SEGD Revision 3'0

1
SEG-D Revision 3.0 Draft
SEG Technical Standards Committee Energist
ics Western Europe Region SIG Meeting 17 jun
2008
2
Why now?
The 2006 2.1 revision was specifically designed
as a targeted upgrade to support very high
capacity tapes. 3.0 was promised in the 2.1
standard to follow soon afterwards and deal with
major overhaul and/or upgrades.
3
Why us?
Jill Lewis (Troika International) Earns living
transferring transcribing data Stewart A. Levin
(Halliburton Energy Services) ProMAX SEG-D
Input support upgrades Rune Hagelund
(WesternGeco) Data acquisition and tape
management software for WesternGeco field
operations Barry D. Barrs (ExxonMobil)
Navigation and positioning expertise
4
SEG-D issues driving 3.0

• An explosion in data being stored in
  vendor-specific headers driven by the need for
  smooth transfer between acquisition
  QC/processing
• Multicomponent surveys
• Higher sampling rates/non-power-of-2 ratios
• Beyond 24-bit sensor sensitivity
• Continuous passive monitoring
• Only non-record format is long byte stream
  cumbersome for QC and transcription

5
SEG-D opportunities driving 3.0

• Nonseismic data?
• Positioning standards coordination
• Using web documents for dynamic information
• Manufacturer list
• Media types and parameters
• API organization codes
• Guidance recommendations and clarifications
• Errata and translations
• Format validation utilities

6
Overview of revision 3.0 proposal

• All features presented at SEG San Antonio/EAGE
  London 2007 have been implemented
• With a few additions (Source Auxiliary Channel
  Ref)
• SEGD Tape Label, General Header 1 and Demux Trace
  Header left as is (with a few exceptions)
• Be an efficient format for transferring
  information between acquisition and QC/processing
• SEGD is not a processing format
• More space for user-defined header blocks

7
Overview of revision 3.0 proposal

• Increased flexibility
• Header blocks tagged to allow shot specific
  (attached to general header) or trace specific
  information
• Block type written in byte 32 of all header
  blocks
• Flexible sample rates, header sizes, reclength
  etc.
• Increased robustness
• Tagged header blocks
• Size of shot/data in General Header
• Simplified encoding/decoding
• No complex datatypes, using 4byte float, int etc
• Explicit values (ex. samples/trace)?
• No dependencies between header blocks
• Additional functionality

8
New features in revision 3.0 proposal

• Positions can be tagged to all traces/equipment
• Multiple positions (in space and time)?
• Format TBD by pos group
• Support for multi-component data
• Node number, trace grouping, orientation header
• New General Trailer format
• Flexible, allows any data-block to be appended
• Standardized edits can be appended to General
  Trailer
• Allows easy addition of edits post-acquisition
• 8080 IEEE 8 byte samples

9
New features in revision 3.0 proposal

• All measurements now properly tagged with
  accurate, absolute timestamp (if available)?
• Positions, traces, shot, etc.
• Allows multiple measurements of same type
• All header sizes extended
• More room for acquisition system defined blocks
• Existing trace ext hdr blocks may be used....
• ...but remember to use correct block tag (byte 32)

10
New features in revision 3.0 proposal

• Support for complex shooting schemes
• Multiple source initiations per shot record
• Multiple sources firing simultaneously
• Support for complex source configurations
• Traces and measurements for sources and parts of
  sources
• Align SEG-D with SPS standard and/or SEG-Y
  revision 1 positioning support

11
New features in revision 3.0 proposal

• Standardizing storage of common survey
  information
• Vessel/crew identification
• Survey area name
• Client identification
• Job identification
• Line identification (Record set ID)?
• Size of record, data, and header explicit in
  General Header

12
New features in revision 3.0 proposal

• Table Of Contents file
• Stored at end (or beginning) of tape
• Lists all SEGD records on a tape
• Enable fast access to data
• May be stored on disk (to simplify data mgt)?

13
New features in revision 3.0 proposal

• Standardizing storage of SEGD
• One header block spanning multiple tape blocks
• A trace spanning multiple tape blocks
• Fixed/variable block devices
• Disk storage
• Transfer across network

14
Thank you for the attention

• Questions?
• More details regarding the proposal follows this
  slide

15
New features in revision 3.0 proposal

• Logical numbering of traces/source points
  extended to handle all operation types
• Line number
• Point number
• Point index
• Group index
• Depth index
• Reshoot index

16
Sample rate

• Sample rate steps of 1 microsecs
• No longer a base scan rate
• Dominant sample interval used for backwards
  compatibility
• Samplerate chosen over frequency
• Backwards compatibility
• Ability to relate sample to point in time
  (timestamp)

17
Timestamp

• Counting microsecs since 6 Jan 1980 (GPS epoch)?
• 8 byte integer
• Negative timestamps allowed
• 292471 years range
• Defined back to 1 jan 1970(SEG-D epoch)
• Leap seconds must be added to UTC
  timestamp(General Header)

18
New features in revision 3.0 proposal

• More sensor types/channel types explicitly
  supported
• Wind, depth, reference signals, source
  measurements etc.
• Sensor sensitivity value stored per channel
• Convert to actual physical unit
• Support non-voltage measurements

19
New features in revision 3.0 proposal

• Line number -gt Record set number
• To be usable for all types of operation
• Only 3 General Header blocks (1-3) required in
  rev 3.0
• Everything else optional
• Allows meta-data records (no traces)
• Allows zero length traces (only trace header)
• Allows trailer only records

20
Header size extensions

• Channel set header block 96 bytes
• Header block sizes
• General header - 65536 blocks (2 MB)?
• Channelset header 65535 blocks (6MB)
• Skew blocks 65535 blocks (2MB)?
• Extended header 16777215 blocks (512MB)?
• External header - 16777215 blocks (512MB)?
• Trace Header Extension 255 blocks (8160 bytes)?
• General Trailer 4294967295 blocks (128GB)?

21
Trace size

• Traces can be 2147 seconds long
• Negative start times allowed
• Extended recording mode allows up to 140,735,340
  seconds (1628 days) of data to be stored in one
  record

22
Filters

• Extended filter definitions
• Frequencies and slopes are IEEE floats (4 bytes)
• Filter type
• Filter delay

23
Positions

• Definition not completed yet (work in progress)
• Placeholders defined
• Projection/datum information part of General
  Header
• Only one projection allowed per SEG-D record
• Positions part of Trace Header or General Header
  (sources)
• SEG-D position will be binary

24
Missing from draft

• Update Introduction chapter
• List of Rev 3.0 changes not complete
• 4 figures in chapter 3 not updated (Record format
  and block layout on tapes)
• TOC file Record description requires proper
  definition
• Several lists needs completion (Channel types,
  Soil types, Sensor types, Physical units, etc.)
• Appendix A (Manufacturer list) and C (API
  Producer code) not updated
• Delete Appendix D (Header descriptors)?
• Appendix E (Examples) not yet completed
• Some questions/topics for discussion listed in
  red.
• All header blocks and Rev 3 concepts completed.

25
Questions, Concerns and Suggestions?

• Volunteers always wanted
• Were taking names
• Multicomponent standards
• Downhole acquisition support
• Metadata additions (temperature, wave height, )
• Opinions too ?
• Were taking notes