Overview

The International Federation of Digital Seismograph Networks (FDSN) defines miniSEED as a format for digital data and related information. The primary intended uses are data collection, archiving and exchange of seismological data. The format is also appropriate for time series data from other geophysical measurements such as pressure, temperature, tilt, etc. In addition to the time series, storage of related state-of-health and parameters documenting the state of the recording system are supported. The FDSN metadata counterpart of miniSEED is StationXML which is used to describe characteristics needed to interpret the data such as location, instrument response, etc.

Versioning

Starting with version 3 of miniSEED, the specification version is a single integer. Any non-backward compatible change to the structure of the header or record results in an increment to this number. The specification version corresponds to field 2, Format version, in the header of records written to conform with this version of the specification. There are no minor revisions. However, the addition of new data encodings or new reserved headers is considered backwards compatible. The FDSN may add encodings and extra headers within the FDSN key or make editorial changes to the specification without change to the major version, resulting only in a new revision of the specification document. Note that all parsing software MUST allow for the potential of unknown, new data encodings and new extra headers as these are subject to extension. This requirement allows older software to parse future records successfully within the same specification major version, even though they may be unable to decompress the data payload for new encodings.

This document is miniSEED v3 rev. 2023-01-18.

Note

This specification defines miniSEED 3. See Background for information on earlier versions.

Record definition

The fundamental unit of the format is a data record. A time series is commonly stored and exchanged as a sequence of these records. There is no interdependence of records, each is independent. There are data encodings for integers, floats, text or compressed data samples. To limit problems with timing system drift and resolution in addition to practical issues of subsetting and resource limitation for readers of the data, typical record lengths for raw data generation and archiving are recommended to be in the range of 256 and 4096 bytes.

Record layout and fields

A record is composed of a header followed by a data payload. The byte order of binary fields in the header must be least significant byte first (little endian).

The total length of a record is variable and is the sum of 40 (length of fixed section of header), field 10 (length of identifier), field 11 (length of extra headers), field 12 (length of payload).

Field	Description	Type	Length	Offset	Content
1	Record header indicator	CHAR	2	0	ASCII ‘MS’
2	Format version	UINT8	1	2	Value of 3
3	Flags	UINT8	1	3
	Record start time
4a	Nanosecond (0 - 999999999)	UINT32	4	4
4b	Year (0-65535)	UINT16	2	8
4c	Day-of-year (1 - 366)	UINT16	2	10
4d	Hour (0 - 23)	UINT8	1	12
4e	Minute (0 - 59)	UINT8	1	13
4f	Second (0 - 60)	UINT8	1	14
5	Data payload encoding	UINT8	1	15	Data Encodings
6	Sample rate/period	FLOAT64	8	16
7	Number of samples	UINT32	4	24
8	CRC of the record	UINT32	4	28
9	Data publication version	UINT8	1	32
10	Length of identifier	UINT8	1	33
11	Length of extra headers	UINT16	2	34
12	Length of data payload	UINT32	4	36
13	Source identifier	CHAR	V	40	URI identifier
14	Extra header fields	JSON	V	40 + field 10
15	Data payload	encoded	V	40 + field 10 + field 11

All length values are specified in bytes, which are assumed to be 8-bits in length. Data types for each field are defined as follows:

CHAR: ASCII encoded character data.
UINT8: Unsigned 8-bit integer.
UINT16: Unsigned 16-bit integer (little endian byte order).
UINT32: Unsigned 32-bit integer (little endian byte order).
FLOAT64: IEEE-754 64-bit floating point number (little endian byte order).
JSON: JSON Data Interchange Standard.

Description of record fields

1: CHAR: Record header indicator. Literal, 2-character sequence “MS”, ASCII 77 and 83, designating the start of a record.

2: UINT8: Format version. Set to 3 for this version. When a non-backwards compatible change is introduced the version will be incremented.

3

UINT8: Flags. Bit field flags, with bits 0-7 defined as:

Calibration signals present. [same as SEED 2.4 FSDH, field 12, bit 0]
Time tag is questionable. [same as SEED 2.4 FSDH, field 14, bit 7]
Clock locked. [same as SEED 2.4 FSDH, field 13, bit 5]
Reserved for future use.
Reserved for future use.
Reserved for future use.
Reserved for future use.
Reserved for future use.

4

Record start time, time of the first data sample. A representation of UTC using individual fields for:

nanosecond

year

day-of-year

hour

minute

second

A 60 second value is used to represent a time value during a positive leap second. If no time series data are included in this record, the time should be relevant for whatever headers or flags are included.

5: UINT8: Data payload encoding. A code indicating the encoding format, see Section 4: Data encoding codes for a list of valid codes. If no data payload is included set this value to 0.

6: FLOAT64: Sample rate/period. Sample rate encoded in 64-bit IEEE-754 floating point format. When the value is positive it represents the rate in samples per second, when it is negative it represents the sample period in seconds. Creators should use the negative value sample period notation for rates less than 1 samples per second to retain resolution. Set to 0.0 if no time series data are included in the record.

7: UINT32: Number of samples. Total number of data samples in the data payload. Set to 0 if no samples (header-only records) or unknown number of samples (e.g. for opaque payload encoding).

8: UINT32: CRC of the record. CRC-32C (Castagnoli) value of the complete record with the 4-byte CRC field set to zeros. The CRC-32C (Castagnoli) algorithm with polynomial 0x1EDC6F41 (reversed 0x82F63B78) to be used is defined in RFC 3309, which further includes references to the relevant background material.

