Contents of one observation

Data, i.e. observational parameters as well as spectra, is organized in the following way:

• One observation is self-contained. All the information needed to reduce it is recorded on the same few disk blocks. It may be one spectrum or one continuum drift.
• Each observation is divided in several sections, containing header parameters or data:
  • General information (date, times, local coordinates, sequence number, ...)
  • Positional information (source, name, astronomical coordinates, epoch, offsets, ...)
  • Spectral information (number of spectra, line names ans frequencies, resolutions, number of channels, ...)
  • Data
  • ...
  Each type of section is known by the system by means of an integer code. It is not required that all sections be present. Some are written by the observing program at the time of the observations. Some may be dropped during the reduction, being replaced by new ones describing the results of the different stages of the reduction procedure.
• Each observation begins with the following words:

  
        length     content 
  
  (bytes) 		 
  
  
  4 		 '2' (identification code) 
  
  
  4 		 number of logical blocks 
  
  
  4 		 number of byte describing the observation 
  
  
  4 		 address of the data section (bytes) 
  
  
  4 		 length  of the data section (bytes) 
  
  
  4 		 the number of header sections 
  
  
  4 		 the index entry of the observation 
  
  
  4*nsec 		 the codes of the sections 
  
  
  4*nsec 		 the addresses of the sections 
  
  
  4*nsec 		 the lengths of the sections  
  

  Thus the first section will e.g. be stored at the address (28+12*nsec), followed by the second section, and so on up to the end of the observation. The length of one particular type of section is not fixed. Adding extra parameters is then possible in each section.