ISIS RELEASE NOTES
For November 15, 1997 Release

This document describes the changes and enhancements that have been made to ISIS since the previous release on February 14, 1997. This is divided into several sections:

1. Notes on Use of the IDL Display Programs
2. Documentation Updates
3. New Programs
4. Program-Specific Updates
5. Updates Applicable to Multiple Programs
6. Data File Updates
7. Changes of Interest to ISIS Programmers
8. Changes of Interest to ISIS Installation Managers

NOTES ON USE OF THE IDL DISPLAY PROGRAMS

8-BIT DISPLAY MODE
The ISIS IDL display programs make use of 8-bit display mode. Thus, they should be run with the 8-bit PseudoColor visual rather than the 24-bit DirectColor or 24-bit TrueColor visuals. If you are running on an 8-bit X terminal or 8-bit workstation display, then you will automatically be running with the 8-bit PseudoColor visual. However, if you are running on a 24-bit workstation display, then you must tell IDL that you want to run with the 8-bit PseudoColor visual. This can be accomplished by setting the IDL_STARTUP environment variable to the name of a file that contains IDL commands that are to be executed each time IDL is started. For example, to define "idl_startup" as your startup file:

    setenv IDL_STARTUP $HOME/idl_startup        # For C/TCSH shell
        - or -
    IDL_STARTUP=$HOME/idl_startup               # For Bourne/Korn shell
    export IDL_STARTUP

The "idl_startup" file in your home directory should then contain the command to select the 8-bit PsuedoColor visual:

    device, pseudo=8
       

RUNNING ISIS DISPLAY PROGRAMS WITH IDL 5.0.2 and IDL 4.0.1b
This release of ISIS contains a set of IDL display programs that is compatible with IDL 5.0.2, which is the latest release of IDL available from Research Systems. This release also contains a set of IDL display programs that is compatible with IDL 4.0.1b. These two sets of programs are functionally identical in this ISIS release. However, note that the 4.0.1b version of the programs will not receive any further development. All future development will be done with the latest available version of IDL. The default ISIS setup will run the 5.0.2 version of the programs. If you are still running IDL 4.0.1b, then you will need to select the 4.0.1b version of the display programs in your .chsrc or .profile files. See the ISIS Installation Guide for details.

The standard ISIS setup defines two aliases that can be used to switch which version of the IDL display programs your are running. Type the following at the operating system prompt to select the desired version:

    setisisidl4         # To select IDL 4.0.1b version
        - or -
    setisisidl5         # To select IDL 5.0.2 version

NEW "ERROR" MESSAGES WITH IDL 5.0.2
You will see two new "error" messages when running with IDL 5.0.2. Neither of these is anything to be concerned about.

The first is "XMANAGER: Error handling is now off." In 5.0.2, IDL's idea of error "handling" is to ignore the error by printing a brief message and attempting to continue running. Thus, the ISIS display programs turn off this error "handling" so that it will be easier to determine what went wrong.

The second message is "XMANAGER: The MODAL keyword to the XMANAGER procedure is obsolete. It is superseded by the MODAL keyword to the WIDGET_BASE function." This results from a change in the design of how IDL widget programs should work. The ISIS display programs will be upgraded to conform to this new IDL design in a future ISIS release. For now, the old design generates the message but will still work OK.

FILE SELECTION WITH IDL 5.0.2
IDL 5.0.2 does not allow file selection with a single click of the mouse button. You must either click on the "OK" button on the file selection widget or double click on the desired file.

WIDGET BACKGROUND COLOR WITH IDL 5.0.2
When running with IDL Version 5.0.2, some of you may find that the background color on widgets displayed by IDL programs has changed (probably to white). If you prefer to have the "official" IDL blue for the backgrounds, try putting the following lines in your .Xdefaults file:

    Idl*background: #7300a100ff00
    Idl*foreground: Black

    alias envi_setup 'source /usgs/idl5.0.2/envi_2.7/bin/envi_setup'

Then, when you want to run ENVI, you should first type envi_setup at the operating system prompt to do the setup and then type envi to start up ENVI.

Note that "sourcing" the envi_setup file will probably reset the IDL_PATH environment variable, which tells IDL where to look for program and subroutine files. Thus, this will make the ISIS IDL display programs unavailable. If you then want to run ISIS IDL programs, you can do so in another terminal window. Or, you could source your .cshrc file to reset IDL_PATH.

2. DOCUMENTATION UPDATES

LOCATION OF DOCUMENTATION
The $ISISDOC directory contains Release Notes for each release of ISIS. These are contained in files named "isisYYMMDD_release.notes", where YYMMDD is the date of the ISIS release. Each Release Notes file describes changes in the system since the previous release.

The current documentation for a release of ISIS is contained in a set of files in the $ISISHTML directory. These are in a format that can be accessed with a web browser. For example, to view the ISIS documentation with netscape, type:

    netscape $ISISHTML/documentation.html
 

PROGRAM NAMES
File names and program names are case-sensitive in Unix and TAE. All ISIS program names are lower case. In some of the documentation, you might find program names spelled with upper case characters. However, please remember that ISIS program names should always be typed as lower case. (IDL is not case-sensitive, so the names of the IDL programs can be typed as either upper case or lower case.)

