Back to the forum: Nintendo - Super Nintendo Entertainment System

 Rev 1
Argument: ROM data , Reference: 0934 , Closed by: xuom2
skaman @ 2017-07-22 21:46:59

I solved the mystery of the Genjuu Ryodan Rev 1.1. The answer is that there is no Rev 1.1.

I dumped 2 copies of the cart with another on the way. A couple other dumpers also shared their cart dumps. All copies of Genjuu Ryodan were the known Rev 1.0.

If you want to arrive at the suspected Genjuu Ryodan Rev 1.1 with CRC32 9DD7B12E, take the Rev 1.0 and change all the bytes to 0xFF starting from 0x3F0000.

This same problem came up in another RetroFreak post on FE Thracia 776 with CRC32 AD2C2FD9. This is the FE 776 Rev 1.1 with all 0xFFs starting from 0x3F0000.

The problem is with the RetroFreak. When it dumps a large LoROM cart, it starts reading from the usual LoROM Bank 0x00. For the larger ROMs when you get to Bank 0x7E, you're actually reading WRAM. On a cart reader, the WRAM area will be 0xFFs that is why you end up with the different CRC. To correctly dump a large LoROM cart, you must start reading from Bank 0x80 to get the complete data at the end.

I noticed this problem with reading the wrong LoROM Banks a long time ago when I was writing the code for my cart dumper but I forgot about it until I recently ran into the issue as I wrote updated firmware for the Retrode.

I'd be careful relying on any RetroFreak info/dump of a LoROM game > 24Mbit. Unless RetroFreak fixed their code, then it'll most likely be bad.