// NEWS
ver 0.59 Fev 16 2008
free_tab() function has been addded in dialog_sinle_chan.c, to avoid crash on non allocated memory of tab_file[]
Major bug corrected in get_path() function for WIN32 application 
(separator can be either '/' or '\').
Free_mem() called after closing the ALL_Channels window (in All_fct() function) ) for recoveringthe allocated memory


ver 0.58 Fev 10 2008
This release gives a portable code for the WIN32 application.

ver 0.57 Dec 07 2007
selection of multiple files is now enable with mouse delection (code from Petr Vyslouzil);
the selection using wild chars (*) in the dialog text box remains also possible.

ver 0.56 Nov 24 2997
Major bugs corrected with alloc() and memcpy() functions for 64 bits machines
 (without influence or problems on 32 bits ones) .

ver 0.53 Oct 16 2997
* In Fourier Series menu: the choice of the fundamental frequency F0, is proposed (as the inverse 
of the window duration), but not imposed;  thus operator can choose F0, in fucntion of 'a priori' 
informations on the frequency content.

ver 0.52 Sep 23 2997
* Documentation has changed with all download made from sourceforge.net
* Computation of Fourier Series was added in Fourier menu. 
* Estimated waveform is plotted from the Fourier series; Coefficients of Fourier and modulus
 of spectrum, are written in text files in the directory pointed by STK_LOG.


ver 0.51 Aug 25 2997
* In STK : a very small bugg is corrected, in function CallBackFile_Selection_Ascii(), that was not 
  storing the path to the selected directory of signal in SAC_ASCII 
* STK_0.51_win32.exe (for Windows) is compiled with optimized code (-O2 option) for acceleration of code
* in Mwp : test-null_channel() function was added to avoid process of signal with null values

ver 0.50 July 4 2007
* major bug corrected in function CallBackFile_Selection_Bin, if the environment variable STK_SAC_BIN_PATH is not defined,the char * Sig_path variable points to NULL. It makes crash of STK on SUN machine (strangely not on PV Linux machine).
 
ver 0.49 May 23 2008 
* Mwp :  a new utility program; Mwp is the moment magnitude computed on P waves.
Mwp is very useful for early seismic warning on P wave. 

ver 0.48 May 2008 
PPOL_MTPA utility program: extension of PPOL in the frequency domain with MTPA method (Multi Prolate Taper Analysis);
same goal than PPOL, but a different tool in the frequency domain.

ver 0.47 07 mar 2007
A minor bug was fixed in PPOL. An error of compilation was produced by a missing line of comments.

ver 0.46 04 mar 2007
PPOL (P wave POLarisation) has been added, which was developed for a study on anisotropy
 of the earth mantle deduced from the anomaly of the azimuth of P wave.
 The principle is to automatically select a window centered on the P wave,
 and to compute the eigen direction of the linear vibration vector.
 The window containing the P wave is calculated from the origin time and coordinates
 of the hypocenter using the IASPEI time tables.

ver 0.45 21 sept 2006
* new version of MM_mag in Utilities program: version 1.09
* MM_mag accepts 7 arguments: dd/mm/yyyy hh:mm:ss lat lon depth
  if these 7args are input, they are taken into account,
  else all coordinates are read in the configuration file MM_mag_xxx.ini 
* tests reinforced for Rayleigh and Love out the time frame of selected signal
* Plots of all the results are performed, using GMT with a single script: gmt_all_plot.bash
  it plots in postscipt files the following results:
  MM_R_all_res_res.txt
  MM_R_results.txt
  MM_R_vs_period.txt
  MM_R_vs_azim.txt
  MM_L_all_res_res.txt
  MM_L_results.txt
  MM_l_vs_period.txt
  MM_L_vs_azim.txt