PROGRAM DOCUMENTATION UPDATES
In addition to new documentation for the new programs and minor updates to documentation for previously existing programs, the following are some additional program documentation updates:

automos.pdf, mospdf.pdf - The documentation for "automos.pdf" and "mospdf.pdf" was cleaned up and level1 and level2 was added to "mospdf.pdf".

base2l1 - Has had an example added to the .pdf showing how to use the program.

boxfilter - The help information in the .pdf has been updated extensively to describe the new features of the program. Also, note that the parameters have been changed, so you might need to modify procedure .pdf files that call "boxfilter".

getfname - The help information in the .pdf has been updated to clarify the operation of the program.

qview - The help information in the .pdf has been extensively revised to describe all the new and old features of the program.

ratio - The pdf has been revised to clarify the various available options in the operation of the program.

OTHER WEB-BASED DOCUMENTATION UPDATES
The ISIS Overview has been updated.

The Introduction to ISIS has been slightly updated.

The Overview of ISIS Architecture has been slightly updated. (This was previously called the ISIS User's Guide.)

The Alphabetical and Menu Category program lists have been updated. In the Menu Category lists, there are now separate categories for the different Mission Specific groups. All of the IDL display programs now have documentation when you click on the program name in the lists. The "tvband", "tvband2", "bandif" and "band2if" IDL display programs are now obsolete and have been removed from the lists. (Their functions are now performed better by the non-IDL "qview" program.) However, these programs are still contained in the distributed system.

The Display Program Summary has been updated.

The description of Viking Image Processing has been added.

The Release Notes for this release (this document) has been added.

The Frequently Asked Questions (FAQ) list has been updated.

The Requested Changes/Features list has been updated.

The Known Bugs list has been updated.

The Distribution and Support Policy has been updated.

The Operation Environment Guide has been updated.

The Installation Guide has been updated.

The Current Activities and Software Updates pages have been updated (These are available only on the ISIS web site.)

PRODUCING A HARD-COPY PROGRAM REFERENCE MANUAL
Note that the "isisdoc" program is intended to allow creation of a file that can be printed to produce a hard-copy reference manual of program descriptions. This will contain the same information that is displayed in the on-line help and also the program descriptions that can be viewed with a web browser as described above. To make a program reference manual file, first make a list of all the ISIS ".pdf" files:

  ls -1 $ISISEXE/*.pdf > isis.lis

Next, run "isisdoc" with PDFLIST=isis and DOCFILE=prog_descr.doc to write the reference manual to the file "prog_descr.doc". Also, SUMFILE=isis and REFFILE=isis will produce files containing a one-line summary of programs and a quick reference for the programs.

Note that most of the ISIS application programs are contained in $ISISEXE, but $PICEXE contains additional programs that operate on PICS format files.

DOCUMENTATION FOR IDL DISPLAY PROGRAMS
The IDL display programs have on-line help that is available by clicking on the "help" buttons. The same documentation is available in the web-based program descriptions. However, the hard-copy manual of programs produced above will not include descriptions of the IDL display programs since they do not have ".pdf" files. The documentation for the IDL display programs is contained in text files in $ISISEXE/idl that have the extension ".hlp". (The ".hlp" files for the IDL 4.0.1b version of the programs are in $ISISEXE/idl4 and are identical to the ".hlp" files in $ISISEXE/idl for this release of ISIS.) Note that files of the form cv_xxx.hlp are the descriptions of the various individual control panels used by "cv". The file "cv.hlp" is a collection of all of these and thus is a complete description of the "cv" program.

3. NEW PROGRAMS

bandstats - This is a new ISIS application that computes statistics on individual bands as well as the entire cube object.

base - A new procedure that runs "base2l1" and "geom" to project a geometrically controlled base mosaic to the same camera view orientation as a LEVEL1 image.

clem0 - A new procedure that takes a list of raw Clementine files and processes them through the level0 processing sequence. This includes "clem2isis" and "naiflab".

cosi - New program that divides the core of a cube by the cosine of the incidence angle (obtained from the INCIDENCE_ANGLE backplane).

getcount - This new program will count the number of lines in a file and return the line count to a TAE procedure.

getrange - A new program that should only be run from a procedure. Given a list of input image filenames for a high resolution quad on Mars or Venus, "getrange" will return the latitude and longitude range of the outer boundary of the images. If no list is entered, the quadrangle latitude and longitude range will be returned. Program "hidim" sets up the call to "getrange" prior to moving into level 2 or higher calls.

hicut - A new procedure that is the final procedure to be run in the Viking high resolution mosaic sequence. The output is a mosaic that shows the actual area covered by each image in a Mars 1:500k quadrangle mosaic.

hidim - HIDIM stands for HIgh resolution Digital Image Mosaic. This is a new program that will create a procedure ".pdf" file that can be executed to perform one of several procedures in the HIDIM project.

hilev3a - New procedure that applies "boxfilter" to individual LEVEL2 images.

hilev3b - A new procedure that creates two mosaics using the output from "hilev3a" mos51.cub and mos251.cub.

hilev4 - A new procedure that adds the HPF mosaic and the HPF mosaic for the Mars 1:500k mapping project. The input to this procedure comes from the output of "hilev3b". The output is the final tone matched mosaic mos4.cub.

imp2prj - This is a new program that creates a TFILE to be used by "geom" for producing a geometric transformation of an Imager for Mars Pathfinder (IMP) image.

isis2std - This new program allows the user to display either a single selected band of an ISIS cube file as a gray-scale image or three selected bands as a Red/Green/Blue color composite image. The stretch of the single displayed band or the stretches of each of the R/G/B color components can then be interactively adjusted. The resulting displayed image can then be written out in a standard image file format. The supported output formats include TIFF (lossless) and JPEG (a lossy compression format with userspecified quality). This is an IDL program that is started by typing isis2std at the IDL prompt.

