Amended 1995 November 3
Philosophy
The Advanced Missile Signature Center (AMSC) standard archive format (SAF) was created for flexible and extensible use in data archiving. Although the data may be ascii or one of several binary formats, the file header is plain ascii text and therefore human readable.
The "magic number" for identifying SAF files is the case-insensitive string: "HdSize ". Therefore, no naming conventions are required, and computer programs can easily identify the the file format.
Both image and non-image data may be stored in the AMSC standard archive format. The type of data is identified by information stored in the header.
Header Format
The standard header is ASCII text consisting of tags, or field identifiers, followed by data fields. The number of tags used in any particular header is variable dependent on the available purtenant information. Additional header information can be added when required with an associated increase in header size. Tags and associated data fields may be listed in any order, except for the HdSize tag which is listed first and identifies the file as a Standard Archive Format file.
Each tag consist of several characters, followed by a space. The tag names and values are case insensitive. Data display and analysis programs can search the header for standard tags and treat the data appropriately.
Header Item Advanced Missile Signature Center Requirements Line Termination LF or CR/LF. HdSize byte count must include the CR and LF characters. HdSize Exact number of header bytes, or the word "auto". If "auto" is used, then the last tag in the header must be "data". Tag Names All standard tags are six characters or less. User defined tags may be from one to 29 characters. Data Fields Variable length with ascii text. Leading and trailing spaces are ignored. Different parameters may have values which are text, integers, or floating point numbers. NumDPs Number of points per parameter in a POD file. If the value is "auto", the value will be determined from the size of the file for binary data, of from the number of line feeds for ascii data. PnSize Existence indicator of parameter names in POD files. A non-zero value indicates that the first line following the header consists of parameter names. Each name must be less than 64 characters long. Embedded spaces are permitted if the name is enclosed in quotes. This tag may be left out of the header if the value is zero. PuSize Existence indicator of parameter units in POD files. A non-zero value indicates that the first line following the (possibly) parameter names is a line of unit names. Each unit name must be less than 64 characters long. Embedded spaces are permitted if the name is enclosed in quotes. This tag may be left out of the header if the value is zero. PcSize Existence indicator of parameter classifications in POD files. A non-zero value indicates that the first line following the (possibly) parameter units is a line of individual unit classifications. Each classification must be less than 32 characters long. Embedded spaces are permitted if the classification is enclosed in quotes. This tag may be left out of the header if the value is zero.
Table I lists generic tags for including information common to all instrument types. Table II lists additional fields commonly used with image data. Table III gives tags for ordered pair (XY) type data, and Table IV adds tags for active sources. Several tags are mandatory, depending upon type of data in the file. Tags should be selected which provide the User with sufficient information to read, understand, and evaluate the data file.
File Names
Data archived at the AMSC will generally have the the ".saf" extension for image files and ".pod" extension for Parameter Oriented Data (POD) files.
Image Files
All image files are preceded by the standard header as described above. The header tag KeyWrd is used to identify the file as one of three types:
IMG -> the data consists of intensity values. CMAP -> the data consists of a color map and indices into the color map. PAV -> the data consists of ordered triplets: x, y, and value.
IMG Image Files
Rectangular image arrays have IMG as the KeyWrd value. Following the header is the two dimensional image data, consisting of an array of XPixls columns by Ypixls rows, stored in row major order (row one, followed by row two, followed by row three, etc.). If the file contains more than one field, the image is assumed to be interlaced. Following the image array is possibly a footer containing background information for Col or Row type background subtraction (Note 1). These data consist of an array of single precision floating point numbers, and this array can be present even if another background subtraction scheme is used (Note 2). The number of elements in this background array is XPixls for Col background subtraction and YPixls for Row background subtraction. The data units are specified by StdUnt (Note 21) for standard units, or by DaUnit for other values.
CMAP Image Files
Color mapped image files have CMAP as the KeyWrd value. Following the standard header are 768 bytes of the color map, with 256 bytes each for red, green, and blue, respectively. Following the color map are the image data. The value of each pixel in the image is a zero-referenced subscript to a color in the color map.