9: UINT8: Data publication version. Values should only be considered relative to each other for data from the same data center. Semantics may vary between data centers but generally larger values denote later and more preferred data. Recommended values: 1 for raw data, 2+ for revisions produced later, incremented for each revision. A value of 0 indicates unknown version such as when data are converted to miniSEED from another format. Changes to this value for user-versioning are not recommended, instead an extra header should be used to allow for user-versioning of different derivatives of the data.

10: UINT8: Length of identifier. Length, in bytes, of source identifier in field 13.

11: UINT16: Length of extra headers. Length, in bytes, of extra headers in field 14. If no extra headers, set this value to 0.

12: UINT32: Length of data payload. Length, in bytes, of data payload starting in field 15. If no data payload is present, set this value to 0. Note that no padding is permitted in the data record itself, although padding may exist within the payload depending on the type of encoding used.

13: CHAR: Source identifier. A unique identifier of the source of the data contained in the record. Recommended to use URI-based identfiers. Commonly an FDSN Source Identifier.

14

JSON: Extra header fields. Extra fields of variable length encoded in JavaScript Object Notation (JSON) Data Interchange Standard as defined by ECMA-404. It is strongly recommended to store compact JSON, containing no non-data white space, in this field to avoid wasted space.

A reserved set of headers fields is defined by the FDSN, see Extra Headers. Other header fields may be present and should be defined by the organization that created them.

15: encoded: Data payload. Length indicated in field 12, encoding indicated in field 5.

Data Encodings

Data payload encodings in the format are identified by a code (number). These codes are assigned by the FDSN. A list of valid codes are as follows:

Code	Description
0	Text, UTF-8 allowed, use ASCII for maximum portability, no structure defined
1	16-bit integer (two’s complement), little endian byte order
3	32-bit integer (two’s complement), little endian byte order
4	32-bit floats (IEEE float), little endian byte order
5	64-bit floats (IEEE double), little endian byte order
10	Steim-1 integer compression, big endian byte order
11	Steim-2 integer compression, big endian byte order
19	Steim-3 integer compression, big endian (not in common use in archives)
100	Opaque data - only for use in special scenarios, not intended for archiving

Overview and description of the Steim-1 and Steim-2 compression encodings may be found in the SEED 2.4 manual, Appendix B.

Retroactive future encodings

New data encodings may be added to the format in the future without incrementing the format version. There is no default encoding, readers must check the encoding value to determine if the encoding is supported.

Retired encoding values, not allowed in this specification

The following numeric codes were used in earlier miniSEED versions and should not be used for encodings defined in the future:

2: 24-bit integers
12: GEOSCOPE multiplexed format 24-bit integer
13: GEOSCOPE multiplexed format 16-bit gain ranged, 3-bit exponent
14: GEOSCOPE multiplexed format 16-bit gain ranged, 4-bit exponent
15: US National Network compression
16: CDSN 16-bit gain ranged
17: Graefenberg 16-bit gain ranged
18: IPG-Strasbourg 16-bit gain ranged
30: SRO format
31: HGLP format
32: DWWSSN gain ranged format
33: RSTN 16-bit gain ranged format

Extra Headers

The extra headers are encoded in the JSON Data Interchange Standard as defined by ECMA-404. All extra headers are optional as far as the format is concerned.

Extra headers must follow these rules:

All extra headers are contained in an anonymous (unnamed) object
All entries are key-value pairs where values can be any valid JSON type
The key value of “FDSN” in the root object is reserved for values defined by the FDSN

Validation

Extra headers are specified and documented in JSON Schema.

FDSN Reserved Headers

The “FDSN” key at the root of the extra headers are reserved for definition by the FDSN.

See FDSN Reserved Headers for documentation of these headers.

See A: Extra header examples for an example of FDSN extra headers.

Guidelines for Extension

Network operators, manufacturers, data centers, users and other agencies may wish to define their own extra headers. The following guidelines should be considered, in particular for data that is expected to reside in a public repository:

All headers defined by a group or agency should be contained in a JSON object that is the value of a key in the root container with a clearly identifiable name, i.e. at the same level as “FDSN”.
Creation of a JSON Schema document describing the field(s) is strongly recommended. The schema should be submitted to the FDSN to be made publically available.
Headers that would be generally useful should be submitted to the FDSN for consideration of being added to the reserved headers for general definition and use.

Multiple JSON Schema documents are easily combined for use in validating extra headers that may contain headers defined in multiple schema documents.

See Non-FDSN extra headers for an example of non-FDSN extra headers.

FDSN Reserved Headers

The “FDSN” key at the root of the Extra Headers are reserved for definition by the FDSN. See A: Extra header examples for an example of FDSN extra headers.

The FDSN reserved headers are defined and documented in JSON Schema.

Download the FDSN Extra Header schema v1.0.

The documentation and schema of these headers may be browsed here:

Schema Docs

Type: object

Extra headers in miniSEED 3 data format

root FDSN

Type: object

Reserved extra headers defined by the FDSN

No Additional Properties

root FDSN Time

Type: object

Headers related to data timing and recording system clock

No Additional Properties

root FDSN Time Quality

Type: integer

DEPRECATED Timing quality. A vendor specific value from 0 to 100% of maximum accuracy. [same as SEED 2.4 Blockette 1001, field 3] It is recommended to use MaxEstimatedError instead.

root FDSN Time Correction

Type: number

Time correction applied to record start time in seconds. [same as SEED 2.4 FSDH, field 16]

root FDSN Time MaxEstimatedError

Type: number

Maximum estimated timing error in seconds.

