixed. The operator mainly adjusts power, frequency, speed and fill. On a MOPA fiber laser, pulse width is an independent adjustable parameter. Pulse width and frequency can be set separately. That gives MOPA one more dimension: how the energy is distributed in time.
[Image 2: Z12 reference configuration comparison—20 W/30 W standard fiber (MAX), 50 W 3D (MAX, RK-CAD), 60 W MOPA (JPT, EZCAD / LightBurn); highlight whether pulse width is independently adjustable.]
Why does that matter? Consider metal surface work. Some surface effects, such as color changes on stainless steel, come from light interference through a thin transparent oxide film. Different film thickness produces different colors. That film is not created by “setting a color value.” It is the result of a heat history. Pulse width directly affects how long each pulse deposits energy, which in turn affects how the oxide film grows. A target color is therefore the result of a set of heat-history conditions, not a single field you can fill in.
[Image 3: MOPA pulse width and heat history—short vs. long pulse energy distribution in time; conceptual diagram only, no specific process values.]
If you take a tuned 60 W MOPA combination of pulse width, frequency, power and speed and move it directly to a 20 W standard fiber machine, the issue is not just “less power.” The pulse-width dimension does not have a comparable adjustment on the standard fiber machine. The reverse is also true: scaling 20 W settings up to a 60 W MOPA machine is not reliable. The two power platforms have different energy budgets, so the same settings can produce different energy density at the material surface—possibly crossing from light surface coloring into deep removal or overburn.
What goes into the record: the equipment condition should include source type + model + power platform, not just “Z12” or “fiber machine.” For MOPA configurations, pulse width should be recorded separately, because it is the key variable that separates MOPA from standard fiber.
3. Field lens: “working area depends on the field lens” is not a footnote
In the Z12 reference configuration, working area is listed as depending on the confirmed field lens. That is technically important: change the lens, and both the working area and the spot size change.
With an F-Theta field lens, a longer focal length generally produces a larger spot. Power density is inversely related to spot area, so as the spot gets larger, the same laser power is spread over a larger area and energy density drops. The result: a set of power, frequency and speed settings that can mark deeply with one lens may not mark deeply with a longer-focal-length lens. That is not a machine fault. The optical condition changed.
[Image 4: Field lens focal length vs. spot size and power density—longer focal length generally means larger spot and lower power density; conceptual diagram only.]
So “Z12 marking settings” is not enough for a record. Field lens focal length, working area and focus position should be part of the equipment condition. Otherwise, after a lens change, the settings “stop working” and the operator can only start over without knowing which condition changed.
4. The C1 CO₂ side: material conditions and assist conditions must be recorded separately
CO₂ lasers and fiber lasers interact with materials differently, so C1 has its own repeatability issues. The reference configuration lists C1 as 45 W, 640 × 350 mm, water-cooled, with an air pump and a built-in positioning camera. Software is CutLabX; LightBurn compatibility depends on the controller and software version.
Wood. Wood cutting and engraving are sensitive to heat input. Assist gas type and pressure directly affect edge charring. An inert environment helps suppress overburn; with insufficient air assist, heat accumulates near the cut and the edge darkens. Even within “wood,” species, density, thickness and moisture content change the best conditions. If the record only says “wood” without thickness, surface condition and air-assist settings, a different batch drifting in result is entirely expected.
Acrylic. Amber edges on acrylic cutting are essentially a sign that heat accumulation has exceeded the energy density needed for clean vaporization, and the excess heat degrades the material. The reference capability says “up to 10 mm acrylic under suitable conditions.” Those suitable conditions include at least power, speed, air-assist pressure and focal setting. If the record only says “10 mm acrylic,” a different C1 or a different batch can turn a clean edge into an amber one. That is not random.
Camera positioning. The built-in camera on C1 helps align the design to the actual material position. It solves a positioning problem; it does not change laser settings. The record can include camera calibration status and alignment method, but keep the distinction clear: accurate positioning does not mean the process is repeatable. One does not replace the other.