PAV Image Files
Position and Value image files have PAV as the KeyWrd value. The data consists of ordered triplets of position in x and y followed by the amplitude. The image reader code will interpolate the data to generate a rectangular array of the size given by the XPixls and YPixls tags. The image boundary points may be specified by using the BndXX tags to specify the clockwise sequence of boundary points. The origin is at the upper left hand corner of the image, with x increasing to the right and y increasing downward. If any boundary points are specified, then there must be at least three of them, in which case a triangular area would be generated.
XY Data Files
Ordered pair data (Radiance vs wavelength, pressure vs time, In-band radiance vs axial distance, etc.) are preceded by the standard header. The two column ordered array is then listed in an xy format. One or more blank spaces separate the values in each row. Parameter names are specified by XParam and YParam. Data units for x and y are specified by XDaUnt and DaUnit.
An alternate form of an xy data file can be used (if desired) when the x values are equally spaced in time, wavelength, position, etc. For this form, only the y values are stored, with the x values calculated using additional parameter tags (XYFRST, XYLAST, NumDPs).
Active Source Files
Active source tags are added to headers for images as appropriate to describe both the active source and detector used in acquiring data.
Parameter Oriented Data (POD) File Format
An additional archive file format is defined for multiple parameters to be stored in the same file, such as time history data. Examples of such data are time, target position, velocity, and aspect angle, or time, motor chamber pressure, turbine speed, and case temperature, etc.
The POD file format has a standard header, an optional line of parameter names, an optional line of parameter units, an optional line of parameter classifications, followed by the data. The data may be arranged in columns (handy for human viewing), or in rows (quicker computer access). Five tags are also required to define the number of parameters in the file and the space required for parameter names, units, and security classifications. Additional information on the POD data file format is included below.
A. Header Information - The header will consist of standard SAF tags. The tag KeyWrd with value POD must be included. The tag NParam will indicate the number of parameters while the tag NumDPs will indicate the number of data points (values) per parameter. B. Parameter Names - A single line of text may be used for the parameter names of the data. Spaces, tabs, commas, colons, semicolons, and vertical bars may be used as delimiters between parameter names. All parameters will have names not exceeding 63 characters. Include the PnSize tag with a positive value in the header to specify the existence of the parameter names. Any parameter name which includes a delimiter must be enclosed in quotes. C. Parameter Units - A single line of text may be used for the parameter units of the data. Spaces, tabs, commas, colons, semicolons, and vertical bars may be used as delimiters between parameter units. All parameters will have units not exceeding 63 characters. Include the PuSize tag with a positive value in the header to specify the existence of the parameter units. Any parameter unit which includes a delimiter must be enclosed in quotes. Use a pair of double quotes ("") as a space holder for a parameter which has no units. D. Parameter Classification - A single line of text may be used for the security classification of each parameter. The Class tag in the header should reflect the classification of the entire file. If this is the same as for each parameter, only the Class tag need be used. Otherwise, include a classification for each column and include the PcSize tag with a