root FDSN Time LeapSecond

Type: integer

If present, one or more leap seconds occuring during this record. The value specifies the number of leap seconds and direction. For example, use 1 to specify a single positive leap second and -1 to specify a single negative leap second. [incorporates SEED 2.4 FSDH, field 12, bits 4 and 5]

root FDSN Time Exception

Type: array of object

List of timing exceptions

No Additional Items

Each item of this array must be:

root FDSN Time Exception Exception items

Type: object
No Additional Properties

root FDSN Time Exception Exception items Time

Type: stringFormat: date-time

Time of timing exeption. [same as SEED 2.4 Blockette 500, field 4]

root FDSN Time Exception Exception items VCOCorrection

Type: number

VCO correction, a percentage from 0 to 100% of VCO control value, where 0 is the slowest and 100 is the fastest. [same as SEED 204 Blockette 500, field 3]

root FDSN Time Exception Exception items ReceptionQuality

Type: integer

Reception quality as a percent of maximum clock accuracy based only on information from the clock. [same as SEED 2.4 Blockette 500, field 6]

root FDSN Time Exception Exception items Count

Type: integer

Exception count, with meaning based on type of exception, such as 15 missing time marks. [same as SEED 2.4 Blockette 500, field 7]

root FDSN Time Exception Exception items Type

Type: string

Description of clock exception, such as MISSING TIMEMARK. [same as SEED 2.4 Blockette 500, field 8]

root FDSN Time Exception Exception items ClockStatus

Type: string

Description of clock status, clock specific parameters such as the station for an Omega clock or satellite signal to noise ratios for GPS clocks. [same as SEED 2.4 Blockette 500, field 10]

root FDSN Event

Type: object

Headers related to event detection and progression

No Additional Properties

root FDSN Event Begin

Type: boolean

An event starts in this record. [same as SEED 2.4 FSDH, field 12, bit 2]

root FDSN Event End

Type: boolean

An in-progress event ends in this record, i.e. the detection algorithm de-triggers. [same as SEED 2.4 FSDH, field 12, bit 3]

root FDSN Event InProgress

Type: boolean

An event is in progress. [same as SEED 2.4 FSDH, field 12, bit 6]

root FDSN Event Detection

Type: array of object

List of event detections

No Additional Items

Each item of this array must be:

root FDSN Event Detection Detection items

Type: object
No Additional Properties

root FDSN Event Detection Detection items Type

Type: string

Type of event detection, e.g. MURDOCK

root FDSN Event Detection Detection items SignalAmplitude

Type: number

Event amplitude of signal in counts. [same as SEED 2.4 Blockettews 200 and 201, field 3]

root FDSN Event Detection Detection items SignalPeriod

Type: number

Event period of signal in seconds. [same as SEED 2.4 Blockettews 200 and 201, field 4]

root FDSN Event Detection Detection items BackgroundEstimate

Type: number

Event background estimate in counts. [same as SEED 2.4 Blockettews 200 and 201, field 5]

root FDSN Event Detection Detection items Wave

Type: string

Event detection wave, use values of DILATATION or COMPRESSION if determined. [same as SEED 2.4 Blockettews 200 and 201, field 6, bit 0]

root FDSN Event Detection Detection items Units

Type: string

Units of amplitude band background estimate, use COUNTS for unprocessed signal. [similar to SEED 2.4 Blockettews 200, field 6, bit 1]

root FDSN Event Detection Detection items OnsetTime

Type: stringFormat: date-time

Event signal onset time. [same as SEED 2.4 Blockettews 200 and 201, field 8]

root FDSN Event Detection Detection items MEDSNR

Type: array of number

Murdock event detection signal-to-noise ratios. [same as SEED 2.4 Blockette 201, field 9]

No Additional Items

Each item of this array must be:

root FDSN Event Detection Detection items MEDSNR MEDSNR items

Type: number

root FDSN Event Detection Detection items MEDLookback

Type: integer

Murdock event detction lookback value: 0, 1 or 2. [same as SEED 2.4 Blockette 201, field 10]

root FDSN Event Detection Detection items MEDPickAlgorithm

Type: integer

Murdock event detection pick algorithm: 0 or 1. [same as SEED 2.4 Blockette 201, field 11]

root FDSN Event Detection Detection items Detector

Type: string

Name of the event detector. [same as SEED 2.4 Blockette 200, field 9 and Blockette 201, field 12]

root FDSN Calibration

Type: object

Headers related to calibrations

No Additional Properties

root FDSN Calibration Sequence

Type: array of object

List of calibration sequences

No Additional Items

Each item of this array must be:

root FDSN Calibration Sequence Sequence items

Type: object

List of calibrations

No Additional Properties

root FDSN Calibration Sequence Sequence items Type

Type: string

Type of calibration: STEP, SINE, PSEUDORANDOM, GENERIC.

root FDSN Calibration Sequence Sequence items BeginTime

Type: stringFormat: date-time

Calibration begin time. [similar to SEED 2.4 Blockettes 300,310,320,390, field 3]

root FDSN Calibration Sequence Sequence items EndTime

Type: stringFormat: date-time

Calibration end time. [same as SEED 2.4 Blockette 395, field 3]

root FDSN Calibration Sequence Sequence items Steps

Type: number

Number of step calibrations in sequence. [same as SEED 2.4 Blockette 300, field 4]

root FDSN Calibration Sequence Sequence items StepFirstPulsePositive

Type: boolean

Step calibration, first pulse is positive. [same as SEED 2.4 Blockette 300, field 5, bit 0]

root FDSN Calibration Sequence Sequence items StepAlternateSign

Type: boolean

Step calibration alternate sign. [same as SEED 2.4 Blockettes 300, field 5, bit 1]