impcal - This new program will radiometrically calibrate Mars Pathfinder IMP data.

implab - this is a new program for limited fixes on the labels of IMP images. The program writes a new MFX_LSC_OFFSET_VECTOR = (0.209732,-0.041858, 0.300000) to the image labels to correct a previously calculated offset. The program also converts the LANDER_ SURFACE_ QUATERNION to the #3 quaternion. LANDER_SURFACE_QUATERNION = (0.2102000,-0.0148000,-0.0317000, 0.9770000)

jigsaw51, jigsaw150, jigsaw1000 - new programs that are versions of the standard "jigsaw" program that are set up to allow processing different numbers of images. The number in the name indicates the maximum number of images that can be processed. Versions that can process a smaller number of images need a smaller amount of virtual memory and swap space to run. However, note that the user should normally just run "jigsaw", which will now determine which version of the program is to be used.

level2.pdf - This new procedure will project cleaned up unprojected images to a sinusoidal equal-area projection. The input images will have the noise along the edges trimmed first before projecting them.

lincont - A new ISIS application that computes the linear continuum or band depth for an ISIS cube. The PDF contains documentation on the formulas used for these operations.

mar10cal - This new program will radiometrically calibrate Mariner 10 data.

msi2isis - This new program will convert NEAR Discovery MSI FITs images into ISIS cubes.

multiimp - New IDL program that is similar to the "multimatch" program except that it runs "tvimp" rather than "tvtie".

outline - New ISIS program. It will draw a border around distinct image sections that have different DN values. The user is allowed to specify the DN value to use to outline each section. (Note that this program operates on ISIS cube files. There is a separate program called "outlin" that operates on PICS format files.)

photosort - New program that takes an input list of filenames and creates a new output file that contains the filenames of the images that fall within the user specified constraints.

slosin - A new program that was brought over from the Vax. It removes periodic noise, in the sample direction, from VIKING images. The images should be run through LEVEL1 before "slosin" is run on the images.

solardiv - New program that divides each spectrum in a cube by the solar flux (obtained from the BAND_BIN_SOLAR_FLUX keyword in the BAND_BIN GROUP).

specfix - a new program that applies a specialized filter to remove noise (despike) from imaging spectrometers such as NIMS.

tran - New procedure that projects an image (e.g., the moscut.cub and mos4.cub) in a map projection to a Transverse Mercator projection. The output mosaics will be trimmed to match the boundaries established for the Mars 1:500k mapping format.

unmix - New program that computes spectral unmixing using a specified set of endmember spectra. The endmember spectra can be obtained from either an Instrument Spectral Library (ISL) file (e.g., average spectra for regions produced by the "cv" program) or a table file.

vlev1.pdf - This new procedure will process raw edr images through "findrx", "remrx", "boxfilter", and "vikcal" or "voycal". This generates LEVEL1 files that have reseau marks and random noise removed and that have been radiometrically corrected. Currently only Viking and Voyager data should be processed using this pdf.

wimp - New IDL program that is a widget interface to the "tvimp" program.

4. PROGRAM-SPECIFIC UPDATES

avg_sd - Has been modified to make it callable from a PDF procedure. When called from a procedure all the output values that were previously written to the screen and/or an output file will be returned to the procedure. Also, the program will now output the following results to the screen, to a file, or to a PDF procedure:
average, standard deviation, number of valid pixels, number of NULL pixels, number of high instrument saturation pixels, number of low instrument saturation pixels, number of high resolution saturation pixels, number of low instrument resolution saturation pixels, minimum pixel value found, and the maximum pixel value found.

bandcopy - Has been modified to handle arbitrary image sizes. There are no longer restrictions on spatial image sizes.

bandtab - Has been modified to read/write the INSTRUMENT_ID keyword as a STRING rather than a LITERAL. Also, a routine in the ISL library (s_isl_specs) was modified for the same reason. This will affect all ISIS ISLs that have been generated previously. They will have to be regenerated.

base2l1 - Is a program that was brought over to the system from vax. It will create the trans formation data file (TFILE) for "geom" (see "geom" for more information) to project a Clementine, Galileo, Mariner, Viking or Voyager "level2" cube or mosaic from a map projection to "level1" geometry using a "level1" cube-to-match. It has been changed so that the user's input, falseno, gets transmitted to "jigsaw" as product ID.

boxfilter - A problem in "boxfilter" has been found and fixed that occured when the cube data storage order is not BSQ. It incorrectly attempted to process the cube by transposing axis dimensions.

There have been two bug fixes made to "boxfilter". These fixes affect the LPF, STDZ, VAR and SUM filter options only.

For the LPF, VAR and SUM filters, the input buffer was not being mapped for data tests because values were perceived to always change. This is *NOT* necessarily true when the minimum number of points are not achieved in the box. "boxfilter" now tests the input value, preserving special pixels or setting the output pixel value to NULL for all other values for this case.