positive value in the header. The maximum length of a classification string is 31 characters. Any classification which includes a delimiter must be enclosed in quotes. E. Parameter Values - The remainder of the pod file contains the parameter values. The data type of these values is defined by the header tag DaType. All values will have the same data type. If DaType is ASCII, then a space, comma, tab, colon, semicolon, or vertical bar may be used to separate parameter values. Non-numeric data may only be stored in ASCII files. Quotes may be used to group sequences of characters. Example: HdSize Auto Class Unclassified DaType ASCII Keywrd POD PcSize 0 PuSize 1 PnSize 1 Nparam 6 NumDPs 5 Data TIME ALTITUDE VELOCITY "ASPECT ANGLE" Filter Camera sec. meters meters/sec degrees "" "" 0.0 0.0 0.0 90. 1 "NIKA 2" 1.0 10.0 1.0 89. 1 "NIKA 2" 2.0 20.0 2.0 88. 1 "NIKA 2" 3.0 30.0 3.0 87. 2 FTS 4.0 40.0 4.0 86. 2 FTSTable I Generic Tags
Tag Description Type HdSize Header Size (in bytes, or it may be AUTO) Integer AqMode Acquisition Mode (Note 10) Text AspAng Aspect Angle (degrees) Float BgFile Background File (Notes 2,6) Text BgType Background Type (Note 1) Text BgValu Background Value (Average or Fixed) (Note 2) Float BPFile Bad Pixel File (Note 6) Text BytOrd Byte Order (Note 3) Text CaFile Calibration File (Notes 2,6) Text ChTemp High Temp of Cal Source (degrees C) Float Class Classification Text CIDay Collection IRIG Day Integer CIHour Collection IRIG Hour Integer CIMin Collection IRIG Minute Integer CISec Collection IRIG Seconds Float ClTemp Low Temp of Cal Source (degrees C) Float COMENT Comment Line (Repeated as Required) Text CSFile Calibration Source File (Note 6) Text Data End of header DaType Data Type (Note 4) Text DaUnit Data Units (Note 2) Text DDOff Data Distribution Office Text DiaFOV Circular FOV (degrees) Float DiStat Distribution Statement Text ElAng Elevation Angle (deg) Float EURAW Processed Level of the Data EU/RAW/Flat-fielded (FF) Text ExpID AMSC Experiment ID# Text Filter Filter Name Text Filtno Filter Number Integer FOVAxl Chamber Test FOV Axial Location (meters) (Note 7) Float FOVRdl Chamber Test FOV Radial Location (meters) (Note 8) Float HdVers Header Version (= 2.0) Text HorFOV Horizontal FOV (degrees) Float IHFOV Instantaneous Horizontal FOV (microradians) Float Itime Integration Time (seconds) Float IVFOV Instantaneous Vertical FOV (microradians) Float LinLog Linear/Log Indicator (LIN/LOG/ASG) (Notes 2,5,16) Text LODAng Observer Line-of-Sight Angle wrt True North (CW is +) Float LogASl Log Amp Slope (Notes 5,16) Float LogOff Log Offset (Notes 5,16) Float Mach Mach Number Float Mdate Mission Date (MM/DD/YY) Text MeasUn Measurement Uncertainty-Nominal (%) Float Miss Mission Text NCoads Number of Coads Integer NEQ Noise Equivalent Quantity (Note 9) Float NodeNo Node Number Text Note01 Numbered Notes, valid from Note01 through Note99 Text OffCor Offset Correction (Note 2) Float PPCNam Platform AMSC Component Name Text RDFile Raw Data File (Note 6) Text RolAng Roll Angle (degrees) Float SBPLo System Bandpass, Lower Wavelength (microns) Float SBPUp System Bandpass, Upper Wavelength (microns) Float SclFac Calibration or Scale Factor (Note 2) Float SDLevl SDIO Data Level (Note 15) Text SecCol Number of Seconds Collected Integer SLFile Spectral Lamp File Text SltRng Slant Range (meters) Float SnsAlt Sensor Altitude (meters) Float SPCNam Sensor AMSC Component Name Text Stage Launch Vehicle Stage Number Float StdUnt Standard Data Units Index (Note 21) Integer SUncLo Independent Parameter Uncertainty Lower Limit Float SUncUp Independent Parameter Uncertainty Upper Limit Float TALO TALO (seconds) Float TAOA Target Angle of Attack (degrees) Float Target Target Name Text TestNo Test Number Text TPCNam Target PC Component Name Text TPFact Transmission Path Factor (Note 2) Float TrgAlt Target Altitude (meters) Float TrgHdg Target