root FDSN Calibration Sequence Sequence items Trigger

Type: string

Calibration trigger, use AUTOMATIC or MANUAL. [same as SEED 2.4 Blockettes 300,310,320,390, field 5, bit 2]

root FDSN Calibration Sequence Sequence items Continued

Type: boolean

Calibration is continued from previous records. [same as SEED 2.4 Blockettes 300,310,320,390, field 5, bit 3]

root FDSN Calibration Sequence Sequence items Amplitude

Type: number

Amplitude of calibration signal. For pseudo-random calibrations, this should be the peak-to-peak amplitude of the steps. [same as SEED 2.4 Blockettes 300,310, field 8 and Blockettes 320,390, field 7]

root FDSN Calibration Sequence Sequence items InputUnits

Type: string

Units of calibration input, usually volts or amps. [same as SEED 2.4 Blockette 52, field 9]

root FDSN Calibration Sequence Sequence items AmplitudeRange

Type: string

Amplitude range of calibration, use values of PEAKTOPEAK, ZEROTOPEAK, RMS OR RANDOM. [same as SEED 2.4 Blockette 310, field 5, bits 4,5,6 and Blockette 320, field 5, bit 4]

root FDSN Calibration Sequence Sequence items Duration

Type: number

Duration of calibration in seconds. For step calibrations this is the duration of the step. [same as SEED 2.4 Blockettes 300,310,320,390, field 6]

root FDSN Calibration Sequence Sequence items SinePeriod

Type: number

Period of sine calibrations in seconds. [same as SEED 2.4 Blockette 310, field 7]

root FDSN Calibration Sequence Sequence items StepBetween

Type: number

Interval between times a step calibration is on in seconds. [same as SEED 2.4 Blockette 300, field 7]

root FDSN Calibration Sequence Sequence items InputChannel

Type: string

Channel containing the calibration input. [same as SEED 2.4 Blockettes 300,310, field 9 and Blockettes 320,390, field 8]

root FDSN Calibration Sequence Sequence items ReferenceAmplitude

Type: number

Reference amplitude. This is a user-defined value that indicates either the voltage or amperage of the calibration signal when the calibrator is set to 0dB. [same as SEED 2.4 Blockettes 300,310, field 11 and Blockette 320, field 10]

root FDSN Calibration Sequence Sequence items Coupling

Type: string

Coupling of the calibration signal, such as RESISTIVE or CAPACITIVE. [same as SEED 2.4 Blockettes 300,310, field 12 and Blockette 320, field 11]

root FDSN Calibration Sequence Sequence items Rolloff

Type: string

Rolloff characteristics for any filters used on the calibrator, such as 3dB@10Hz. [same as SEED 2.4 Blockettes 300,310, field 13 and Blockette 320, field 12]

root FDSN Calibration Sequence Sequence items Noise

Type: string

Noise characteristics for pseudo-random calibrations, such as WHITE or RED. [same as SEED 2.4 Blockette 320, field 13]

root FDSN Recenter

Type: object

Headers related to sensor recentering (mass, gimble, etc.)

No Additional Properties

root FDSN Recenter Sequence

Type: array of object

List of recentering sequences

No Additional Items

Each item of this array must be:

root FDSN Recenter Sequence Sequence items

Type: object

List of recenterings

No Additional Properties

root FDSN Recenter Sequence Sequence items Type

Type: string

Type of recenter: Mass, Gimbal, etc. If omitted a mass recenter may be assumed.

root FDSN Recenter Sequence Sequence items BeginTime

Type: stringFormat: date-time

Recenter begin time.

root FDSN Recenter Sequence Sequence items EndTime

Type: stringFormat: date-time

Estimate of recenter end time.

root FDSN Recenter Sequence Sequence items Trigger

Type: string

Calibration trigger, use AUTOMATIC or MANUAL.

root FDSN Flags

Type: object
No Additional Properties

root FDSN Flags MassPositionOffscale

Type: boolean

Sensor mass position is offscale as defined by the vendor or operator.

root FDSN Flags AmplifierSaturation

Type: boolean

Amplifier saturation detected. [same as SEED 2.4 FSDH, field 14, bit 0]

root FDSN Flags DigitizerClipping

Type: boolean

Digitizer clipping detected. [same as SEED 2.4 FSDH, field 14, bit 1]

root FDSN Flags Spikes

Type: boolean

Spikes detected. [same as SEED 2.4 FSDH, field 14, bit 2]

root FDSN Flags Glitches

Type: boolean

Glitches detected. [same as SEED 2.4 FSDH, field 14, bit 3]

root FDSN Flags FilterCharging

Type: boolean

A digital filter may be charging. [same as SEED 2.4 FDSH, field 14, bit 6]

root FDSN Flags StationVolumeParityError

Type: boolean

DEPRECATED Station volume parity error possibly present. [same as SEED 2.4 FDSH, field 13, bit 0]

root FDSN Flags LongRecordRead

Type: boolean

DEPRECATED Long record read (possibly no problem). [same as SEED 2.4 FDSH, field 13, bit 1]

root FDSN Flags ShortRecordRead

Type: boolean

DEPRECATED Short record read (record padded). [same as SEED 2.4 FDSH, field 13, bit 2]

root FDSN Flags StartOfTimeSeries

Type: boolean

DEPRECATED Start of time series. [same as SEED 2.4 FDSH, field 13, bit 3]

root FDSN Flags EndOfTimeSeries

Type: boolean

DEPRECATED End of time series. [same as SEED 2.4 FDSH, field 13, bit 4]