ver 0.44 05 sept 2006
* the WIN32 version seems to be stable
* code is very much slower in WIN32, than UNIX
* code also does not support optimization 
* in UNIX version the FFT produce segmentation fault with optimization (-O) for large number    of points.
* compilation with libfonyx.lib (a native format for our Lab)
  this library will not be distributed as we are alone to use it
* the frequency in mHz are written in Mm_vs_period
* minor bug: in MM_mag in Utilies: the file Mm_vs_periodwas opened and not closed
* code cleaned: 
	change g_malloc() to malloc() 
	changed f_free() to free() to avoid error with WINDOWS runtime	

ver 0.43 30 Aug 2006
* new version of MM_mag in Utilies : version 1.07
* major bug corrected in cal_mm.cpp with the selection of period
  which is function of focal depth
* A new file Mw_vs_period containing the average of Mw 
  at each period is computed.


ver 0.42 12 Aug 2006: 
* new version of MM_mag in Utilies : version 1.06
* major bug corrected list_dir_unix3 (basic.cpp )
* results for Love are presented like for Rayleigh
* seimic moment are in the last column (more convenient to plot
  them vs azimuth for example)

ver 0.41 13 jun 2006
 + compilation of whole project STK for Windows  (using {MinGW and mysis} ).
   except GMT_plot and SVDS in utilities programs.

ver 0.40 June 10 2006
 + compilation of an executable of MM_mag for Windows; also in the same way a win32 executable 
	is available for  AZI, SEED_RESP, SAC2WAVE and MECA in the Utilities Programs.
	Compilation was made using {MinGW and mysis}.



ver 0.39 May 26 2006
 + MAJOR BUG corrected in TIME-FREQUENCY ANALYSIS: the step in frequency was wrong, and so introduced randomly until 20-30% of bias in frequency estimation. This bug is probably inside the 10 last version of STK. DO NOT use time-frequency with version older than 0.39.

ver 0.38 May 25 2006
 + doc corrected in file Doc/Install.html
 + new version of Mm_mag 1.05 : extrema values of magnitudes are removed 
   in fonction of standard error criterion

ver 0.37 some Docs cleaned

ver 0.36 dec 21 2005
 + code cleaned in SEED_RESP Utility program
 + change in the computation of frequency Step df : it becomes a geometrical suite, instead of regular step between FMIN and FMAX;
  consequently the modulus do not need a large number of point for describing the entire spectrum (a small number of point 
  as 512 is enough (against at least 25000, with old method using the regular step)   

ver 0.35 dec 12 2005
 + major bug corrected in Utilies\MM_mag : in conversion from SAC_BIN to SAC_ASCII
 + new version of MM_mag 1.03
  
nov 04 2005 ver 0.34
 + time-frequency greatly improved with overlapped window
 + also numerical rounding errors on the frequency step were corrected  
 + bug corrected in prog. plot_time_freq.c in Utilies/GMT_plot


oct 27 2005  ver 0.33 
 + major bug corrected in the COMPILATION of Polar/Spectra utility program
 + new version Polar_0.05 has been created

 oct 17 2005, ver 0.32
 + minor changes and code cleanup in utility program: MM_mag

 oct 10 2005, ver 0.31
 + new utility program: MM_mag for computation of Mm, Mw, Mo
   Mw is the moment magnitude, and Mo the scalar seismic moment
   Mm is the mantle magnitude. This program needs signals in SAC_ASCII format

Aug 27, 2005 ver 0.30
 + New utility program for converting signal in SAC_ASCII files into audio files (*.wav)
  The program, allows to convert an entire directory of signals; wildchar can be used to select files. 
  The sampling frequency can be changed, with the choice of mono/stereo channels, and signals can be filtered 
  or not before conversion in wav format. 

