i found out how the program reads.
the two big files are only for Movies, in JAP there are 12, for USA i can count 14.
Code:
it starts reading from
Offset 72 (4 Byte) = Offset of Movie_01 (calculate it with * 2048), then Offset 76 (4 Byte) = Size of Movie_01
the last is Offset 92 (8 byte) = ???
then it starts again, only 32 Byte later
Off. 104 = Offset * 2048
Off. 108 = Size
Off: 124 = ???
...
but when you look at the offsets, you find garbage data, this is maybe a encryption or something else, atm don't know.