root FDSN Flags MissingData

Type: boolean

DEPRECATED Missing data. [same as SEED 2.4 FDSH, field 14, bit 4]

root FDSN Flags TelemetrySyncError

Type: boolean

DEPRECATED Telemetry synchronization error. [same as SEED 2.4 FDSH, field 14, bit 5]

root FDSN Logger

Type: object

Description of data logger

No Additional Properties

root FDSN Logger Model

Type: string

Model of equipment

root FDSN Logger Serial

Type: string

Serial number of equipment

root FDSN Sensor

Type: object

Description of sensor

Same definition as Logger

root FDSN Clock

Type: object

Description of clock system

Same definition as Logger

root FDSN ProvenanceURI

Type: string

An identifier for a provenance description

root FDSN DataQuality

Type: string

Data quality indicator, use D, R, Q or M. [same as SEED 2.4 FSDH, field 2]

root FDSN Sequence

Type: integer

Data record sequence number. [same as SEED 2.4 FSDH, field 1]

When not present, the boolean values in the FDSN reserved headers should be considered to be false unless otherwise documented. Such values do not need to be included when the value is false.

Background

The Standard for the Exchange of Earthquake Data (SEED) was adopted by the FDSN in the 1987 and served as the dominant standard for seismological research data archiving and exchange for many decades.

The previous specification of miniSEED, latest 2.4, is defined as a subset of SEED that contains only data records.

Changes relative to 2.4

An overview of significant changes between miniSEED 2.4 and this specification:

Adoption of FDSN Source Identifiers, replacing independent SEED codes (network, station, location, channel).
Incorporate critical details previously in blockettes (actual sample rate, encoding, microseconds) into the fixed section of the data header
Increase sample rate/period representation to a 64-bit floating point value
Increase start time resolution to nanoseconds
Specify fixed byte order (little endian) for the binary portions of the headers and define a byte order for each data encoding
Drop legacy data encodings and reserve their values so they are not used again in the future
Add a format version
Add a data publication version
Add CRC field for validating integrity of record
Add a “mass position off scale” flag
Add “Recenter” (mass, gimbal, etc.) headers
Add “ProvenanceURI” header to identify provenance documentation
Replace the blockette structure with flexible extra header construct:
- Specify a reserved set of extra headers defined by the FDSN, provide schema for validation
- Previous flags and blockette contents defined in reserved extra headers
- Allow arbitrary headers to be included in a record
Remove the restriction on record length to be powers of 2, allow variable length

Near complete preservation of miniSEED 2.4 data. Information that is not retained is limited to: clock model specification per timing exception (current specification only allows a single clock model specification per record), Blockettes 400 (Beam) & 405 (Beam Delay) and Blockette 2000 (Opaque Data).

Appendix

A: Extra header examples

The example extra header structures below include additional white space and formatting for readability, which should not be used in an actual record.

FDSN reserved extra headers

A relatively simple example of reserved FDSN extra headers that contains a timing quality value and an event detection is:

{
  "FDSN": {
    "Time": {
      "Quality": 100,
      "Correction": 1.234
    },
    "Event": {
      "Begin": true,
      "End": true,
      "InProgress": true,
      "Detection": [
        {
          "Type": "MURDOCK",
          "SignalAmplitude": 80,
          "SignalPeriod": 0.4,
          "BackgroundEstimate": 18,
          "Wave": "DILATATION",
          "Units": "COUNTS",
          "OnsetTime": "2022-06-05T20:32:39.120000Z",
          "MEDSNR": [ 1, 3, 2, 1, 4, 0 ],
          "MEDLookback": 2,
          "MEDPickAlgorithm": 0,
          "Detector": "Z_SPWWSS"
        }
      ]
    }
  }
}

An example of the reserved FDSN headers defined by the FDSN is provided below. In this (pathalogical) illustration, all reserved fields are represented.

