Malvern Zetasizer .dts

The Malvern Zetasizer software (Nano, Pro and Ultra) saves dynamic light scattering and zeta potential measurements as .dts files. OpenReadout returns one table row per size or zeta record: sample, time, temperature and the results the software computed (Z-average, PDI, peaks, zeta potential, mobility, conductivity). No prior art exists for this binary format. It was derived from public files of several depositors, one of whom also published the Zetasizer’s own table of zeta results. Provenance: docs/provenance/malvern-zetasizer.md. Crate: openreadout-biophys.

format id	files	reads	confidence
malvern-zetasizer-dts	.dts measurement files of the Zetasizer software (Nano, Pro, Ultra; versions 7.02-8.01)	size and zeta records: header, sample, results	evidence rubric (docs/assurance.md)

Not read: ZS Xplorer .zmes files; the size distributions, correlation functions, zeta distributions and phase plots stored in each record; molecular-weight and other record kinds (listed only).

Layout

A compound file ([MS-CFB]):

• Header: u16 1, a 16-byte identifier (c0 fe e4 a2 b2 26 d6 11 99 1b 00 90 27 9b 77 0c), u16 1, u32 the next record number.
• Deleted: the deleted records.
• REC<n>: one stream per record, n its record number.

A record is a sequence of fields without tags. Strings are u32 byte length, 0x01, UTF-16LE ending in NUL.

offset	field
0	u16 record schema (10 in software 7.02; 13 in 7.12-8.01)
2	u16 kind: 1 size, 2 zeta
12	u32 record number
16	string: software version (7.13, 8.00.4813)
then	f64 measurement date: OLE Automation days, local time
then	string: instrument serial (MAL…)
then	u32, f32, f32 measured temperature (°C)

Further on: the operator, the SOP path, the dispersant (one or more strings with their properties), the material and a 46-byte block that begins 01 00 00 00 01 00 and ends 02 00 00 00 01 00 00 02 00 00 00 01 00 00, then the sample name. The reader takes the string after the first material block of that shape.

Size results are located by their structure: f32 Z-average (d.nm), f32 PdI, u32 n and n f32 (a fit of the correlation function), then for the intensity, number and volume distributions in turn: u32 k + k peak means (d.nm), u32 k + k areas (%), u32 k + k widths (d.nm); a distribution’s areas sum to 100.

Zeta results: u32 5 + 5 f32 zeta peak areas (%, largest first), u32 5 + 5 zeta peak means (mV), f64 conductivity (mS/cm), then ten f32: applied voltage (V), zeta potential (mV), zeta deviation (mV), mobility (µm·cm/V·s), mobility deviation, five not identified; then u32 5 + 5 zeta peak widths and the mobility peak areas, means and widths (u32 5 + 5 each).

A result block is returned only when exactly one block of that structure is in the record; a record with none (an aborted measurement) or several is listed without results and with a finding.

What the reader returns

Table records, one row per record in record-number order:

column	unit
record	—	record number
kind	—	size, zeta or other
sample_name, measured_at	—	ISO local time
temperature	°C	measured
z_average, pdi	nm, —	size records
peak1_mean … peak3_width	nm, %, nm	intensity peaks (mean, area, width)
zeta_potential, zeta_deviation	mV	zeta records
mobility, mobility_deviation	µm·cm/(V·s)
conductivity	mS/cm
voltage	V	applied voltage (inferred)
zeta_peak1_mean … zeta_peak3_width	mV, %, mV	zeta peaks

Table peaks: every stored peak (record, distribution intensity/number/volume/zeta/mobility, peak, mean, area, width, unit).

Experiment: vendor Malvern Panalytical, model Zetasizer, serial and software version (first record), start and end (earliest and latest record), the sample name when every record has the same one, technique dynamic light scattering (all records size) or zeta-potential measurement (all zeta).

The export’s number mean, volume mean and diffusion coefficient are computed by the software from the distributions and the Z-average; they are not stored and not returned.

Validation

tests/series_oracle/mod.rs against the depositor’s Zetasizer results table (oracle/ zetasizer_oracle.py): 36 zeta records (software 7.12) — record, sample name, date (within the export’s second), temperature, zeta potential, mobility and conductivity — agree within the export’s rounding, 252 cells. The other files (7.02, 7.13, 8.00, 8.01) are read for structure: every record’s header walks and every size record with results has exactly one block of the size structure.

Size results are withheld. The size-result layout above was mapped on one depositor’s size records whose export matched it (Z-average, PdI, peaks), but that record is the source of a held-out file of another format and cannot be development evidence; no other public export of size records was found. Until one is in the corpus, z_average, pdi and the intensity peaks are NaN and a finding (size_results_withheld) says so; the records, samples, dates and temperatures of size records are returned.

Vocabulary (public API of openreadout-biophys, Zetasizer)

identifier	meaning
ZetasizerReader	reader of Zetasizer .dts files
ZETASIZER_FORMAT_ID	malvern-zetasizer-dts
FORMAT_ID	malvern-zetasizer-dts (module constant)
Peaks	one distribution’s peaks: mean, area, width
Size	size results: z_average, pdi, peaks
Zeta	zeta results: zeta, zeta_deviation, mobility, mobility_deviation, conductivity, voltage, zeta_peaks, mobility_peaks
Record	one record: stream, schema, kind, number, version, measured_at, serial, temperature, sample, size, zeta, matches

---

How this reader was derived, file by file: provenance log.