Field guide
Separate actual revisions from ordering and formatting
Keep the matching rule, numerical differences and source evidence together.
Track revisions before using a replacement point file
Compare survey point files when a revised export arrives and you need a clear account of what changed. Choose the earlier file as the old version and the replacement as the new version. The result separates added points, removed points, modified records, differences that do not exceed your reporting thresholds, and records that cannot be resolved safely. This makes a useful handover checklist before a crew loads revised setting-out data or a drafter replaces an earlier point table.
This is a file revision review, not an assessment of construction accuracy. There is no acceptance percentage, design approval or statement that an observed position is correct. A coordinate difference may represent an intentional design revision, a changed export setting or an error requiring investigation. Keep the version information and the reason for a revision with the project; this tool cannot infer those facts from the numbers.
Read each source using its own field layout
Import UTF-8 CSV or TXT files, or paste delimited text. Each source can use comma, semicolon or tab separation and its own header arrangement. Confirm point ID, northing, easting, elevation and description separately for both versions. A PNEZD file and a PENZD file can be compared after the correct mapping, because the comparison uses the meaning of a column rather than its physical position. Quoted separators and multiline descriptions retain their source line ranges. XLSX workbooks are also accepted for this input: explicitly choose the worksheet, and replace formulas, dates or error cells with plain values before importing.
Each input may contain up to 100,000 point records and 10 MiB of text. Original fields, additional columns, filenames and source lines remain in the complete report. Only the five mapped survey fields participate in change classification. Additional columns and header wording are explicitly outside that classification, even if a note in an extra column differs. Review those fields separately if they are important to the handover.
Confirm the reference before interpreting differences
Both files must describe the same planar coordinate reference, local origin and height datum. Horizontal and elevation unit labels must agree. Selecting a unit label does not perform a conversion, and the tool will not translate or rotate one dataset to make it fit the other. Write a short reference note and confirm which file is older before running the comparison.
All reported differences are new minus old. A positive easting difference means the new value is farther east in the common coordinate frame; a positive height difference means the new value is higher. The horizontal difference is the length formed by the easting and northing differences. It is a coordinate displacement, not a ground distance, movement certification or excavation instruction.
Use exact identifiers instead of row positions
Unique complete point IDs match automatically, irrespective of file order. If a new export simply sorts the same points differently, that alone produces no coordinate change. Leading zeros, letter case and spaces remain part of the identifier: 001 and 1 are different text IDs in this comparison. The software does not renumber them or assume that points close together represent the same object.
Repeated IDs are not resolved by taking the first occurrence or overwriting an earlier observation. Missing IDs and duplicate groups remain visible as unresolved records unless you supply unambiguous source-row pairs. A changed ID without a mapping usually appears as a removal and an addition. That is more limited than recognizing a rename, so use a documented mapping when you know two differently named records correspond.
Map renamed or repeated points explicitly
The optional mapping is a two-column CSV with the exact header old_line,new_line. Enter original starting line numbers, counting the header and any earlier lines in each source. Every listed old record and every listed new record can be used only once. A quoted multiline description still belongs to one record identified by its starting line, rather than to several separate points.
Explicit pairs are processed first. Remaining records then match by unique exact IDs. If an explicit pair uses a counterpart that another record would otherwise claim, that leftover is unresolved rather than being silently paired twice. Incorrect, repeated or nonexistent line references stop the comparison. When a confirmed pair has different IDs, the ID change is reported even if its coordinates and description are identical.
Choose thresholds without hiding the underlying numbers
The default numeric thresholds are zero, so every representable nonzero coordinate difference is reported as a modification. You may enter a horizontal threshold and a separate absolute elevation threshold in the source units. A difference exactly equal to a threshold does not exceed it. This inclusive boundary is used consistently in the result; display formatting does not decide the classification.
Differences at or below a nonzero threshold are retained in their own category, with their values and source records available for review. They are not included in the modified-point download unless another field, such as the description or ID, is modified. The report therefore distinguishes no detected change from a small real difference that your chosen reporting rule treats as insignificant. Thresholds do not establish measurement accuracy or replace a project specification.
Distinguish missing data from an actual zero
When elevation columns exist in both sources, changing a blank height to zero is a change in whether a value is present. It is not treated as a zero height difference. Two absent heights do not produce a fabricated difference. If one source has no mapped elevation column, elevation is marked as not compared, even when the other source contains values. Description columns follow the same availability principle.
Invalid source rows remain unresolved and retain their raw fields. They do not become usable added points or apparently unchanged pairs. Numeric differences that cannot be represented reliably are also marked unresolved. Equivalent decimal spellings such as 1, 1.0 and 1.000 are numerically equal; formatting-only differences can be noted without classifying the point as moved. Description comparison, by contrast, uses the original text exactly.
Reproduce the four-point example
Load the paired synthetic example and read both inputs after confirming their columns. The old file contains IDs 001 through 004. The new file reorders the first three IDs, omits 004 and adds 005. Point 001 only changes the spelling of one number, so it has no detected change. Point 003 gains an explicitly recorded elevation of zero, so it is modified.
Point 002 moves by 0.03 east and 0.04 north and gains 0.01 in elevation. With the example thresholds of 0.05 horizontally and 0.01 vertically, it belongs to the at-or-below-threshold category. Expect three matched pairs, one modified point, one small-difference point, one unchanged pair, one addition and one removal. Set both thresholds to zero and compare again: point 002 becomes modified, while the addition and removal counts stay the same.
Inspect the table and local point overview together
Filter the preview by category or search either the old or new identifier. These controls change the visible rows only; they do not redefine the exported change list. Select a source-line button to inspect both complete raw records and highlight their positions. Clicking a plotted position locates its result row, and left or right arrow keys inspect positions using the keyboard.
The overview shows north upwards and east to the right, with grey old positions and blue new positions. It draws up to the first 2,000 available positions plus the currently selected record. This is an explicit display limit, not a comparison limit. Every accepted record still participates in matching and reports. The overview has no basemap, does not recognize a coordinate system, and is unsuitable for deciding physical site boundaries.
Export the evidence and the appropriate source records
Download the complete change CSV for a spreadsheet review, and the JSON report to retain both original inputs, mappings, thresholds, reference note, raw fields and unresolved records. The review CSV protects values that spreadsheet software might interpret as formulas. It is a report, not a replacement instrument point file. JSON retains the original values separately from the spreadsheet presentation.
There are three point-file downloads: modified points using new values, added points using new values, and removed points using old values. Each retains the appropriate source column order, header presence and record order, using comma separation. Each point output is limited to 10 MiB. Preview filters do not restrict these files. Modified records can also be passed locally to the checker for another field review; save the change report if you need the old values later.
Keep the review reproducible
Changing either input, its mapping or the comparison rules invalidates old results. A running task can be cancelled while retaining both inputs. Clearing the workspace removes the two current sources and comparison settings from this session; it does not delete reports already downloaded to your computer. The workflow does not automatically save project coordinates in browser storage or upload them to a processing service.
Keep a copy of both dated inputs alongside the complete comparison report. Before distributing replacement data, resolve duplicate IDs, missing fields and unexpected coordinate changes using project records. An unresolved item is a request for a decision, not permission for the program to guess. Test the chosen point format in the receiving application, and do not assume that a clean file comparison proves the revised design is approved.