Data format

DBF converter

A .dbf file is an xBase database table: a fixed-width binary table whose header describes every field's name, type, width and decimal places, followed by records of exactly that width. It is old, simple and still everywhere — most notably as the attribute half of an ESRI Shapefile.

Private for supported formats — processed in your browser

Convert supported files

Runs on your device

Drop DBF files here

Batch files can each use a different output. Nothing uploads for local conversions.

Working inputs include camera RAW, browser-local audio/video, PDF, CBZ/CBR comics, office documents, ebooks, markup, 3D models, structured text, images, and archives.

Where DBF comes from

The format arrived with dBASE II in 1980 and dBASE III in 1984, when dBASE was the dominant PC database and Ashton-Tate one of the largest software companies in the world. FoxPro and Clipper cloned it, the family became known as xBase, and the format outlived every product that defined it. It survives because ESRI chose it in 1998 for Shapefile attribute tables, so a format from the era of the IBM PC is still being written by GIS software today.

How DBF works

  • A 32-byte file header records the version byte, last-update date, record count, header length and record length.
  • Each field gets a 32-byte descriptor giving an 11-byte name, a one-character type (C, N, D, L, M, F), a width and a decimal count.
  • Records are fixed-width and begin with a one-byte deletion flag: a space means live, an asterisk means deleted-but-still-present.
  • There are no binary numeric types in the classic dialect — numbers are stored as right-aligned ASCII text and parsed by the reader.
  • Memo fields (type M) hold only a pointer into a separate .dbt or .fpt file, so a .dbf alone cannot resolve them.
  • The character encoding is a code page byte that is frequently absent or wrong, which is the most common source of mangled non-ASCII names.

When to use DBF

  • Reading Shapefile attribute tables without GIS software
  • Recovering data from legacy dBASE, FoxPro or Clipper systems
  • Extracting a records table produced by older municipal or scientific software
  • Converting a fixed-width table to CSV, JSON or a spreadsheet

Strengths and limitations

Strengths

  • Trivially simple structure that is still readable forty years on
  • Fixed-width records make random access a multiplication
  • Universally supported by GIS and legacy database tooling

Limitations

  • Field names are capped at ten characters, which mangles descriptive columns
  • Character encoding is unreliable, so non-ASCII text often needs a manual guess
  • Memo fields are useless without their sidecar file, which is easily separated

Compatibility

QGIS, ArcGIS, LibreOffice Base, GDAL/OGR and libraries for Python, R and JavaScript all read DBF. Novus Convert decodes bounded dBASE III and IV tables locally to JSON, CSV or XLSX, keeps deleted records separate from live ones instead of quietly including them, warns explicitly when the code page byte is missing or ambiguous, and never attempts to reach for a memo sidecar it was not given.

Field names, types, widths and decimal counts are read from the header and reported; the last-update date and code page are surfaced as warnings where they matter.