The main code is in the folder pss_sfdb It is crea_sfdb.c To compile it: ./compila RELATIVE To run it: 1- ./crea_sfdb.out < input 2- ./crea_sfdb.out and then at the prompt answer to the questions... Input files are in the ffl list of the frame files of each detector Output files are SFDB files. You can produce FFTs of any length (in practice the closer power of 2 will be used). Follows examples on how we use the code to produce 4 different sets of FFTs. The four frequency bands and their length are described in the All Sky O2 paper and referecnes therein. Example of input files (or input to give at the prompt) H1 % (or L1 ) 2 %if 2 creates only SFDB files. With 3 creates also PEAKMAP files. 2 2016-11-30_2016-11-30_2016-12-30.ffl %name of the ffl file name of the science file %data not in science periods will be put to zero H1:DCH-CLEAN_STRAIN_C02 % channel name. For Livingston H1 --> L1 64 % subsampling factor. Use 64 for the first band, up to 128 Hz. Use 16 for the second (128-512 Hz). 3rd: 8 4th: 4 1 134217728 %number of samples in the FFT. This is for the first band. For the second: 67108864. 3rd: 33554432. 4th: 16777216 2 % with 2 FFTs overlaps. This is what we use. 1000000 % put here the total number of FFTs you want to produce. If very high then all the possible FFTs will be done 128 % this is a subsampling factor, used to store a reduced version of the power spctra in the header of files 5 % window used 100 % frequency of the veto to clean the data -1 % use always -1 100 % total number of FFT in the file Format of the files: header data (Real, Imaginary part) Tools to read them are in the public software SNAG (matlab based) The code extract_band.c gives an example on how to read SFDB files and extract a bandwitdh. The function used is readfilesfdb, and it is in (the library) pss_sfdb.c