Alright, so I did a bunch of tests with Tales of Phantasia, and I achieved some pretty interesting results!1) The saving with cp4+ patches actually do work!
There is just a delay when the game actually decides to write to the SRAM.cp4c:
Test 1: Started a game with zero SRAM files, saved, played a bit more, saved again. Only first save was actually recorded.
Test 2: Started a game with zero SRAM files, saved and then quit. Save was not recorded.1-beta:
Exact same results as cp4c.cp3a:
Saved and then quit right after. Save was recorded.SlowROM check fix does not affect any of this.
I tried both no slowROM check fix as well as with slowROM check fix for all of my tests.The game actually does not write anything to SRAM when you choose to save (save slot blocks within SRAM are still blank) until after a delay.
After the delay, then the game will actually decide to write to SRAM.
What's it waiting for to write to SRAM?2) There are 5 instances of slowROM checks with the patch according to uCON64.
According to uCON64, the 5 instances occur at the following offsets:01 0D 422 instances
x5FB136A9 01 85 0D3 instances
I knew about 0D42, but I had never heard of the other one.
The only information I could find on it was located within the uCON64 source code.
There's a comment next to A901850D which says:
// special one (used by Konami and Jaleco? sometimes)
I did some testing with each of these offsets.The ONLY one that actually did anything was x70DD.
It fixed the sound issues.
The rest provided zero results, or at least no results that were immediately apparent.
They did not fix the sound issues.
These other ones might not be slowROM checks, maybe just a coincidence?