I believe that the Collection of Mana version of FFA is running around 0.4% faster. Here is a video comparing different versions of the game side by side:
All 4 versions of the game are synced at the press of start after typing in the girl's name. What we are looking for is how well the 4 versions stay in sync. We aren't interested in frame-perfect synchronization all the way through, just roughly at where the Jackal is on the screen (left/center/right). The Collection of Mana footage quickly de-syncs and accumulates an 8-second lead over the course of 30 minutes. The very same footage slowed down to about 95.55% stays in sync.
Additionally, the above video also demonstrates that the Analogue Pocket emulates the game accurately in docked mode.
The Game Boy runs at a weird framerate - 59.727500569606fps. Most TVs can't display that kind of video signal. So some emulators round up to a more common speed of 60fps. This speed-up is not noticeable with the naked eye and it's fine in casual play. But it can impact the timing on a speedrun. This is a phenomenon similar to but separate from the Super Game Boy issue. It looks like Collection of Mana on the Nintendo Switch is using that kind of inaccurate emulator.
To counter the effect the actual time needs to be multiplied by 1.004562377929684687237604521
(Can be safely rounded up to 1.00456237)
So if a Collection of Mana run is 30m 05s: That is 1805 seconds That is about 1813.235 seconds with the multiplier applied Which gives us an adjusted time of 30m 13s 235ms - over 8 seconds longer.
In order for Collection of Mana footage to appear at the original speed in a video editing program, the footage needs to be slowed down to 99.54583428267666%
Collection of Mana v 1.0.1 on the Nintendo Switch
Special thanks to @aquova@mastodon.gamedev.place for providing the Analogue Pocket footage Analogue Pocket running the game on openFPGA Library Version 74A501E8 - Mar. 8, 2023
I would appreciate it if another person could confirm these findings and the math. I could have made an error somewhere.
A simple test anybody could do is to start a stopwatch after typing in the girl's name. When the stopwatch reaches exactly 24 minutes, take note of where the Jackal is on the screen. A game running accurately should have the Jackal hit the right edge about at that time. A game running 0.4% faster will have the Jackal on the left edge.
If the findings are verified I would propose adding the re-timing to the rules. The Collection of Mana is a convenient way to play the game on modern consoles, especially for new players. I think there should be a clear official ruling on hiw this platform should be handled
