After a few months of cutting daily with the P3, we now have a parameter notebook that works. This is the public version of that notebook, and the idea is to keep it up to date: every time we master a new material, it goes in here.
A warning before copying anything: these values are a starting point, not a universal truth. Change the wood supplier and everything changes. Always make a test grid on a scrap piece before putting in a full sheet.
First of all: xTool presets do work
And we say this because it's fair, not to make ourselves look good: the settings included in xTool Studio are well designed and provide a good starting point. They're not just for show. What you need to understand is that they require prior validation for two honest reasons:
- In some cases, they're calibrated for xTool-branded materials, and your plywood from the supplier around the corner isn't exactly the same.
- For others, there are factors that xTool can't control from a preset: ambient temperature, humidity, and the material's source. Two sheets that are "identical" from different suppliers behave differently, and that's not a preset failure—it's physics.
Even so, having a starting value that's close to the right one is a luxury. To understand the difference: in LightBurn, when faced with a new material, you start with a blank map and have to explore from scratch. With Studio, you start from a sensible reference and only need to fine-tune it. By far, we prefer fine-tuning to making things up from scratch.
What you need to know before the table
Machine limits
- Engraving: up to 1200 mm/s.
- Cut: limit of 300 mm/s.
- Score (surface marking): limit of 400 mm/s.
That cutting limit is a software limitation, not a hardware limitation. It's worth knowing because it explains a trick we use quite often: with thin materials, switching the operation from cut to score gives you an extra 100 mm/s and a considerably less charred edge.
The three lenses, and which one to actually use
We have all three, and the actual distribution of use is very uneven:
| Lens | What it's for | How much we use it |
|---|---|---|
| S (1.5") | Fine engraving, small details | Occasional |
| M (2.0") | Everything else | The vast majority of the time |
| L (2.5") | Thick-material cutting | Barely used |
If you're deciding whether to buy the complete lens set [xTool EU], here's our honest experience: the standard M lens will work for almost everything. The S makes sense if you regularly do very fine detail work. The L is only worthwhile if you often cut thick material, and we've used it so little that we can't give you our own settings for it.
Affiliate link: if you buy through it, we may earn a small commission at no cost to you. The review is independent.
Air assist is global, not per layer
A software detail worth keeping in mind because it affects the parameters: air assist is configured for the entire job, not material by material or layer by layer. It is in the device settings.
This matters especially with acrylic, where excess air makes the result worse: if you’re going to cut acrylic, you need to go into the machine settings and disable it, then remember to turn it back on afterward. No profile will do it for you.
The table
Plywood — our main material
| Material | Lens | Power | Speed | Air assist | Notes |
|---|---|---|---|---|---|
| Plywood (production) | M | 90 % | 30–50 mm/s | Low | Our daily setting. Cuts perfectly in one pass |
| 3 mm plywood | M | 40 % | 10 mm/s | — | Reduces scorching, but doesn’t eliminate it. Very slow |
| 6 mm laser-safe plywood | M | — | 40 mm/s | — | Community data |
| 6 mm DIY plywood | M | — | 12 mm/s | — | The glue is what slows you down. More than 3× slower |
| 3 mm solid wood | M | 80 % | 25 mm/s | — | Watch out for the machine’s left-to-right asymmetry |
The most profitable conclusion from this entire table is in the comparison of the two middle rows: the same thickness takes three times longer if the glue is not laser-safe. It is explained in the article on scorching.
These settings produce the engraved wooden plaques and many of the wedding favors we sell, which account for most of our machine hours.
Acrylic
| Material | Lens | Power | Speed | Air assist | Notes |
|---|---|---|---|---|---|
| Cast acrylic 25 mm | L | 100 % | 1 mm/s | Disabled | 1 pass, autofocus lowered by 3 mm. “It really is the limit” for 80 W |
This is community data, not ours, and we include it because it marks the machine’s upper limit: 25 mm of acrylic is the most you can expect from 80 W, at 1 mm/s. If your work depends on cutting thick acrylic, this is not the machine for you.
Two details are consistent with acrylic: air assist is disabled —air makes the finish worse— and it is the material that produces the most odor when you open the lid. If you work with an air purifier, leave it running for a couple more minutes before opening.
Thin materials
| Material | Operation | Speed | Notes |
|---|---|---|---|
| Thick cardstock | Score (not cut) | 400 mm/s | Edges noticeably less burned than when cutting at 300 |
What we haven’t achieved
Soft gray tones on slate. It is not a matter of finding the right parameter: the CO₂ tube has a minimum firing power below which it does not emit consistently. This makes the fine gradient that a diode can achieve impossible. There is a "minimum output power" setting you can adjust, but do not expect miracles. It is a limitation of the type of source, not a defect in this machine.
Glass, leather, and doormats
Besides plywood, these are the materials that regularly go through our machine:
- Glass. A lot, especially during the summer season. The RA3 rotary attachment [xTool EU] changes everything here: it has made working with bottles and stainless-steel tumblers enormously easier for us, and it is one of the tools we use most.