{
  "FDSN": {
    "Time": {
      "Quality": 100,
      "Correction": 1.234,
      "MaxEstimatedError": 1e-06,
      "LeapSecond": 1,
      "Exception": [
        {
          "Time": "2022-05-06T20:32:41.12Z",
          "VCOCorrection": 50.7812,
          "ReceptionQuality": 80,
          "Count": 23,
          "Type": "Valid Timemark",
          "ClockStatus": "SNR=48,51,51,50,50,48,46,48,48,45,45"
        },
        {
          "Time": "2022-05-06T20:32:42.185Z",
          "VCOCorrection": 44.1313,
          "ReceptionQuality": 55,
          "Count": 19690,
          "Type": "Missing timemarks",
          "ClockStatus": "SNR=50,48,46,48,48,45,45"
        }
      ]
    },
    "Event": {
      "Begin": true,
      "End": true,
      "InProgress": true,
      "Detection": [
        {
          "Type": "GENERIC",
          "SignalAmplitude": 80,
          "SignalPeriod": 0.4,
          "BackgroundEstimate": 18,
          "OnsetTime": "2022-05-06T20:32:39.120000Z",
          "Detector": "Dalek STA/LTA"
        },
        {
          "Type": "MURDOCK",
          "SignalAmplitude": 80,
          "SignalPeriod": 0.4,
          "BackgroundEstimate": 18,
          "Wave": "DILATATION",
          "Units": "COUNTS",
          "OnsetTime": "2022-05-06T20:32:39.185000Z",
          "MEDSNR": [ 1, 3, 2, 1, 4, 0 ],
          "MEDLookback": 2,
          "MEDPickAlgorithm": 0,
          "Detector": "Z_SPWWSS"
        }
      ]
    },
    "Calibration": {
      "Sequence": [
        {
          "Type": "Step",
          "BeginTime": "2022-05-06T20:32:39.120000Z",
          "Steps": 12,
          "StepFirstPulsePositive": true,
          "StepAlternateSign": true,
          "Trigger": "AUTOMATIC",
          "Continued": false,
          "Amplitude": 1345,
          "InputUnits": "COUNTS",
          "Duration": 603.456,
          "SinePeriod": 5.0,
          "StepBetween": 500.0,
          "InputChannel": "CAL",
          "ReferenceAmplitude": 45.8,
          "Coupling": "RESISTIVE",
          "Rolloff": "Description of rolloff"
        },
        {
          "Type": "Step",
          "EndTime": "2022-05-06T20:32:39.120000Z"
        },
        {
          "Type": "Sine",
          "BeginTime": "2022-05-06T20:32:39.120000Z",
          "EndTime": "2022-05-06T20:32:39.120000Z",
          "Trigger": "MANUAL",
          "Continued": true,
          "Amplitude": 1345,
          "InputUnits": "COUNTS",
          "AmplitudeRange": "PEAKTOPEAK",
          "SinePeriod": 5.0,
          "InputChannel": "CAL",
          "ReferenceAmplitude": 45.8,
          "Coupling": "RESISTIVE",
          "Rolloff": "Description of rolloff"
        },
        {
          "Type": "PseudoRandom",
          "BeginTime": "2022-05-06T20:32:39.120000Z",
          "EndTime": "2022-05-06T20:32:39.120000Z",
          "Trigger": "MANUAL",
          "Amplitude": 0.0001,
          "InputUnits": "M/S",
          "Duration": 300,
          "InputChannel": "CAL",
          "ReferenceAmplitude": 45.8,
          "Coupling": "CAPACITIVE",
          "Rolloff": "Very randomly",
          "Noise": "White"
        },
        {
          "Type": "Generic",
          "BeginTime": "2022-05-06T20:32:39.120000Z",
          "EndTime": "2022-05-06T20:32:39.120000Z",
          "Trigger": "MANUAL",
          "Amplitude": 1345,
          "Duration": 100
        }
      ]
    },
    "Recenter": {
      "Sequence": [
        {
          "Type": "Gimbal",
          "BeginTime": "2022-05-06T20:32:39.120000Z",
          "EndTime": "2022-05-06T20:32:39.120000Z",
          "Trigger": "AUTOMATIC"
        },
        {
          "BeginTime": "2022-05-06T20:32:40.120000Z"
        }
      ]
    },
    "Flags": {
      "MassPositionOffscale": true,
      "AmplifierSaturation": true,
      "DigitizerClipping": true,
      "Spikes": true,
      "Glitches": true,
      "FilterCharging": true,
      "StationVolumeParityError": true,
      "LongRecordRead": true,
      "ShortRecordRead": true,
      "StartOfTimeSeries": true,
      "EndOfTimeSeries": true,
      "MissingData": true,
      "TelemetrySyncError": true
    },
    "Logger": {
      "Model": "DM24",
      "Serial": "A4567"
    },
    "Sensor": {
      "Model": "T240",
      "Serial": "123123"
    },
    "Clock": {
      "Model": "P273T11N16",
      "Serial": "24A00000"
    },
    "ProvenanceURI": "prov:sp001_wf_f84fb9a",
    "DataQuality": "D",
    "Sequence": 123456
  }
}

Non-FDSN extra headers

An example of reserved FDSN headers combined with other top-level headers, illustrating how custom headers may be added.

{
  "FDSN": {
    "Time": {
      "Quality": 90
    }
  },
  "Manufacturer123": {
    "Metadata": {
      "FilamentCurrent": 16.4,
      "HyperCoordinates": "1.1789:965402:73324@3.14159"
    }
  },
  "OperatorXYZ": {
    "DSP": {
      "PeakRMS": 2067,
      "RMSWindow": 10.5
    }
  }
}

B: Reference data

The reference data set is intended as an illustration of properly constructed miniSEED 3 and to be used by software implementors during development and testing.

The set contains multiple examples of data records illustrating different characteristics and data encodings of the from. Each example is available in miniSEED 3, JSON, and human-readable text representations.

The whole data set may be downloaded from documentation repository.

All time series in the reference set that contain series are the same expanding sinusoid signal.

#	Description	Download
1	Text payload	`mseed3` `JSON` `Text`
2	Event detection headers only, no data payload	`mseed3` `JSON` `Text`
3	Sinusoid series encoded as 16-bit integers	`mseed3` `JSON` `Text`
4	Sinusoid series encoded as 32-bit integers	`mseed3` `JSON` `Text`
5	Sinusoid series encoded as 32-bit IEEE float	`mseed3` `JSON` `Text`
6	Sinusoid series encoded as 64-bit IEEE float	`mseed3` `JSON` `Text`
7	Sinusoid series encoded as Steim-1 compressed integers	`mseed3` `JSON` `Text`
8	Sinusoid series encoded as Steim-2 compressed integers	`mseed3` `JSON` `Text`
9	Series with time quality, correction, event detections headers	`mseed3` `JSON` `Text`
10	Series with some FDSN and non-FDSN extra headers	`mseed3` `JSON` `Text`
11	Series with all FDSN extra headers (unrealistic)	`mseed3` `JSON` `Text`

C: Mapping from miniSEED 2.4

The following list of miniSEED 2.4 format structures specifies the mapping of all fields to this specification. In this listing the following abbreviations are used:

EH: Extra Header (of this specification), with path
SID: Source Identifier (of this specification)
FSDH: Fixed Section Data Header (in either specification)

miniSEED 2.4 Fixed Section Data Header (FSDH)

Field

Description

This specification

1

Sequence number

EH: FDSN.Sequence

2

Data header/quality indicator

EH: FDSN.DataQuality

A data center may choose to

translate miniSEED 2.4 quality

values to publication versions

with the following mapping:

2.4 quality

Pub. version

R

1

D

2

Q

3

M

4

3

Reserved byte

[no mapping]

4

Station identifier code

Incorporated into SID, FSDH field 13

5

Location identifier

Incorporated into SID, FSDH field 13

6

Channel identifier

Incorporated into SID, FSDH field 13

7

Network code

Incorporated into SID, FSDH field 13

8

Record start time

FSDH field 4

9

Number of samples

FSDH field 7

10

Sample rate factor

Incorporated into FSDH field 6

11

Sample rate multiplier

Incorporated into FSDH field 6

12

Activity flags:: 0 = calibration signals

1 = time correction applied

2 = begining of event

3 = end of event

4 = + leap second included

5 = - leap second included

6 = event in progress

FSDH and Extra headers:: FSDH field 3, bit 0

[no mapping]

FDSN.Event.Begin

FDSN.Event.End

FDSN.Time.LeapSecond

FDSN.Time.LeapSecond

FDSN.Event.InProgress

13

I/O flags, bits:: 0 = Sta. volume parity error

1 = Long record read

2 = Short record read

3 = Start of time series

4 = End of time series

5 = Clock locked

FSDH and Extra headers:: FDSN.Flags.StationVolumeParityError

FDSN.Flags.LongRecordRead

FDSN.Flags.ShortRecordRead

FDSN.Flags.StartOfTimeSeries

FDSN.Flags.EndOfTimeSeries

FSDH field 3, bit 2

14

Data quality flags, bits:: 0 = Amp. saturation detected

1 = Dig. clipping detected

2 = Spikes detected

3 = Glitches detected

4 = Missing/padded data

5 = Telemetry synch. error

6 = Digital filter charging

7 = Time tag questionable

FSDH and Extra headers:: FDSN.Flags.AmplifierSaturation

FDSN.Flags.DigitizerClipping

FDSN.Flags.Spikes

FDSN.Flags.Glitches

FDSN.Flags.MissingData

FDSN.Flags.TelemetrySyncError

FDSN.Flags.FilterCharging

FSDH field 3, bit 1

15

Number of blockettes that follow

[no mapping]

16

Time correction

EH: FDSN.Time.Correction

17

Beginning of data

[no mapping]

18

First blockette

[no mapping]

Blockette 100 (Sample Rate)

Field	Description	This specification
3	Actual sample rate	FSDH field 6
4	Flags (undefined)	[no mapping]
4	Reserved byte	[no mapping]

Blockette 200 (Generic Event Detection)

Field	Description	This specification
3	Signal amplitude	EH: FDSN.Event.Detection.SignalAmplitude
4	Signal period	EH: FDSN.Event.Detection.SignalPeriod
5	Background estimate	EH: FDSN.Event.Detection.BackgroundEstimate
6	Event detection flag bits: 0 = dil./comp. wave 1 = units	Extra headers: FDSN.Event.Detection.Wave FDSN.Event.Detection.Units
7	Reserved byte	[no mapping]
8	Signal onset time	EH: FDSN.Event.Detection.OnsetTime
9	Detector name	EH: FDSN.Event.Detection.Detector

Blockette 201 (Murdock Event Detection)

Field	Description	This specification
3	Signal amplitude	EH: FDSN.Event.Detection.SignalAmplitude
4	Signal period	EH: FDSN.Event.Detection.SignalPeriod
5	Background estimate	EH: FDSN.Event.Detection.BackgroundEstimate
6	Event detection flag bits: 0 = dil./comp. wave	Extra headers: FDSN.Event.Detection.Wave
7	Reserved byte	[no mapping]
8	Signal onset time	EH: FDSN.Event.Detection.OnsetTime
9	Signal-to-noise values	EH: FDSN.Event.Detection.MEDSNR (array)
10	Lookback value	EH: FDSN.Event.Detection.MEDLookback
11	Pick algorithm	EH: FDSN.Event.Detection.MEDPickAlgorithm
12	Detector name	EH: FDSN.Event.Detection.Detector

Blockette 300 (Step Calibration)

Field	Description	This specification
3	Start calib	EH: FDSN.Calibration.Sequence.Begintime
4	Num of calibs	EH: FDSN.Calibration.Sequence.Steps
5	Calibration flags: 0 = first pulse 1 = cal alt sign 2 = cal auto 3 = cal cont	Extra headers: FDSN.Calibration.Sequence.StepFirstPulsePositive FDSN.Calibration.Sequence.StepAlternateSign FDSN.Calibration.Sequence.Trigger FDSN.Calibration.Sequence.Continued
6	Duration of step	EH: FDSN.Calibration.Sequence.Duration
7	Time between steps	EH: FDSN.Calibration.Sequence.StepBetween
8	Amp. of cal. signal	EH: FDSN.Calibration.Sequence.Amplitude
9	Channel cal. signal	EH: FDSN.Calibration.Sequence.InputChannel
10	Reserved byte	[no mapping]
11	Reference amplitude	EH: FDSN.Calibration.Sequence.ReferenceAmplitude
12	Coupling of signal	EH: FDSN.Calibration.Sequence.Coupling
13	Rolloff of filters	EH: FDSN.Calibration.Sequence.Rolloff