For the STDZ filter, pre-existing special pixel were not being preserved when tests on limits were applied. Existing special pixel values are now propagated to the output file or set to NULL for all other values.

boxfilter - The following changes have been made:

1. The TOL parameter has been renamed to TOLMIN and now applies only to the STD and STDZ filters. It has been removed from use in the NOIS filter option. Its default has also been changed to 0.0 which will essentially result in all valid pixels being passed through this initial check onto the standard deviation test. Use TOL2H and TOL2L for the NOIS filter.
2. The TOL2H and TOL2L parameters apply only to the NOIS filter option. It has been removed from use in the STD and STDZ filter options. The default value for both values is the NULL TAE value. This will require the user to enter an appropriate value or boxfilter will terminate with an appropriate error message. Use TOLMIN for STD and STDZ filtering.
3. A correction was made in how TOLMIN, TOL2H and TOL2L are computed internally when processing 8 and 16 bit data types. From the user perspective, these values are the differences between the input pixel value and the low pass component of the box centered at the input pixel value.
4. For the STD and STDZ filter options, 8 and 16 bit data is now converted to floating point when the CORE_BASE is not 0.0. This is required because of how CORE_BASE affect squares of numbers. The algebra is really ugly and was not previously handled properly!
5. New features have been added to give a count of the number of pixels affected by the filter. Reports the number of pixels NULLed due to insufficent points in the boxcar, the number of pixels replaced by the filtering option and the number of pixels that could not be represented in the output file due to minimum or maximum representable values exceeded (if incurred).
6. For the LPFV and HPF filters when the pixel values are lower than LOW or higher than HIGH as specified by the user, the output pixel value is set to LRS and HRS, respectively. They were set to NULL previously.
7. New parameters have been added to give the user more control over how special pixels are processed/handled in the filtering process. MINREPL allows the user to select the special pixel value to replace the center pixel when insufficient points are in the box (subject to MIN or FRACTION). REPLACE allows the user to specify the value to replace input DN when the filtering process encounters a case where the output DN cannot be determined. Both MINREPL and REPLACE allows a value of "NONE" that will propagate the input value unchanged to the output file where appropriate. The SPECIAL parameter allows the user to select which special pixel values are subject to change in a given filter. This parameter can be used to construct a "mask" of sorts that will specify which special pixel values can be modified by the filter. See the level 2 help section in the PDF for details on these parameters. Note that not all filters utilize all these parameters. The affected filters are noted in the new parameter help sections. Also note that defaults are provided for these parameters that preserve previous special pixel processing behavior in "boxfilter".
8. Added a new parameter called MEMORY. This parameter allows the user to select the amount of memory boxfilter uses for internal buffering. The default (of no value for this parameter) is 12 megabytes. Values provided by the user must be expressed in megabytes (multiples of 1,024,000 bytes). This parameter most likely will not be used. It is for extreme cases where large files cannot be processed with 12MB or for systems that cannot allocate this amount of memory. Note that a large amount of memory does not necessarily result in more efficient processing as it could invoke excessive system page faulting, so use with caution! However, 12MB may not be the optimal memory allocation for some systems. Older versions of boxfilter referred to the value of the environment variable BOXFILTER_MEMORY to override the 12MB default. This version no longer refers to this environment variable.
9. Numerous documentation changes were made to the PDF.

One other note: I have extensively tested processing native pixel data types versus 32-bit REAL values of the same data. The most noticable differences were found when running STD, STDZ and NOIS filters on these cases. There will most likely be discrepancies reported in the number of pixels that were modified by the filter in these cases. This can be attributed to rounding of numbers to the representable values in data types other than 32-bit. More accurate results can be attained when processing 32-bit data except when your data starts out in 8 or 16 bit data and you then convert to 32-bit REALs. These cases will actually yield the same results with the only difference being a preservation of accuracy in the output modified DN values. Note that the same statement applies to all filters. The resulting values of the filters will not be rounded when processing in 32 bit.

The other big gotcha I noticed is when you change the output representable ranges for 8 and 16 bit data types. This will re-distribute the data over different values thus altering filtering results. It should go without saying that if you stretch your data to different ranges, this will also alter filtering results.

cd2isis.pdf - Added an option to the PDF to specify the temporary PICS file name. This file could be large and should be directed to a disk with sufficient space. It defaults to a file called "cd2pics.pic" in the current directory. It will be removed at the completion of the application.

cd2pics - Was not processing 16-bit images correctly. Support for other pixel data types has been added.

cd2pics - Modifed to handle larger images (like Magellan). It previously was limited to around a 2MB file. It is now limited only by the amount of memory available on the computer system (RAM, disk space and mostly swap space).

cpylab - Modified to fix the number of arguements used to call the subroutine CHAR2I.

cpylab - Modified so that it can handle more than 50 keywords in the BAND_BIN group. It now handles up to 200.

cubeit - Has been changed to allow stacking input cubes with multiple core planes and back planes.

cubelook - Has had a bug fixed that appeared on Dec Alpha systems. It appeared as extraneous characters when "painting" a shift in the column or row direction of the window. It appears to be a bug in DEC's implemention of curses. Unfortunately, the fix results in an appearance of "jumpy" scrolling, but this fix only applies to DEC Alphas and is only used on this system.

cv - Several changes were made to the "cv" program:

1. On the Data Plot windows (e.g., spectrum plots), the "hidden" Functions menu has been replaced by a menu bar. Clicking with the RIGHT mouse button to the LEFT of the right hand plot axis will now toggle whether the plot names are displayed.
2. User control of the plotting range has been added to the Data Plot windows. When you first create a plot window to display a spatial profile, spectrum or average spectrum, the "Y Mode" is set to "Display Range," which means that changing the image display range will automatically change the Y range on the plot. You can also set the "Y Mode" to "Auto Range," which will adjust the Y range according to the data being plotted.
3. The "Plot Parameters" control panel has been added (under "Options" on the menu bar). This allows modifying the plot titles, plotting range, and selecting horizontal/vertical grid lines.
4. Zooming in/out on a plot can be accomplished with the MIDDLE mouse button. Hold the button down and drag to draw a box around an area of the plot. When you release the button, the plot range is set to the area within the box. You can zoom multiple times. Clicking the MIDDLE button will back up one zoom level. There are also functions in the "Options" menu to reset/backup the zoom and also to reset the plot X and/or Y range to the range used when the window was first created.
5. The "Save Copy" button on the plot window menu bar will create a new plot window and copy the current plot (average data for an average spectrum) into it. Each time you hit "Save Copy," the current plot will be copied to the new window and plotted with a different color. When plots are "saved" in this manner, a box is drawn on the Front View windows to mark the location of the "saved" plot. The Marker Control panel contains functions to enable/ disable the drawing of these overlay boxes and also to delete all the current boxes.
6. By default, cubes are now deleted from memory after they are not being displayed. If you start the program with "cv,/retain" then all cubes are retained in memory until you exit from "cv."
7. A bug was fixed that was causing the spectrum plots to sometimes not be updated properly when the cursor was moved on the Side View or Bottom View windows.
8. A bug was fixed that caused the list of data plot names to sometimes be displayed incorrectly when the user typed in a change to a plot name.
9. A bug was fixed in the Instrument Spectral Library viewer function.
10. A bug was fixed that was causing the cube file to not be properly closed, which was a problem when the user had written a set of ROI definitions to a backplane.
11. A bug was fixed in the polygon drawing for the Region of Interest (ROI) function. (It was not properly handling the case where the user held the middle mouse button down and then pressed the right button before releasing the middle button.)
12. Two functions have been added to the "Write File" menu in the Region of Interest Control Panel. These new functions will write the statistics for each defined region into a separate text file.
13. The View Control panel now has a new button that allows enabling/disabling the automatic display of the Subsampled Front View window when the Main Front View window does not contain the entire spatial-spatial area. Disabling the Subsampled window will increase the speed of displaying a new band image when the cube contains a large spatial area.
14. Changed to allow (but not require) double quotes around the value of the NATIVE_START_TIME label keyword. (This is used when reporting values of the NATIVE_TIME backplane on NIMS cubes. The change is needed because NIMS cubes on CD-ROMs will use the double quotes, but the double quotes were not previously used.)
15. Changed to allow the program to be run with an IDL "runtime" license rather than requiring a full IDL license.

fixlucas - Fixed a minor bug. Changed MOON to 'MOON' in the conditional statements.

fixlucas - Corrected to reverse the orientation of the longitude for lunar cubes only.

fixlucas - Has been updated to conform to comments from the external review.

fixlucas - Some of the Lucas vtubs arrive with an unrealistic longitude range - -10^22 -> +10^22 deg. "fixlucas" has been modified to test for and skip vtubs with longitude range greater than 360 deg.

geom - Had a bug in the way it processed the labels from the tfile. The problem has been fixed.

geom - Has been fixed to correctly handle "base2l1" labels.

getfname - This program is called from TAE procedures and will now return filenames that are listed in multiple columns. It will return all columns from a requested line. (The previous version of this program could handle only a single column). The help information in the .pdf was updated to clarify the operation of the program.

histplane - The program previously switched the high instrument saturation and high representation saturation counts in the session log file. The problem has been corrected.

imp2isis - This program has been deleted. It is replaced by the "uax2isis" program.

isis2pics - Modified to fix the number of arguements used to call the subroutine CHAR2I.

isisdoc - A bug which didn't allow writing to one output file has been fixed.

jigsaw - A new version of "jigsaw" has been checked into ISIS. It includes the following changes:

1. A new option has been added to create a file listing images and pixel errors when the error tolerance is exceeded. This file is created if it doesn't exist already, otherwise it is appended.
2. "jigsaw" now checks for islands (images without match point connections to any other images in the list)
3. "jigsaw" now checks the number of input files and runs the most efficient version of the program based on the input file count.

jigsaw - was updated to run on only 125 files on SGI systems. The change should not affect other systems.

labels - Will now allow you to change keyword values which are arrays. You must change the entire array, not just a single element.

labels - Will now allow you to enter a keyword in scientific notation.

labels - Added option to enter a text file containing keyvalues to be changed in the labels. This allows setting a keyword value that has an array of more than 50 values. Also added option to allow changing a specific value in a keyword array. Changed the default value of the BAND parameter to YES so that the default is to list the entire label.

magcube - A new version of magcube has been checked into the system which fixes a problem that occurred when scaling down cubes.

makeflat - Will now allow the input cube to be multi-band in addition to being able to enter many single-band input cubes.

matchpt - Has been changed to include a new TAE paramater. The new paramater, CHKOFF, is used to determine whether the user wants to check the offset calculated with the LTOL and STOL. The default value of YES will perform the check and this is the way the program worked in the past. If CHKOFF is set to NO, this check will not be performed.

Also, the program will now not end in an error when 3 match points are not found. A warning message will be written to the session log and the screen and the two filenames will be written to the NO3MATCH file, but the program will terminate normally. Previously when this condition occurred the program termintated with an error condition.