- Leather. Only for large batches. For individual pieces, another machine is more worthwhile for us.
- Doormats. Occasionally. It is a rewarding material to work with, with the drawback that it generates a great deal of fibrous residue: cleanup is required afterward.
Working with cylindrical objects using the RA3: what you need to know
Since cylindrical glass and steel are an important part of what we do, and the RA3 has its quirks, it gets its own section here.
Our assessment is very positive—it is one of the tools we use most—but there are legitimate complaints on the owners' forum that are worth knowing before buying it, because almost all of them have a workaround:
- It does not anchor to the bed. It rests in a cradle, without a lock. At high speed, it can shift and ruin the engraving halfway through the piece. If you work quickly, check that it has not moved between pieces.
- Automatic diameter measurement fails in two specific cases: when the T-shaped clamps are aligned with the top, and with bottles that have a cap or thread—it measures one inch and stops. Solutions: rotate the holder, put the cap on, or switch to manual mode.
- The chuck mode is slow: about 15 minutes per glass, whereas rollers are considerably faster. The official explanation is that, in automatic cylindrical mode, three axes move at once. If the surfaces are not parallel, the system asks you to raise them, and doing so greatly increases the speed. It is worth taking that minute to adjust them.
- Workshop lighting matters. One documented case from another owner: with the fluorescent light on, the system did not request the 3° tilt, and the job took 55 minutes; after turning off the light, it dropped to less than 25. It is the same interference issue that affects autofocus.
- For full-wrap designs, the automatic diameter measurement falls short by 30 to 40 mm. If the design goes all the way around the piece, measure it yourself.
And the community trick that saves the most material: wrap the piece in masking tape and run the test pass over the tape. You can see the result, make adjustments, and avoid using one glass for every iteration.
Regarding glass, there are two things we have learned that do not appear in any table: the finish depends much more on the consistency of the focus across the entire pass than on power, which is why the RA3 parallelism adjustments we mentioned above are not just a quirk—they show up in the piece. And with dark glass, it is advisable to test the design in a low area before running the whole thing: the actual contrast varies considerably depending on the shade of the glass.
The workbed: honeycomb or bars
For the past few weeks, we have also had a honeycomb panel [xTool EU], and since it directly affects the result, it belongs here.
The good part: it is very convenient for keeping small pieces from falling to the bottom of the machine. When you cut forty tiny pieces from a sheet, this alone justifies the accessory.
What you need to know: it does not completely eliminate flashbacks—those rebound marks on the underside of the piece—so don’t buy it expecting that problem to disappear. And the clips for securing the material are not particularly convenient.
In practice, we alternate between the honeycomb and the metal bars depending on the job: honeycomb for small, loose pieces; bars for large sheets where the piece is not going to fall anywhere.
How long it really takes
To put the settings in a production context, here are two references that help give you a sense of scale:
- A sheet measuring approximately 30 × 60 cm with around twenty engravings plus their cuts: about 40 minutes.
- A job that took three hours on a 10 W diode machine can be completed here in less than half an hour.
That difference is what justifies a machine at this price in a production workshop. It is not that it does impossible things; it is that it does the same things in a fraction of the time, and that changes which orders you can accept. Corporate orders, which arrive in large quantities and with tight delivery deadlines, stop being intimidating.
How we use this table
Our method with an unknown material, in case it helps as a template:
- Look for the closest matching row above.
- Test grid on a scrap piece, varying power and speed around that value.
- Start quickly and work your way down, never the other way around.
- Write it down. The notebook is what turns a machine into a process.
We will keep updating this page as we get the hang of new materials.
August 2026 update: parameters can now be exported
The “write it down” part from point 4 has just improved. With xTool Studio 1.8, released on August 3, 2026, material parameters no longer live only inside the program:
- They work offline. The parameters support local storage, so you do not depend on the cloud for processing.
- They sync between local and cloud storage.
- They can be imported, exported, and managed in batches, instead of one at a time.
-
There is an exchange format,
.mxs, which also saves the processing preview image. It is useful for backups and for sending your table to someone else.
For what matters on this page, the change matters for two reasons. First, a table like the one above can now be shared as a file, not as a screenshot. Second, if you have two machines or switch computers, you no longer have to manually recreate what took you weeks to measure.
The same version adds project notes —written in the file itself and able to pop up as a reminder before processing—, which is exactly where it makes sense to write “this piece uses 90/12 and two passes.” It also increases free cloud storage from 500 MB to 2 GB.
The values in our table do not change: the machine still measures them, not the software. What changes is that they can now be saved and moved without copying them by hand.
Related xTool materials
Check the finish, thickness, composition and batch before production. A purchase link does not make settings or adhesion automatically transferable.
Affiliate links: Picasita may receive a commission without changing your price. The [xTool EU] code identifies the actual destination store.
Noemi and Javi from Picasita here. We are xTool affiliates, and our workshop is a brand demonstration space; we bought our P3 and paid for it on the day it went on sale. More about this blog and our relationship with xTool.
0 comments