PhaPacker format (also called ProPacker V1.0 by NoiseConverter & Exotic Ripper)
                 (also called HanniPacker and/or StrangePlayer)
Created by Azatoth / Phenomena
Analysed by Asle / ReDoX (asle@mygale.org)

Source :
 - "In The Kitchen" demo by Anarchy
 - "Deja Vu" demo by Anarchy

Offset    size (byte)    Comment
------    -----------    -------

      ***************************************
      * the following is repeated 31 times  *
      * with 14 bytes description for 1 smp *
******************************************************
                                                     *
 0             2         Sample Size / 2             *
 2             1         ??                          *
 3             1         Volume (0 - 40h)            *
 4             2         Loop Start / 2              *
 6             2         Loop Size / 2               *
 8             4         address of this sample      *
 12 ($C)       1         finetune - $b (?seems ok?)  *
 13 ($D)       1         ??                          *
                                                     *
******************************************************

434 ($1B2)    14         ?!?!?!? Couldn't understand
                         nothing out of these bytes.
                         Though, it's fairly safe to
                         say that there might be the
                         number of pattern in pattern
                         list ... . It's possible to
                         convert without this info, but
                         it's VERY hard !.

448 ($1C0)    512 ($200) pattern list (128 dword)

 it's a table of pointers which points at the adresses
of the pattern data in the "PhaPacked" file. So, the
pattern 0 will be the one starting right after the smp
data. Beware of the unused patterns when converting !.


960 ($3C0)    ??         sample data

 stored the same way as Protracker.


??            ??         pattern data

 Stored like protracker .. meaning:
 row 01: voice 1 | voice 2 | voice 3 | voice 4
 row 02: voice 1 | voice 2 | voice 3 | voice 4
 ...

 each voice is 4 bytes (ot two bytes .. see below).

Two cases depending of first byte:

 0000-0000  0000-0000
 \       /  \       /
   $ff        value

 If the first byte is $ff, then the following byte is
the number of times the previous value in THIS row will
be re-used. Well, in fact, it's ($ff-value) number of
times !.
 example :

  01-02-0A-0A  00-00-0F-06  02-00-00-37  FF-F5
  1st four bytes : Sample #1, Note #1, Fx Vol Slide, FxArg $0A.
  2nd four bytes : no sample, no note, Fx Set Speed, FxArg $06.
  3rd four bytes : sample #2, no note, no Fx or arpeggio, FxArg $37.
  4th block : $ff meaning previous block (02-00-00-37) will be repeated
             FF-F5=0A times IN THE VOICE #3 !!!.



 0000-0000  0000-0000  0000-0000  0000-0000
 ||\     /  \       /  |  | \  /  \       /
 ||sample   relative    \/   Fx    Fx Arg
 ||number     note    unused
 ||          number
 ||
 description bits, both set to 0 in this case.

relative note number: (ProTracker)

C-1 to B-1 : 856,808,762,720,678,640,604,570,538,508,480,453
C-2 to B-2 : 428,404,381,360,339,320,302,285,269,254,240,226
C-3 to B-3 : 214,202,190,180,170,160,151,143,135,127,120,113

becomes : (Pha Packer)

C-1 to B-1 : 02,04,06,08,0a,0c,0e,10,13,15,17,19
C-2 to B-2 : 1a,1c,1e,20,22,24,26,28,2a,2c,2e,30
C-3 to B-3 : 32,34,36,38,3a,3c,3e,40,42,44,46,48