Aug 2005 ver 0.29
  + code cleaned
  + mouse selection of the button activated only for zoom in all press_event(), release_event() funtions,
  + in preparation of future uses of the other buttons

Aug 2005 ver 0.28
  +Major bug corrected in grad_x2, if the signal length is equal exactly to 1 day (8640s)
  +export to ASCII added in Polar_3D menu
  +prostcript file showing polarisation, can be generated by the new program:
  	plot_3comp_and_polar.c (compilation ./compil in Utilities/GMT_plot )
  +example of GMT script added for plotting polarisation, in Utilities/GMT_plot (example_of_polar.bash)

July 2005 ver 0.27
  +MTPA method added in Utilies, for spectral analysis and polarizarion in the frequency domain
  +GMT scripts for plotting results of MTPA in postscript
  +GMT script for plotting transfert functions in SEED_RESPONSE utility
    
May 2005 ver 0.26
  + Minor bug corrected in instantaneous polarisation

Apr 2004 ver 0.25
  + Minor bug in Integrator_Fc : the gain was not correct for signal with high_sampling rates 
  + minor bug in Utilities/plot_all_chan.c if the channel is not standard (ie N,Z,E), the min and max was not read
	
Mar 2004 ver 0.24

* export to ascii menu for 
	current_channel plotted
	NZE channels
	ALL channels
	
* in Utilities/GMT_plot programms:
	prostcript files can be generated by the new programs:
	+ plot_3comp.c
	+ plot_current_chan.c
	+ plot_all_chan.c
 they are compiled with the command ./compil
 These programs create a command file for GMT (Generic Mapping Tool), 
 which is able to generate itself postscript graphics.
 
 	
feb 2004
** major bug corrected in SEED_RESPONSE Utility (seed_resp.c) to take into account type B response
  (must multiplicate pole and zeros by 2 PI and A0 by Pi**(Np-Nz) for type B)


jan_2004: 
* add exponantial taper to time_frequency & spectra menu
* instantaneus measure of azimuth and incidence on 3_componants signals
* minor bug corrected for VH* channel (dt <= 10 s) 
  for time_frequency, polar_evolutive and instantaneous_polar menus 
  (minimum length of windows adapted to sampling period)


oct 2004:
* new menu : export to ASCII file, for writting results in ASCII in the FILE menu (main window)

* in time_freq.c : the results are written in Log directory,so the time-frequency graph can be plotted
        with other high-level programm(GMT, Math_lab for examples). This is realized with the Export to ASCII menu.

*in Utilities:
      a new module is added: GMT_plot; it contains programms for plotting the results
        directly in postscript (if high quality graphics are needed, for plubications for example)
        ->plot_time_freq programm create a postscript file in the Log directory,via GMT software
        (GMT :http://gmt.soest.hawaii.edu).
        MECA: a useful small program to plot focal mecanism with  beach-ball in text mode (in the console)
        
        

Sept 2004:
*in SVDS : possibilty to integrate signal or not before polarisation analysis
   one argument is added: INT or NO_INT
*in STK: in the time-frequency function : the choice among 4 taper functions were added
  (Hanning, Hamming, Blackman or Hanning generalized (the default)
  also a possibility to smooth or not the spectra


Dec 2003 :  ver 0.13 Evolutive polarization

June 2003 : search the comp N, Z,E in the header file
            add the compensator order 1 filter in menu
            
April 2003: bug with the optimization option, the new compiler gcc 3.2:
            -O, -O1 and -O2 produce a segmentation fault in the FFT,
            for large number of points ( > 132000 pts)
            so do not use - O option.

            bug fixed in the SVDS utility in version 0.10
            
feb 2003    - Utilities programs:
		SWAP  for swapping SAC_BIN files
		SEED_RESPONSE: to compute instrumental response from response file with zeros and poles
		AZI: a simple program to calculate azimuth, back_azimuth and distance
		SVDS  Singular Value Decomposition of Signal, to compute eigen directions of 3 comp. signal from its eigen vectors.
									
25 jan 2003 - differents taper windows (hannning, Hamming, Blackman) in Fourier menu

july 2002    Time-frequency representation in Fourier menu

xxxxxxxxxx   Possibility to read 3 componants in select_file
						(putting * instead of the axis)