Heading with respect to True North (CW is pos) Float TrgTyp Target Type (Liquid/Solid) Text TrgVel Target Velocity (meters/second) Float TrlNum Trial Number Text TZDay T-Zero Day Integer TZHour T-Zero Hour Integer TZMin T-Zero Minute Integer TZSec T-Zero Seconds Float USRCON User-Specified Parameters Header Filename (Note 17) Text VrtFOV Vertical FOV (degrees) Float Warn01 Warning to the user, valid from 01 thru 99 Text XUncUn Independent Parameter Uncertainty Units Float YUncLo Dependent Parameter Uncertainty Lower Limit Float YUncLo Dependent Parameter Uncertainty Upper Limit Float YUncUn Dependent Parameter Uncertainty Units FloatTable II Imager Tags
Tag Description Type ACFile Area Correction Factor File Text ADJFAC Non-Linear Adjustment Factor Float ApSize Aperture Size Float BGBLLX Lower Left X for Background Box (Note 2) Integer BGBLLY Lower Left Y for Background Box (Note 2) Integer BGBLRX Lower Right X for Background Box (Note 2) Integer BGBLRY Lower Right Y for Background Box (Note 2) Integer BGBULX Upper Left X for Background Box (Note 2) Integer BGBULY Upper Left Y for Background Box (Note 2) Integer BGBURX Upper Right X for Background Box (Note 2) Integer BGBURY Upper Right Y for Background Box (Note 2) Integer Bnd01 Boundary Point for PAV files, valid from 01 thru 99 Integer Bx1Int Box 1 Integral Value Float Bx1LLX Box 1 Lower Left X Integer Bx1LLY Box 1 Lower Left Y Integer Bx1LRX Box 1 Lower Right X Integer Bx1LRY Box 1 Lower Right Y Integer Bx1ULX Box 1 Upper Left X Integer Bx1ULY Box 1 Upper Left Y Integer Bx1URX Box 1 Upper Right X Integer Bx1URY Box 1 Upper Right Y Integer Bx2Int Box 2 Integral Value Float Bx2LLY Box 2 Lower Left X Integer Bx2LLY Box 2 Lower Left Y Integer Bx2LRX Box 2 Lower Right X Integer Bx2ULX Box 2 Upper Left X Integer Bx2ULY Box 2 Upper Left Y Integer Bx2URX Box 2 Upper Right X Integer Bx2URY Box 2 Upper Right Y Integer CentMX Center Cursor X Position Integer CentMY Center Cursor Y Position Integer CGain Camera Gain Float Colr1 Color 1 Value Text Colr2 Color 2 Value Text Colr3 Color 3 Value Text Colr4 Color 4 Value Text Colr5 Color 5 Value Text Colr6 Color 6 Value Text Colr7 Color 7 Value Text Colr8 Color 8 Value Text Colr9 Color 9 Value Text Colr10 Color 10 Value Text Colr11 Color 11 Value Text Colr12 Color 12 Value Text Colr13 Color 13 Value Text Colr14 Color 14 Value Text Colr15 Color 15 Value Text Colr16 Color 16 Value Text ComPrs Compression Type (Note 25) Text DGFld Digitizer Gamma setting (non-linear adj.) (Note 14) Float DiType Digitizer Type (Note 14) Text DPtNum Data Point Number Integer DSGain Digitizer System Gain Float DSOff Digitizer System Offset Float FldFrm Fields per Frame Integer FlorFr Field or Frame Data in this File Text FRate Frame Rate (frames/second) Float FrstCl First Color Integer EMFile Emissivity File Text IDFile Image Display Control File (Note 6) Text ImDisp Pre-digitized image display format (Note 14) Text ImQual Image Quality (Note 11) Text ImSig Pre-digitized image signal type (Note 14) Text ImSize Image Size (in bytes after any compression) Integer Intrlc Frame Data Are Interlaced in the File Text NClrs Number of Colors Used Integer PLeftX Profile Left X Integer PLeftY Profile Left Y Integer PRghtX Profile Right X Integer PRghtY Profile Right Y Integer ProCX Profile Cursor X Integer ProCY Profile Cursor Y Integer SpecFn Special Function to apply to the image (Note 23) Text SRFile Slant Range Correction Factor File Text VrtAtt Vertical Attitude (degrees) Float XMag X Compress or Enlarge Factor Float XPixls Digitized Data Image Width (pixels) Integer XPxWid X (horizontal) Pixel Width (meters) Float YMag Y Compress or Enlarge Factor Float YMax Maximum y value Float YMin Minimum value Float YPixls Digitized Data Image Height (pixels) Integer YPxWid Y (vertical) Pixel Width (meters) FloatTable III Data Tags
