How we compare
What we check before a figure goes on the page
A number in a Laser Workbench guide is quoted from a named document with the date we read it, worked out from inputs we show, or marked as not stated.
A comparison table looks authoritative whether or not it is. Four machines, six columns, tidy figures — and nothing on the face of it tells you that two of those figures were measuring different things.
That is the failure this page exists to prevent, and it is easier to fall into than it sounds. So rather than describe a method in the abstract, here is one field, on two real product pages, read this week.
One field, two machines
The field is the simplest one a buyer looks up: how big a piece the machine takes. Both pages below are xTool’s, both were read on 29 September 2026, and the middle column is their wording throughout.
| Page | What it says | What we do with it |
|---|---|---|
| xTool S11 | “a large bed size of 23.93" x 15.16" inches and a spacious work area of 19.6" x 12.5" inches”Product description. The specification table then gives the working area per module: 19.6 × 12.5 in with the 40 W, and 19.6 × 13 in with the 20 W, 10 W and 2 W infrared modules. | Keep both figures, in separately named fields, and record that the working area depends on which module is fitted. Do not collapse them into one “size”. |
| xTool P2S2 | “Up to 26 x 14 inches bed size”; Product JSON-LD also says “The largest 23.6" × 12" Bed Size”; and the page separately gives “600mm x 305mm working area”Two different values carry the same bed-size label. The page title reads P2 while the product heading reads P2S. | Keep all three figures under the labels the source gives them. Treat the exact bed-size, model and variant relationship as unresolved, preserve the P2/P2S naming difference, and do not use either bed-size figure as a settled comparison value. |
The S1 page gives a bed size and a working area without defining the relationship between them. The P2/P2S page currently gives two different values under the same “bed size” label and a third value as “working area”, without defining the physical relationship among them or which model or variant each bed-size figure belongs to. Those labels alone do not support a like-for-like comparison.
So the rule is dull and it is the whole method: keep the source’s own term, and compare a figure only with a figure that was measured the same way. Where a page gives a figure a name, we use that name. Where it gives conflicting figures the same name, or additional figures under a different name without defining their relationship, we carry them all, mark the relationship unresolved and stop the comparison there.
An earlier version of this page said the xTool P2/P2S listing no longer contained a same-label size conflict. A fresh check on 29 September 2026 found that was wrong. The current first-party page gives 26 × 14 inches as “bed size” in visible and metadata copy and 23.6 × 12 inches as “Bed Size” in its Product JSON-LD, while separately giving 600 × 305 mm as “working area”.
The page title says P2 while the product heading says P2S. Until the model and variant relationship between those figures is established, we keep all three under the labels the source gives them and do not use either bed-size figure as a settled comparison value. This notice remains as part of the page’s correction history.
The words we use for evidence
Behind a figure sits a question the reader cannot see: where did this come from, and what is it good for? These are the labels we use to answer it. They matter because they change what a claim can be used for, not because a taxonomy is impressive.
| Label | What it means, and what you can do with it |
|---|---|
| Manufacturer-stated | The maker’s own page, manual or data sheet gives this value. We quote it in their words and attribute it. It tells you what is being sold, not what was measured independently. |
| Supplier document | A safety data sheet or equivalent from whoever sells the material. Authoritative about the product as supplied — its composition, hazards and disposal — and it need not say what happens under a laser. |
| Independent evidence | A third party publishes a result and the method, equipment, configuration and conditions behind it. Used only where those conditions match the comparison being made. |
| Laser Workbench tested | A documented test carried out by us, published with the machine, method, material, observations and limits. This label has never been used. No machine has been bought, borrowed or handled for this publication, and until one is, it stays unused. |
| Calculated | Worked out from inputs we show, with a formula you can repeat. Shown with its units, its assumptions and the date its inputs were priced — and built so you can replace our numbers with yours. |
| Editorial judgement | Our reading, against criteria stated before we looked. It can order a list or frame a decision. It cannot overrule a safety requirement, and it cannot fill a gap where evidence is missing. |
| Not stated | The document does not say. Printed as “not stated” and never quietly turned into zero, into a dash, or into a figure from a different model. |
| Open question | Current sources disagree, or use definitions that cannot be reconciled. Both readings stay visible, and anything that depends on the answer waits. |
When sources disagree
Published documents disagree in several ways. A product page and its own metadata can differ; a manual is revised and the listing is not; a figure applies to one module of a machine sold with four.
When it happens, we record both readings, say which document each came from and when we read it, and stop anything that depends on the answer. We do not average them, and we do not pick the one that makes a better story — including when the tidier reading would suit us. The alternative is a confident table nobody can defend, which is worse than an honest gap.
What a conclusion is allowed to say
Where a guide reaches a conclusion, it is bounded: best fit for a stated use, on stated criteria, as of a stated date. Not best for everyone, and not best in general.
In practice none of the six guides names a best machine. What to work out first hands you six questions rather than a shortlist. The requirements builder writes your brief and then lists what you still have to ask a supplier. That is deliberate: we can see what the documents say, and we cannot see your material, your room or your week.
Money, and how it is kept out
Whether a product carries a commission plays no part in whether it appears, how its evidence is labelled, or what a guide concludes. We will write about a machine we earn nothing from, and leave out one we would. Any commercial relationship is disclosed separately, on the affiliate disclosure page.
Dates, and getting it wrong
Consequential figures carry the date we read them, because manufacturers change pages without notice — as the correction above shows. When a source changes, we re-read the field and the conclusion that rested on it rather than editing one number and moving on.
If something here is wrong, the corrections page explains how to tell us and what happens next. About sets out who publishes this and what we can honestly claim.
Sources
Both pages and their current first-party product data were read in full on 29 September 2026. Each figure above is attributed to the page surface or product data that states it. All product figures are the manufacturers’ own.
- xTool UK, S1 product page. Manufacturer’s product page. Read 29 September 2026. Source of the quoted bed-size and work-area sentence and of the per-module working areas. The page does not say what distinguishes the two terms.
- xTool UK, P2S product page. Manufacturer’s product page and current Product JSON-LD. Read 29 September 2026. Source of the two bed-size values, the separately labelled working-area figure and the difference between the page title (P2) and product heading (P2S). The exact bed-size, model and variant relationship remains unresolved.
Publication standards: editorial policy · corrections · affiliate disclosure · contact.