Blockette 310 (Sine Calibration)

Field	Description	This specification
3	Start calib	EH: FDSN.Calibration.Sequence.Begintime
4	Reserved byte	[no mapping]
5	Calibration flags: 2 = cal auto 3 = cal cont 4 = PtoP amp 5 = ZtoP amp 6 = RMS amp	Extra headers: FDSN.Calibration.Sequence.Trigger FDSN.Calibration.Sequence.Continued FDSN.Calibration.Sequence.AmplitudeRange FDSN.Calibration.Sequence.AmplitudeRange FDSN.Calibration.Sequence.AmplitudeRange
6	Duration of cal	EH: FDSN.Calibration.Sequence.Duration
7	Period of signal	EH: FDSN.Calibration.Sequence.Period
8	Amp. of cal. signal	EH: FDSN.Calibration.Sequence.Amplitude
9	Channel cal. signal	EH: FDSN.Calibration.Sequence.InputChannel
10	Reserved byte	[no mapping]
11	Reference amplitude	EH: FDSN.Calibration.Sequence.ReferenceAmplitude
12	Coupling of signal	EH: FDSN.Calibration.Sequence.Coupling
13	Rolloff of filters	EH: FDSN.Calibration.Sequence.Rolloff

Blockette 320 (Pseudo-random Calibration)

Field	Description	This specification
3	Start calib	EH: FDSN.Calibration.Sequence.Begintime
10	Reserved byte	[no mapping]
5	Calibration flags: 2 = cal auto 3 = cal cont 4 = Random amps	Extra headers: FDSN.Calibration.Sequence.Trigger FDSN.Calibration.Sequence.Continued FDSN.Calibration.Sequence.AmplitudeRange
6	Duration of cal	EH: FDSN.Calibration.Sequence.Duration
7	PtoP amp. of steps	EH: FDSN.Calibration.Sequence.Amplitude
8	Channel cal. signal	EH: FDSN.Calibration.Sequence.InputChannel
9	Reserved byte	[no mapping]
10	Reference amplitude	EH: FDSN.Calibration.Sequence.ReferenceAmplitude
11	Coupling of signal	EH: FDSN.Calibration.Sequence.Coupling
12	Rolloff of filters	EH: FDSN.Calibration.Sequence.Rolloff
13	Noise type	EH: FDSN.Calibration.Sequence.Noise

Blockette 390 (Generic Calibration)

Field	Description	This specification
3	Start calib	EH: FDSN.Calibration.Sequence.Begintime
10	Reserved byte	[no mapping]
5	Calibration flags: 2 = cal auto 3 = cal cont	Extra headers: FDSN.Calibration.Sequence.Trigger FDSN.Calibration.Sequence.Continued
6	Duration of cal	EH: FDSN.Calibration.Sequence.Duration
7	Amplitude of signal	EH: FDSN.Calibration.Sequence.Amplitude
8	Channel cal. signal	EH: FDSN.Calibration.Sequence.InputChannel
9	Reserved byte	[no mapping]

Blockette 395 (Calibration Abort)

Field	Description	This specification
3	End calib	EH: FDSN.Calibration.Sequence.Endtime
10	Reserved byte	[no mapping]

Blockette 400 (Beam), Blockette 405 (Beam Delay)

No mapping for these blockettes

Blockette 500 (Timing)

Field	Description	This specification
3	VCO correction	EH: FDSN.Time.Exception.VCOCorrection
4	Time of exception	EH: FDSN.Time.Exception.Time
5	Microsecond offset	[included in record start time]
6	Reception quality	EH: FDSN.Time.Exception.ReceptionQuality
7	Exception count	EH: FDSN.Time.Exception.Count
8	Exception type	EH: FDSN.Time.Exception.Type
9	Clock model	EH: FDSN.Clock.Model
10	Clock status	EH: FDSN.Time.Exception.ClockStatus

Blockette 1000 (Data Only SEED)

Field	Description	This specification
3	Encoding format	FSDH field 5
4	Word order	[no mapping, no longer needed]
5	Data record length	[no mapping, no longer needed]
6	Reserved byte	[no mapping]

Blockette 1001 (Data Extension)

Field	Description	This specification
3	Timing quality	EH: FDSN.Time.Quality
4	Microsecond offset	Incorporated into FSDH field 4
5	Reserved byte	[no mapping]
6	Frame count	[no mapping]

Blockette 2000 (Variable Length Opaque)

No mapping for this blockette. The opaque data encoding may be used to specify an opaque payload for nearly equivalent functionality.

Unsupported miniSEED 2.4 content

The following defined information in miniSEED 2.4 cannot be represented in this specification:

Clock model specification per timing exception. Current specification only allows a single clock model specification per record.
Blockettes 400 (Beam) & 405 (Beam Delay)
Blockette 2000 (Opaque Data)

Software

A variety of open-source software packages are available to that support the miniSEED format.

Validation

The mseed3-utils project includes the following programs:

mseed3-validator - a validator for the miniSEED 3 formatted files
mseed2text - convert miniSEED 3 format to a text representation
mseed2json - convert miniSEED 3 format to a JSON representation

Download mseed3-utils releases

mseedconvert

The mseedconvert program converts miniSEED between versions and data encodings.

Download mseedconvert releases

libmseed

A general library (C language) for reading and writing miniSEED. Release versions 3.x support both this specification and version 2 of the format.

Download libmseed releases

Changes

Changes to this specification are listed below.

April 2023

Specification revision 2023-01-18 adopted by FDSN.
Reference data were refined following the specification revision.

2021

Initial specification of miniSEED 3