A reliable estimate starts with a clear image and a scale you can check. FloorSQ uses a valid reference line first; if there is no valid line, it uses the selected reference room as a fallback.
The five errors that matter most
The biggest sources of error are the source image, the distance used to set the scale, text recognition, matching the right fallback room, boundary placement and the choice of which spaces to include. These errors can compound: a reference line drawn over the wrong distance or a slightly oversized fallback room creates the wrong scale, while an oversized property outline adds pixels to the total.
High resolution does not guarantee a good source image. A large, blurry screenshot can be less useful than a smaller original plan with crisp wall lines and readable dimensions.
- Image distortion: perspective, cropping, blur, compression or unequal scaling.
- Scale error: a reference line uses the wrong endpoints or known length, or text recognition misreads or mismatches a fallback room dimension.
- Room-boundary error: the calibration box does not follow the same wall face as the printed dimension.
- Outline error: a garage, patio, void or neighbouring plan is included or omitted.
- Definition error: the estimate and comparison figure use different area conventions.
Step 1: start with the best source
Use the original digital plan when possible. Crop out browser controls and unrelated listing content without clipping the plan. Keep every storey visible and avoid photographs taken at an angle.
Before uploading, find at least one distance you can trust. A room with two readable metric dimensions works well as the automatic fallback. If you know another exact distance, write it down so you can add a reference line later. If the plan contains multiple buildings or levels, decide which ones should be included.
Step 2: choose and check how the plan is scaled
If you know a trustworthy real-world distance, draw a reference line over exactly that span and enter its length in metres. A valid reference line takes priority over the selected room fallback. Do not infer the scale from assumed furniture, door, tile or appliance sizes.
Without a valid line, choose a large rectangular fallback room with clear opposing walls and an unambiguous label. Verify the printed numbers against the detected text. Select another suitable room to cross-check the result; a large difference is a warning to inspect both room boxes, their dimensions and the consistency of the drawing scale.
Step 3: place the reference line or room box accurately
For a reference line, place both endpoints on the exact span represented by the entered length. For the room fallback, zoom into the selected room, decide which wall boundary its printed dimensions describe, then move the room box or individual edges to follow that convention.
Do not move a line or room box just to produce the number you hope to see. It should follow a visible distance that you can confirm from the plan or another reliable source.
Step 4: review the whole-property outline
Inspect every part of the plan from top to bottom. Move, add or remove outline points wherever the detected outline does not follow the visible exterior boundary. Pay particular attention to garages, open balconies, patios, narrow connecting areas, stairs and detached structures.
Use the same inner, centre or outer wall position as the reference line or fallback room. If you want separate totals for internal and ancillary spaces, record which spaces are included so the meaning of each figure remains clear.
Step 5: decide whether the estimate is stable enough
A result is more credible when a measurement line confirms another known distance, several clear dimensions give similar answers, the outline follows the plan and small boundary adjustments cause only small changes. Warnings, missing walls, uncertain text readings, or a large difference between the reference line and room fallback all reduce confidence.
If you need a record of your checks, note the plan source, chosen convention and spaces included. For any decision with significant financial consequences, or for professional use, treat the result only as an estimate and obtain an independent measurement.