A Satellaview research blog.

I knew I was forgetting something all this time… EDIT: Now with No-Intro comparison

So, as I was going through testing ROMs to add to the Java emulator, I came across a couple redumps I never actually got around to showing on the blog!

For two years!

Oopsie-daisy!

The game they’re for has been put on the Java emulator already. It’s Dr. Mario!

I don’t remember the details too hard on these redumps, but if my memory serves they have the ol’ “Deleted Headers”. The Java embed is of the first of the two – it appears to play identically to the commonly circulating ROM.

I deeper analysis might come around.. if I don’t forget for another 2 years.

Dr. Mario (BS)
ROM Download

EDIT Oct. 15, 2012: I’ve attempted to do a more in-depth comparison with the previous dump, HOWEVER…

It appears said previous dump, as it appears in the no-intro set, has a hacked header. I’m not entirely sure how this happened, but regardless…

Here’s 7FC0-7FDF on the previous ROM:

42 53 5F 44 72 20 4D 61 72 69 6F 00 FF 00 00 00
00 00 00 00 00 00 FF FF 20 10 33 02 CE B7 31 4B

On the same addresses, these two redumps have the following bytes in common:

82 63 82 92 81 44 B3 7D 83 8A 83 49 82 61 82 72
FF 00 00 00 00 00 ** ** 20 10 00 02 00 00 00 00

(The ** are differing bytes. Their location is in the “date” area of the ROM, anyhow.)

Notably, while the latter two’s headers are consistent with a “Deleted” game, the former has a header that just plain looks screwed up, especially with spelling “BS_Dr.Mario” the way it does.

Furthermore, none of the ROMs load in BSNES, but while this is expected of the deleted header ROMs, the “hacked” header ROM causes an odd behavior which causes the ROM loader to crash.

To boot up Dr. Mario closer to the way it was meant to be played, the header should look something more like this;

82 63 82 92 81 44 B3 7D 83 8A 83 49 82 61 82 72
FF 00 00 00 00 00 ** ** 20 10 33 02 C7 B7 31 4B

Note how the “hacked header” ROM does appear to have the checksum and maker values correct. Either the hacker attempted to restore these, or the header hack was based off a previously “pure” dump. This may need to be investigated.

… To think we don’t even have as simple a ROM as this in it’s purest state? Sheesh…



9 thoughts on I knew I was forgetting something all this time… EDIT: Now with No-Intro comparison

  1. Well done. The old dump clearly has its internal name hacked.

    BSX headers have been a real pain in the neck since they were even designed to be modified by the system for some extent. I’ve come to a conclusion that the most important bytes that we need to have unmodified and original in the header are at $7/F:FD0 (Allocation bytes) and $7/F:FD9 (ROM type), among the internal checksum of course.

    Of those the checksum is the only one that could get overwritten by the BIOS but calculating a new one is easy if we have those two other bits of info correct.


    By the way, if you’re ever reading this, LuigiBlood, do you know if it could be possible to download a part of the streamed broadcast data into PSRAM for additional graphics (Inishie no Sekiban dungeon tile data) while running most of the game from the FLASH memory? By judging how those two memory chips can be mapped independently at different memory addresses it could theoretically work, but is this actually possible?

Add Your Comment

* Indicates Required Field

Your email address will not be published.

*