Sample Details

 (Offset 20)

 There can be either 15 or 31 samples (the number of samples can be
 determined from the sample count marker - see below). There is an entry
 here for each sample with the following information:

 ■ A 22 byte character string (padded with null characters) holding the name
   of the sample. There should be at least one null character.
 ■ The length of the sampled sound in words. The value is stored in a
   word in AMIGA format, i.e. most significant byte first - opposite to the
   way a word is stored on an 80?86.
 ■ The fine-tuning information for the sample stored in a byte. Only 4 bits
   are actually used. The number is signed, so a value of F (hex) would be
   a fine-tuning of -1.
 ■ The default volume of the sample stored in a byte. The maximum the volume
   can be is 64 (decimal).
 ■ Two words in AMIGA format holding the loop start and the loop length.
   These are used for samples that need to be continous. Since a sample
   can only be a finite length, the loop start value specifies where to
   continue playing from when the end of the sample is reached, and the
   loop length value specifies how far to continue playing once you have
   looped (you may not want to continue to the end).
   * Note that these values are word offsets, e.g. a loop offset of 4 means
     the sample will loop back to byte number 8 (word number 4).
   * I have also seen occurences of these values being out of range. It is
     possible that certain trackers interpret the loop length as a loop end.

 Note that the actual sampled sound data is stored in the sixth section at
 the very end of the MOD file.