Tag Description Type Calc01 Calculated Parameter, valid from 01 thru 99 (Note 24) Text FreRsp System Electrical Frequency Response (Hz) Float Keywrd XY File Keyword (Note 12) Text NParam Number of Parameters in File Integer NumDPs Number of Data Points in File Integer PcSize Parameter Classification Size (bytes) Integer PltSub XY Plot Subtitle Text PltTtl XY Plot Title Text PnSize Parameter Name Size (bytes) Integer PodOrd Data Ordering in POD Files (Note 20) Text PuSize Parameter Units Size (bytes) Integer SampRa Sample Rate (Hz) Float XCFile X-Axis Calibration File Text XDaUnt X-Axis Data Units Text XFName X-Axis File Name (a 1 parameter POD file) Text XParam X-Axis Parameter name Text XScFac X-Axis Scale Factor (applied to XFName values) Float XYFNum XY File Number Integer XYFrst XY First Value Float XYLast XY Last Value Float YParam Y-Axis Parameter name TextTable IV Active Source Tags
Tag Description Type SrcNam Active Source AMSC Component Name Text SrcWav Active Source Wavelength or Frequency (Note 13) Float SrcWid Active Source Scan Width (GHz) Float SrcWUn Source Wavelength or Frequency Units (Note 13) Text SrcRat Active Source Scan Rate (Hz) Float
NOTES
1. Background Type Code (BgType)
Col Column Average (each column has an average background value) Row Row Average (each row has an average background value) Avg Floating Box Average File File Subtraction Fix Fixed Value None Assume Zero Background2. Linear Data EUD Calculation This information was originally provided for image data sets, but can be adapted as required for other data set types. Instrument specific variations in this equation should be documented with the data set in an ASCII file.
The linear data mode is indicated by LIN in the LinLog field. The equation for converting the pixel values to engineering units is
EUD = (PixelValue - Background)*SclFac*TPFact+OffCorThe units of the resulting value are defined in DaUnit. The individual pixel values and the background value(s) are required to have the same units (i.e., counts of volts). SclFac is responsivity factor for the instrument which obtained the data, and is defined in radiance per detector output units (i.e., W/sr- cm2-count or W/sr-cm2-volt). Scale Factors (SclFac) which vary on a pixel by pixel basis are stored in the file identified in Cafile. If neither CaFile or SclFac are specified, SclFac defaults to 1.0 for all pixels.
TPFact is a dimensionless correction factor to adjust for transmission path effects, and defaults to 1.0 if not present. OffCor is an optional data correction offset factor (with a default of 0.0) and has the units defined in DaUnit. If the background type (BgType) is set to None or not present, the background defaults to 0.
For Avg and Fix backgrounds, the background value, which is the same for all pixels, is defined in BgValu. In addition, for the Avg type of background, the corners of the box used to compute the background are given in (BGBULX, BGBULY), (BGBURX, BGBURY), (BGBLLX, BGBLLY), and (BGBLRX, BGBLRY).
If the background type is Row, the footer contains an array of floating point number of dimension Ypixls. Then for all pixels in row 1 of the image, the first element of the footer array is used as the background; for all pixels in row 2 of the image, the second element of the footer array is used as the background, etc.
If the background type is Col, the footer contains an array of floating point number of dimension XPixls. Then, for all pixels in column 1 of the image, the first element of the footer array is used as the background; for all pixels in volume 2 of the image, the second element of the footer array is used as the background, etc.
If the background type is File, the BgFile field must contain a name for a background image file of the same format as all standard images. For file background subtraction, the background value for any pixel in the image file is the corresponding pixel in the background image.