The new version will choose the best three matchpoints from the 9 matchpoints that are looked at.

mosaic - There is a new version of "mosaic" with the following changes:

1. There is an option to allow mosaicking of input low and high sat values. See the parameter, mossat, in the pdf for more information.
2. Added capability for IMP mosaicking.
3. If the BANDCHK parameter is set to YES, the input cube does not need to have the same number of bands as the mosaic cube. "mosaic" will look at the BAND_BIN_CENTER keyword to determine which output band the input band should be mosaicked into.
4. Had a bug in the tvreg option which has been fixed.

multimatch - Has been updated so that the default saturation percent to use for the automatic display range computation has been changed from 0.5 percent to 0.25 percent.

Note that when you run "multimatch," you can override the default "tvtie" parameters by specifying desired values on the command line. The names of these parameters are the same as the names of the "tvtie" parameters. For example, to use a different set of saturation values:

    IDL> multimatch, 'no3match.fil', losat1=0.1, hisat1=0.3, &
        losat2=0.1, hisat2=0.3

naiflab - has been updated with the following changes:

1. It will only load c-kernels that cover the time span of the input image and avoid exceeding the 20 file limitation on loaded kernels.
2. The GMT warning message will no longer appear unless no time values are found on the labels.

noproj - The CAMERA_DISTORTION keyword label generated by "noproj" has been corrected.

pics2isis - Modified to fix the number of arguements used to call the subroutine CHAR2I.

photom - has been changed to allow the "MOONPR" function and the "albedo" option to be used for planets other than the moon. The default values used when these two options are chosen are still for the moon, however. A warning message is issued to the log file and the terminal if a user chooses MOONPR and albedo just to let the user know they are using lunar default values for a planet other than the moon. For lunar images the program works exactly as it did before.

photompr - code was added to check to see if the user input a CLAT or CLON value. If the user did not input a CLAT or CLON, the program will read the spice information. If the user did input a CLAT or CLON the program will not read the spice information. Also, if the user has chosen the TRIM function, entering a CLAT or CLON value will affect the way the image is trimmed.

photostat - Has been updated to include the target body in the output.

photostat - additions to "photostat" include a report on the filter number and the user's option to input the band number for which the filter will be reported. A bug that caused an error when the center of the image is not on the planet has been fixed. Now, instead of "photostat" bombing, the observer/solar azimuth will default to 0.0, whenever the center of the image is not on the planet.

pics2isis - Was fixed to report mapping errors to the log file PRINT.PRT as well as the screen.

planlab - During the SGI port a problem was found in planlab when it loaded the NAIF SPICE kernels. The problem has been fixed.

qview - "qview" has the following changes:

1. Multiple files and min/max values can be entered on the TAE command line or tutoring qview and using the FROM and MIN and MAX parameters.
2. The file created when saving listings now writes one line of header information.
3. The directional movement in the roam option has been changed. For example, if the user clicks on the right side of the display, the image will move to the right, click on the top of the display and the image will move up.
4. The help information in the .pdf has been extensively revised to describe all the new and old features of the program.

ratio - There is a new version of "ratio" which fixes a few bugs. The pdf has also been revised.

size - The scale calculations in the program have been corrected.

spam2isl - Changed so that it assumes the value of the INSTRUMENT_ID label keyword is a string rather than a literal

specpr2isl - Changed so that it assumes the value of the INSTRUMENT_ID label keyword is a string rather than a literal.

spice2rand - Had a bug that caused it to not work correctly when the nomat and err=BOTH options were used. The error occurred in the nomat file if the first image in the list did not have any match points in its label. The problem has been fixed.

spice2rand - Has been modified to write a more accurate error message when the SPICE file cannot be opened.

spicelab - A parameter called VFLAG defaults to "NO", in which case the program runs as it always did. If VFLAG is set to "yes", and EXT = "rmb", then for Viking and Voyager missions, the program will look in the .rmb files and then if the file is not found will look in the .gem files.

ssical - Changed the check for valid CAL_TARGET_CODE for C10 and E11 to look for "10" and "11" rather than "C10" and "E11".

tvimp - Has been updated to include the same features as the latest version of "tvtie," i.e., uniformly scales the display range into the available range of display colors/intensities, displays special pixels as user-controllable colors, and allows user to control the color used for drawing marks on the images.

tvimp - A bug was fixed which was causing it to not record the FSC (IMAGE_NUMBER label keyword) for the second image in the MATCH_POINT_n keyword in the first image.

tvimp - Removed the "Xpalette" function from the Tools menu. (Changes to the program awhile ago caused this function to quit working properly. Choosing Xpalette caused the program to crash and this function was not very useful anyway.)

tvimp - Has been updated to allow storing 99 match points on the label of the image (rather than the previous limit of 20).

tvtie - The image display and marker display have been improved. When displaying the images, the pixel display range is now uniformly scaled into the number of screen intensities or colors that are available to the current IDL session. This corrects the problem of saturating more than the requested percent of pixels at the high end of the display range (which resulted from a simple-minded approach to using the IDL display functions).
Display of special pixels is now handled the same as in the "cv" program. By default, null pixels and both types of low saturation are displayed as black. By default, both types of high saturation are displayed as white. Note that by default, valid pixels that are outside the display range are displayed as red or green. A new function button labelled "Special Pixel Colors" creates a control panel that allows changing the displayed colors for the various types of special pixels. Note that changing the stretch or color table for the displayed images will not change the colors that are used for the special pixels.
The user now has the option of selecting the color to be used for displaying the markers and the planet's limb. The default color is now red rather than white/black.

