File formats for data export
This article explains the formats used in files exported from the SENS motion web application: the timestamp columns and number formats in CSV files, and the binary file format for raw accelerometer data. For how to export data, see Data export. For example scripts in R and Python, see Processing of raw accelerometer data.
Timestamps and number formats#
When you export to CSV, you choose how time and numbers are written in the file. The options are explained below.
Timestamp columns#
Each row in the file has a time. You can include it in one or more of these formats. All three describe the same moment, here 8 January 2026 at 01:00 Danish time:
| Column | Example | What it is |
|---|---|---|
| Unix timestamp | 1767830400015 | The number of milliseconds since 1 January 1970 at 00:00 UTC. Easy for computers and scripts to calculate with, but not readable for people. |
| UTC ISO | 2026-01-08T00:00:00.015Z | Date and time in UTC (Coordinated Universal Time). UTC is the same everywhere and has no daylight saving time. |
| Local time ISO | 2026-01-08T01:00:00.015+01:00 | Date and time in the participant’s timezone, with the difference to UTC at the end (here +01:00). This is the time of day the participant experienced. |
If you are unsure, choose Local time ISO — it is the easiest to read and to relate to the participant’s day. Choose Unix timestamp if you will process the data with your own scripts. Only include the columns you need, as each column makes the file larger.
Delimiter & number format#
A CSV file is plain text, where a delimiter separates the columns. Countries write decimal numbers differently, so you can choose the combination that suits your software:
| Option | Example | Choose it if… |
|---|---|---|
| Semicolon, comma decimal | 0,27;-0,30;-0,66 | you open the file in Excel with Danish or other European settings, where a comma is used as the decimal sign. |
| Comma, dot decimal | 0.27,-0.30,-0.66 | you open the file in Excel with English settings, or process it in R, Python or similar tools. |
If everything ends up in one column when you open the file, try the other option, or see Making a CSV file readable.
Binary file format#
When you export raw data with the file format Binary, the file has no header — it consists only of samples, stored back to back. Each sample is 12 bytes, with the most significant byte first (big-endian):
- 6 bytes = Unix timestamp in milliseconds (UTC)
- 2 bytes = X, signed 16 bit integer x 0.0078125 G
- 2 bytes = Y, signed 16 bit integer x 0.0078125 G
- 2 bytes = Z, signed 16 bit integer x 0.0078125 G
Range: -4G to +4G, 10 bit resolution
Example (three samples, with the bytes shown in hexadecimal):
- 016C87D85F170022FFD9FFAC
- 016C87D85F6E0022FFD9FFAC
- 016C87D85FC50022FFD9FFAC
The first line decodes as:
- Timestamp 016C87D85F17 = 1565647200023 ms = 12 August 2019, 22:00:00.023 UTC
- X 0022 = 34 × 0.0078125 = 0.27 G
- Y FFD9 = −39 × 0.0078125 = −0.30 G
- Z FFAC = −84 × 0.0078125 = −0.66 G