[Image 5: C1 CO₂ process record—material, thickness, air assist, focal setting, speed and power should be stored as one condition set.]
What goes into the record: the material field should be split into material name, thickness/spec, surface condition, assist gas type and pressure. C1 records should also include focal/focus setting. A broad material category alone is not a repeatable condition.
5. Source files, delivery files and software parameter portability
A source file is for continued editing. A delivery file is for viewing or exchange. That distinction is correct, but in laser work there is one more point: a delivery file usually does not carry settings. An exported outline file can tell someone the shape and dimensions. It cannot tell them whether that outline is a cut or an engrave, what the hatch spacing is, or which parameter library was used. Settings must be recorded separately in process fields. Do not expect someone to “see” them in a delivery file.
The same applies to preview images. A preview helps identify content quickly, but it loses dimensional accuracy, layer ownership, and the distinction between fill and outline. It is fine to note in the preview file name whether it is shown at actual scale and which version it corresponds to, but do not use the preview as a substitute for structure and dimensions in the source file.
On the software side, Caverjoy’s reference configurations include several environments: Z12 20 W/30 W and 60 W MOPA reference EZCAD / LightBurn; 50 W 3D references RK-CAD; C1 uses CutLabX, with LightBurn compatibility depending on controller and software version. Parameter library formats are not universal across software. Directly importing project files or parameter libraries between environments is often limited. Even with conversion tools, fields such as pulse width and frequency may not map completely.
[Image 6: Source file, delivery file, preview and parameter library relationships—delivery files do not carry settings; parameter libraries should be archived with the record.]
So the record should include source file format, software and version, whether the parameter library is saved with the record, and whether any format conversion was performed. Otherwise, a MOPA setting tuned in EZCAD can be “re-entered” in another software environment with key fields missing, and the reason for the failed reproduction will be hard to find.
6. What to record: define fields by asking “what happens if we omit this?”
Instead of an abstract rulebook, work backward from consequences. These are the groups most worth keeping for Caverjoy Z12 and C1 projects.
Equipment condition. Model (Z12 / C1), source type (standard fiber / MOPA / CO₂), power platform, field lens focal length (Z12) or work area (C1), focus position.
If omitted: after a power-platform, lens or focus change, settings cannot be transferred and the source of the difference cannot be identified.
Laser settings. Pulse width (required for MOPA; mark “fixed” for standard fiber), frequency, power, speed, fill/hatch spacing, number of passes.
If omitted: recording only a power percentage cannot reproduce the heat history or show whether energy density has crossed a threshold.
Material and assist conditions. Material name, thickness/spec, surface condition, assist gas type and pressure (especially for C1).
If omitted: same-name materials with different batches, density or air assist can drift in cut and engrave quality.
File and software state. Source file format, software and version, whether the parameter library is archived, whether format conversion was used.
If omitted: key fields can be lost in conversion, or the parameter library is missing, and the record becomes unreliable after migration.
Verification status. Mark clearly: verified on this machine and material / reference only / needs retest.
If omitted: settings from another machine or power platform may be treated as verified and put into production.
[Image 7: Project record field framework—equipment condition, laser settings, material and assist conditions, file and software state, verification status.]
That last item matters most. A complete layout and a nice preview do not mean the process is verified. A file that has not been tested on this machine and this material should not be called verified just because it looks finished. Files reused across machines should be marked “reference only” or “needs retest” so another machine’s settings are not carried over directly.
7. Build a maintenance habit that lasts
A record is valuable not because it is written once, but because it is maintained. Set one update entry point and one archiving rule for each project so the same note is not edited in multiple places. Before publishing externally, check again for personal or internal information. When a project is complete, archive the record and note where it is stored so the team can find it later and hand it off cleanly.
None of this is complicated. The hard part is treating it as part of the process rather than an administrative burden. When a record captures conditions instead of conclusions, it becomes technically useful: the next time the same file does not produce the same result, you can troubleshoot from the record instead of starting from scratch.