tvtie - Updated so that a proper error message is displayed if an input file is in the wrong format, e.g., you are running on a Sun machine and the input file is in DEC Alpha format.

tvtie - Removed the "Xpalette" function from the Tools menu.

uax2isis - Is a replacement of the "imp2isis" application. The name has been changed because it better describes the origin of the input data. These files are those from the University of Arizona in its own format. The data is most likely from the IMP (Imager for Mars Pathfinder) camera. It now supports all the possible data formats and generates many more ISIS specific keywords. Output ISIS cube files generated from this application may be projected using "imp2prj". This application does not yet support the proposed PDS format of the IMP files.
It has been modified to allow for more flexibility in parsing IMP UAX keyword values. Some problem values are now parsed correctly and undefined or unknown keywords associated with a value are reported as a warning.

uvviscal - The problem with "uvviscal" on the Alphas has been corrected. You will now get the exact same values when you run "uvviscal" on the Suns and Alphas.

5. UPDATES APPLICABLE TO MULTIPLE PROGRAMS

TAE Menus - The program menus (obtained by typing "menu" at the TAE prompt) have been updated to include the new programs and to use the various Mission-specific categories.

Instrument Spectral Library Files - The INSTRUMENT_ID label keyword was changed from a LITERAL value to a STRING value. This affects all ISL programs. Also, this affects all ISIS ISLs that have been generated previously. They will have to be regenerated.

A bug in u_error parsing error messages was fixed.

The subroutine spi_distort.F has been changed to check that the distortion flag "IDIS" has been set to 0 -- undistorted to distorted -- before the reseaus are read. This change should eliminate the 'error searching for RESEAU_SAMPLE keyword' message that some processors were getting when picking tie points in "wtie" or "tvtie" using level1 outputs from "base2l1".

6. DATA FILE UPDATES

GENERAL DATA FILES

The projection_name.sav file has been updated to add/subtract IMP specific projection keywords to/from the labels. This will change the state of IMP projection labels ("imp2prj").

Changes have been made to several libmap routines and the projection_name.sav file. The changes relate to writing to labels and extracting from labels the keyword MAP_PROJECTION_FRAME which will now be a three character literal.

Some typos have been corrected in the projection_name.sav file in $ISISDATA.

GALILEO DATA FILES

Galileo SPICE files have been updated.

The Galileo SSI spacecraft kernels have been updated to include reconstructed E4 data and updated predict for the remainder of the mission. Also the predict c-kernel for day 12/19/96 had been updated.

A missing Galileo SSI Callisto c-kernel file has been installed on the alphas.

The file s970223a.bsp replaces s970109a.bsp and has reconstructions for E4 & E6.

The spacecraft clock kernel has been updated for Galileo.

The file cke06bmg.plt is an E06 predict.

The file ckmn21aa.plt is an E06 predict. The file ckg01b.plt is a G01 update.

The G01 c-kernel has now been installed on the alphas.

The file ckg07a1g.plt is a G07 predict. The file ckg07a2g.plt is a G07 predict. The file ckg07a2h.plt is a G07 predict. The file ckmn23aa.plt is a G07 predict. The file ckg07bmd.plt is a G07 predict. The file ckg02b.plt is the G02 final.
The file ckotm6.plt is a G01 update.
The file s970401a.bsp is reconstruction of G1 to E6 and a predict for Mar 4, 97 to Oct 10, 99.
The file s970408a.bsp is reconstruction of initial orbit to E6 and a predict for Apr 7, 97 to Jan 1, 98.

The files:

ckg01ajh.plt
s970103a.bsp
s970223a.bsp
have been deleted.

The number of c-kernels allowed in the SPICE kernel table listing has been increased to accomodate all the Galileo encounters. This increase corrects a problem naiflab was having in getting to the Clementine c-kernels.

The file ckc03b.plt is a C03 final.

The files:

ckc03ah1.plt
ckc03ai2.plt
ckc03ak2.plt
have been deleted.

The spacecraft clock kernel has been updated for Galileo. The leapseconds file has been updated.

Galileo SPICE files have been temporarily updated with old predicts for C03 because the final ckc03b.plt has gaps in it.

The files ckc03ah1.plt and ckc03ak2.plt are the C03 predicts.

The naif_kernels.sav file has been updated with access to E06-Callisto.

The file ckg08aph.plt was added and is a G08 predict. A new naif_kernels.sav file was added also.

The file s970509a.bsp is a partial G08 reconstruction. It is in addition to s970417a.bsp, which should not be deleted.

The file s970513a.bsp is a reconstruction of initial orbit to G07, preliminary reconstruction of post G07 & G08, and the rest is predict for 97-5-12 to 98-1-1.

The file:

s970509a.bsp
has been deleted.

The file s970627a.bsp is a temporary C09 reconstruct not to be used after 10 Jul 97.

The file s970709a.bsp is a reconstruction of initial orbit to G08, preliminary reconstruction of C09 through 97-06-27, and the rest is predict for remainder of orbit 9 to end of GEM mission. It replaces s970513.bsp & s970627a.bsp which have been deleted. Temporary file s970606a.bsp has also been deleted.

The file ckg07b.plt is a G-07 final.

The spacecraft clock kernel has been updated for Galileo. A clock kernel for C09 only has been added, but until there is clarification on how to use it, naif_kernels.sav will only look at the regular clock kernel.

The file ckc10aod.plt is a C-10 predict.

