FreeUSP Release 1.02 Patch 2
This patch contains program fixes and enhancements as follows:

trcgp/usp/src/cmd/Makefile -

Removed wavest target. Added saneusp target. (program underwent name change).



Program symname -

Fixed coding bug. This is not a user program, but is used in building the system.



Program eign -

Fixed bugs in Makefile.



Program saneusp -

Name changed from wavest.



Program ikp -

Changed the default verbosity off. Program attempted to print character strings that had not been allocated. This worked on some architectures, but not all.



Program crossfilt -

Program to do filtering on two data sets based on their amplitude relationships defined with a polygonal zone.



Program ampspec -

Added -PWR flag to optionally compute the power spectrum rather than the default amplitude spectrum. Also corrected bug on SGI that would not allow -O [] type output.



Program diffr -

Added normalization to the differential operator.



Program tim2hed3d -

Fixed a bug in the xsd_xy subroutine that caused the pickfile read to bomb on the SGIs.



Program attin -

Fixed the bug of handling velocity trace extension.



Program attsel -

Fixed an error in setting the last record value resulting in reporting a stop 100. Made a change so program does not try to check for velocity file when running in an ikp net and the -at flag is set.



Program replace -

Fixed -s[] -e[] resolution to assume that sample 1 occurs at time zero [unless otherwise noted] which is the USP standard. This code for whatever reason assumed that sample 1 occured at time delta t, which caused an off by one error when replacing zones of data using the -s[] -e[] capability. Added option to do salt velocity replacement. Velocities greater than that of salt will be replaced with zeros, and velocities ajacent to the salt bodies will be replaced with reasonable sediment velocities, instead of suspicious velocities resulting from velocity grid decimation.



Program xyz2sis -

Add -believe option to allow use of the -flat2 option and have the sample number be the actual sample desired as opposed to the sample as counted from the base of the trace which is the default.



Program hdrswap -

Upped the size of space allocated for the line header initially. Current 10000 byte allocation was exceeded by a dataset and failure resulted. New allocation is 32768 bytes which is then reduced after header is read in.



Program clean -

Modified the code to get rid of hard-coded 10000 byte limit on input line headers and/or trace data. Dynamic allocation is now used.



Program smedian -

Changed -rs[] -re[] to refer to processing limits as all data is now passed. Also fixed 1 sample shift in output when -median was flagged.



Program rmo -

Program rmo computes and applies residual moveout corrections. Input data must be NMO corrected and CDP ordered.