LH Low order byte first (Typical DOS application) HL High order byte first VX VAX byte order4. Data Type Codes (DaType)
ASCII ASCII Int8 1-byte integer (0-255 unsigned) Int16 2-byte integer Int32 4-byte integer Int64 8-byte integer Flt32 Single precision floating point Flt64 Double precision floating point RGB24 24-bit color image (3 bytes per pixel, RGB)5. Log Data EUD Calculation (LogASl, LogOff) This information was originally provided for image data sets, but can be adapted as required for other data set types. Instrument specific variations in this equation should be documented with the data set in an ASCII file.
This mode of operation is indicated by the presence of LOG in the LinLog field. The equation for converting log camera data to engineering units is
EUD = SclFac*TPFact)*(10^(LogASl*(PixelValue-LogOff)) - 10^(LogASl*(Background-LogOff)))where SclFac, TPFact, OffCor, and DaUnit are defined the same as for linear data and LogOff is a tape recorder offset adjustment which defaults to 0. If the background type is set to None or not present, the second term in the equation (enclosed in braces []) defaults to 0.
6. Ancillary Files Any files named in the standard header are themselves required to have a standard header.
7. Captive test FOV Axial Location (FOVAxl) This parameter is defined as the distance from the nozzle exit to the center of the FOV (center pixel in the digitized image) parallel to the direction of flow and normal to the nozzle exit plane, with positive values in the direction of flow. The axial location is typically specified only for near-field test chamber or static test measurements.
For target aspect angles of 0 deg (nose-on) or 180 deg (tail- on), this parameter represents the horizontal distance to the target center.
8. Captive Test FOV Radial Location (FOVRdl) This parameter is defined as the distance from the plume centerline to the center of the FOV (center pixel in the digitized image) in a line normal to the direction of flow and parallel with the nozzle exit plane, with positive values being up and to the right in the digitized image. The radial location is typically specified only for near-field test chamber or static test measurements. For target aspect angles of 0 deg (nose-on) or 180 deg (tail-on), this parameter represents the horizontal distance to the target center.
9. Noise Equivalent Quantity (NEQ) This parameter has the units defined in DaUnits. The intent is to identify the rms noise level in the data. Nominal values should be specified and separate files archived as required for spatially or temporally dependent noise equivalent parameters.
10. Acquisition Mode Codes (AqMode)
FM FM Tape LV Live (Real time digitization) OP Video Optical Disk UM Umatic Tape VR VCR SV Super VHS D2 D2 Digital Video Tape11. Quality Codes (ImQual)
OK Preferred Image, scan or data point NO Non-preferred Image, scan or data point12. SAF Keyword Codes (KeyWrd)
Data Format KeyWrd Description Image IMG 2-D array of values (This is the default tag value) CMAP 2-D array of color map indices PAV Position and value triplets Parameter Oriented Data POD Multiple dependent parameters with the same independent parameter xy pairs XYPT (x,y) pairs, y vs. points XYFN (x,y) pairs, y vs. file number XYTM (x,y) pairs, vs. time XYDI (x,y) pairs, y vs. distance (in meters) y values only YPT y vs. points YFN y vs. file number YTM y vs. time YDI y vs. distance (in meters) YWL y vs. wavelength YWN y vs. wavenumber13. Active Source Wavelength and Units (SrcWav, SrcWUn) Not a required parameter for lamps or blackbodies. Standard units are micron, GHz or cm-1.
Digitizer system settings are required to maintain traceability and repeatability. Pre-digitized image information to be listed includes the display type and the signal type. The Gamma setting is required for Quantex-digitized imagery. Options are listed in the table below.