The file s970922a.bsp is a reconstruction of initial orbit to C09, 97-08-22, preliminary reconstruction of C10 through 97-09-17, and the rest is predict for remainder of orbit 10 to end of GEM mission. It replaces s970801a.bsp which has been deleted.

The file s970801a.bsp is a reconstruction of initial orbit to C-09 through 97-07-27 with the rest predict through remainder of orbit 9 to end of GEM mission. It replaces s970709a.bsp which has been deleted. The file s970417a.bsp has also been deleted.

The files gll_gain.sav and grns.cal03 were updated.

The file s971016a.bsp is a reconstruction of initial orbit to C09, 97-08-22, preliminary reconstruction of C10 through 97-09-17, and the rest is predict for remainder of orbit 10 to end of GEM mission. It replaces s970922a.bsp which has been deleted.

ckmn35aa.plt - a new Galileo mission C-10 predict in $ISISGALDATA which replaces the previous C-10 predict, ckc10bmg.plt

The Galileo calibration file for the green filter had a problem in the labels. This has been corrected and a new version of the file has been put into $ISISGALDATA.

cke11aig.plt - a new Galileo mission E-11 predict in $ISISGALDATA. cke11bel.plt - a new Galileo mission E-11 predict in $ISISGALDATA.

ckmn37aa.plt - is a new Galileo mission E-11 predict which replaces ckmn36ab.plt in $ISISGALDATA.

The file ckmn36ab.plt has been deleted from $ISISGALDATA.

CLEMENTINE DATA FILES

A new version of CLEMENTINE_1_UVVIS_SORT.sav has been installed because the previous version was missing some data.

An updated control file that contains additonal files for the Clementine mosaic has been added to the $ISISCLEMDATA area of ISIS.

IMAGER FOR MARS PATHFINDER (IMP) DATA FILES

This set of data files is new in this release of ISIS.

MARINER 9 DATA FILES

This set of data files is new in this release of ISIS.

MARINER 10 DATA FILES

This set of data files is new in this release of ISIS. These include the latest spice files from Mert Davies for Mariner 10.

7. CHANGES OF INTEREST TO ISIS PROGRAMMERS

The C interface to DOCUBE is now available for use. See the C DOUSER example program $ISIS/apps/gen/bandstats.c for insight in how to use this new interface. There is also a document in $ISISDOC called "c_douser.txt" that has information in it as well.

There are two new FORTRAN dynamic memory routines available in the ISIS system. u_allocate_buffer.F will allocate dynamic memory for the size and number of elements specified by the caller. This routine is recommended to replace all occurances of u_allocate_memory CALLs in FORTRAN source because it now computes the proper *index* to the allocated memory based upon the size of the desired data array. Note that this routine will also allocate CHARACTER arrays and provided proper access to the allocated memory. There are examples in the source code documentation. u_release_buffer.F is the accompanying memory deallocation routine to u_allocate_buffer.F. The arguments MUST be the same as was used in the call to u_allocate_buffer. See the documentation for both these routines for more information. Note that this routine only applies to FORTRAN...it is not available to C programmers. Also, all FORTRAN source code that uses these routine must include isis_memory.inc and declare the appropriate variables (_D_MEMNDX) in order to preserve portability. The "bandcopy" program, bandcopy.F, shows an example of how to use u_allocate_buffer.

ISIS programmers can now filter data with the same filtering techinques used in the boxfilter application. See the file isisfilters.c in $ISIS/subs/libisistools for usage.

implib.h - This include file contains function prototypes for IMP mission specific library routines. It should be included in all C applications that call these routines.

The fix in map_projt in libmap changes a format statement reference in an encode line which writes a label. The fix puts pole latitiude as latitude rather than longitude.

The following map projection data & include files have been updated to include information for the Imager for Mars Pathfinder, IMP.

prj.h
prj.inc
projection_name.sav

Several new routines were added to the ISIS utility routine library. u_roi_setxy and u_roi_setline draw into a region of interest buffer. up_clip, up_clcode, up_getticloc, up_numtics and u_format are utility routines for graphics plotting and coordinate grid drawing.

imp_photoch.F has been updated to test for az/el points that are off the image. It will now return an error code of -3 if a point is more than 90 degrees away from the line of boresight.

The isisdef.inc FORTRAN include file has been updated to properly declare INTEGER*4 (as opposed to INTEGER) variables.

8. CHANGES OF INTEREST TO INSTALLATION MANAGERS

To allow having both an IDL 5.0.2 version and an IDL 4.0.1b version of the ISIS IDL display programs, the IDL source files are no longer stored in $ISISEXE (which is the same as $ISISR/exe). The IDL 5.0.2 version of the programs and their associated ".hlp" files are now stored in $ISISR/exe/idl. The IDL 4.0.1b version of the files are now stored in $ISISR/exe/idl4. The "isispgmdef" setup file has been modified to put $ISISIDL in the IDL_PATH environment variable. "isispgmdef" also sets $ISISIDL to point to the IDL 5.0.2 version of the programs in $ISISR/exe/idl.

Most of the IDL display programs are compiled at run-time. However, the one exception is the "cv" program. Due to its large size, "cv" is pre-compiled and appropriate versions of a "cv.sav" file are stored in $ISISR/exe/idl and $ISISR/exe/idl4. The various ".pro" source files for "cv" are also stored in these directories. However, note that the main "cv" source file is stored as "cv.pro.src" rather than as "cv.pro". This allows IDL to run the pre-compiled "cv.sav" and also allows re-compiling the program in the unlikely event that that is necessary.

<End of Release Notes>