Sondex MITcal Readme
====================

Release History
===============

V2.60  07 December 2006
=======================
Added the facility to convert new cal files to old.

V2.50  21 July 2005
===================
Added the option of not interpolating the calibrations for bad arms.

V2.40  24 May 2004
==================
Added the option of temperature compensation using the sample file
instead of the coefficient file.

V2.30  16 July 2003
===================
Corrected a bug in the deviation calculation for LAS output.
Corrected a bug dealing with negative linespeed.
Added the facility to convert old cal files to new.
Sample files are now at 5 degree intervals above 130 degrees.

V2.21  9 July 2002
==================
Corrected a bug when dealing with interpolated arms.  If the calibration
file also had bad values for this arm then the interpolated values for the
arm could end up wrong.  Now the calibration for the interpolated arm is
interpolated as well as the data.

V2.20  24 April 2002
====================
Facility added to splice MIT data files.

Can now skip stripes with checksum errors when creating both MIT and LAS
files.

Added the facility to skip stripes marked as collars.  This allows the
user to manually despike a log by marking individual stripes as "collars"
in CAL3 and then use MITcal to remove those stripes.

Centralising algorithms changed with some new configuration options.
Details are in the help file.

V2.10  10 October 2000
======================
Facility added to depth match a log by stretching or squeezing.

Warning now given when changing the output depth units and writing an MIT
data file and not resampling depth, as the resultant pulses per unit depth
value will be non-integer.

If an arm gain is not sensible (eg if an arm has failed during calibration),
a nominal value is used instead so that processing is not aborted.

V2.03  1 September 2000
=======================
Warning now given when resampling depth and writing an MIT data file if the
selected output interval will result in a non-integer pulses per unit depth.

If the MIT data file header has a blank for pulses per unit depth, 100 pulses
per foot is assumed and a warning is given.

Nominal ID value is now used correctly in the centralising routines when the
normal centralising algorithms fail to find the centre.

V2.02  23 November 1999
=======================
Original 32-bit version.