Maybe a Bug in Yamaha FM Registers

[crisp]

Hello, I surfed the net for some informations about Soundhardware and found an interesting PPT file which has a very interesting information on Page 20 with following readings...:

Yamaha FM registers in apparently random  order (like Japanese House numbers)

Bug found by Williams/Watts in 1985 in YM3526

Did a workaround so quick time to market for Commodore computers

Bug found  again by Linux programmers when trying to emulate Soundblaster registers

2004 - OPL2 core now  integrated into the South Bridge of the motherboard's chipset

Bug now  in every PC? (>1 Billion?)

The twenty-two registers (20-35, 40-55, etc.) have an unusual

order due to the use of two operators for each FM voice.

Channel        1    2    3  4    5    6  7    8  9

Operator 1    00  01  02  08  09  0A  10  11  12

Operator 2    03  04  05  0B  0C  0D  13  14  15

you can find the PPT here: http://research.microsoft.com/users/lyn/soundcardhistory.ppt

I just wanted to know if this bug even exist in the opl3 used in the MIDIBOX FM?

Thanks a lot... I am building it at the moment... also a SID Module...

Greets and Thumbs up for the good work!

[TK.]

This "bug" has been declared as a feature ;-) You will find these addresses in the specification, and MBFM uses translation tables based on this spec in order to address the operators correctly.

But note also, that MBFM uses the OPL3 in 4 operator mode only, 2 operator mode (which would result into more voices, but also with less quality) is not supported!

Best Regards, Thorsten.