PARAMETER DIGITIZER IMAGE DISPLAY TYPE SIGNAL TYPE Tag DiType ImSig ImDisp Possible Quantex RGB Color Values Matrox NTSC RCI_RTD_8* PAL RCI_RTD_10* SECAM YUV_RTD_8** *RCI Real-Time Disk, 8-bit or 10-bit Resolution **RCI YUV File Format, 8-bit Resolution15. SDIO Data Levels (SDLevl)
Level Description 0 Telemetry data as received by ground stations. Both digital and analog. 1 Data processed by ground station into digitized, decommutated data, in proper time order, separated by sensor, time-tagged, with collateral data attached. Data is placed into a raw database. 2 Data reduced to scientific and engineering units by application of calibration data. Time-tag converted to GMT. Data is assembled into a reduced database. 3 Fully characterized, validated level 2 data. Characterization includes target characteristics, range, fuel type, etc, atmospheric conditions; as well as a complete sensor characterization.16. ASG IRIS DATA EUD CALCULATION (from Flat-Fielded Data)
EU = Exp[{ln(x-bg)-a} / b] a = LogASl in header b = LogOff in header bg = Background Value x = Pixel Value17. User Controlled Tags (USRCON)
Certain experiments may have specific and exceptional information to be included in the header. For such programs, a second header file is created in the same format as the standard file which includes tags and associated information. This file is named as specified by the parameter USRCON in the standard header. It is the responsibility of the archivist to provide sufficient documentation describing the user-specified tags and definitions.
This tag is used to indicate which stage is thrusting during powered flight. For most data sets and targets, the use of this tag will be straight forward. Nomenclature or conventions are not specified by this document for instances when the target coasts between stages, there are combination burns of different stages (e.g. Titan 34D), or when prominent interstage events occur. This tag is used by some AMSC Software tools to graphically indicate the approximate operating times or regions for the various stages.
19. Target Type Definitions (TRGTYP)
Liquid Liquid Rocket Engine Solid Solid Rocket Motor Liquid/Solid Dual System using a combination of both types of propulsion systems20. Data Ordering in POD Files (PodOrd)
This tag is used to indicate the storage order in POD files. The default value is COL, which is most suitable for ASCII files. In this case the data for each parameter is in a column, with each row having one value for each parameter. For binary files, the alternate value Row is more suitable. In this case all the data for each parameter occupies one row in the file. For binary files this means that all the data for each parameter is contiguous, allowing faster access.
21. Standard Data Units (StdUnt)
Standard data units are specified using this tag. For human readability of the header information, the data units may be placed in the header using the DaUnit tag also. The StdUnt tag will take precedence over the DaUnit tag.
Index Units Standard Abbreviation 0 Nonstandard, use DaUnit 1 Counts cnt 2 Volts V 3 Amperes A 4 Watts W 5 Fahrenheit Temperature deg F 6 Celsius Temperature deg C 7 Rankine Temperature deg R 8 Kelvin Temperature K 9 Meters m 10 Centimeters cm 11 Kilometers km 12 Micrometers um 13 Seconds sec 14 Steradians sr 15 In-Band Radiant Intensity W/sr 16 In-Band Irradiance W/cm^2 17 In-Band Radiance W/(sr cm^2) 18 Spectral Radiant Intensity W/(sr um) 19 Spectral Irradiance W/(cm^2 um) 20 Spectral Radiance W/(sr cm^2 um) 21 Station Radiance W/(sr cm)22. AMSC Default SAF Parameters Default values for common header parameters utilized by various AMSC analysis tools are listed below. These are supplied by the software if parameters are not included in the SAF file.
TAG VALUE BgType NONE BgValu 0.0 Class Unclassified EURaw RAW FldFrm 2 FlorFr FRAME Frate 30 Intrlc NO Keywrd IMG LinLog LIN OffCor 0.0 PodOrd Column SclFac 1.0 TPFact 1.0 ComPrs None23. Special Function Applied to Images
The text value of the SpecFn tag is an algebraic expression which is to be applied to the image after all other processing is completed. Header values are referenced as $tag where 'tag' is the name of the desired header tag. Each pixel value is referenced as x. Thus a sample expression is
"1.01 + x + sin(x*pi) / ($TALO/2)^3".
The analysis software will parse the text value of each CalcXX tag for the name, units, and the algebraic expression which defines this parameter. The parameter values will be calculated. Header values may be referenced by $tag where 'tag' is the name of the desired header value. The value of other parameters included in the file are referenced by placing the case sensitive name in single quotes. There may be up to 99 such parameters.
A sample text value is
"Mach Number", "", "'speed' * 1024.2 / $Temp^2"25. Compression Type
To save space, image data may be compressed. The only compression presently supported is gzip. The possible values are None and GZIP.