Wii64 (Rice) vs Not64 (glN64) 60FPS framerate/performance test with Mario Kart 64
Posted: Mon Jul 06, 2020 4:49 am
I did a test with FIX94's fork of Wii64 and Extrems Not64, using the Mario Kart 64 60FPS Hack to measure which had better framerate/performance and basically the Wii64 version with the Rice video plugin had far better performance than the glN64 GFX plugin both on Wii64 and Not64. Not only is able to keep 60FPS on 2-Player Grand Prix, it also keeps a steady 30FPS on 4-Player Versus, on the most demanding track, which it would be DKJP, only losing near 8 frames on slowndown, but most of the time the average framerate is 58-60, the other track being KD, which lags quite a lot on console, the Rice video plugin can keep above average framerate. All the other tracks run at a solid 60FPS on 2-player GP and 30FPS on 4-player versus, with the worst slowndown being, like, 2 frames? Anyways, I will post the results for both Not64 and Wii64 Rice.
Before doing this test I thought the Not64 GFX plugin was the best it could get the Wii to achieve maximum framerate and stability, but I was wrong. I am quite impressed and still can't believe this Rice video plugin can achieve lagless 60FPS Single and Two player GP, and lagless 30FPS 3-4 multiplayer with Mario Kart 64.
Note: LimitVIs = 1 in settings.cfg must be set on Wii64 since it doesn't feature the VI/s frame limiter on the GUI yet.
Note2: Lagfix is MANDATORY for 60FPS, specially with Rice GFX where the 60FPS code glitches the graphics sometimes, Lagfix increases performance overall and prevent the graphic mess on Rice but breaks the final lap music, this seems to happen on all emulators and not only mupen64gc, so I developed a workaround to fix that with more Gameshark codes.
Wii64 (Rice GFX)
2-Player Grand Prix @60FPS
KD: worst 52FPS, average 58-60.
DKJP: worst 52FPS, average 56-60.
All other tracks: 60FPS, with pretty much null frame loss.
Wii64 (Rice GFX)
4-Player Versus @30FPS
KD: worst 26FPS, average 29-30. (1.0x)
DKJP: worst 26FPS, average 28-30. (1.0x)
All other tracks: 30FPS (1.0x), with pretty much null frame loss.
Bug encountered: When loading the game for the first time everything works correctly, however if you restart game or reload the same game again some textures like the skybox and the item boxes won't load properly. It happens with Mario Kart 64 only on Rice GFX plugin.
https://prnt.sc/tcui1w
Wii64 (glN64 GFX) and Not64
(Mushroom Cup) LR@50+++, MMF@39+, KTB@52+, KD@40+
(Flower Cup) TT@54++++, FS@48+++, CM@60+++++, MR@50+++
(Star Cup) WS@57++++, SL@57++++, RRy@52+++, BC@54++++
(Special Cup) DKJP@40+, YV @52+++, BB@50+++, RRd@57++++
All tracks on 2P Versus reach 60FPS unless stated otherwise.
4-player Versus@30FPS
Instead, I use the Speed Multiplier next to the VI count, on some tracks I include them.
(Note: 60VIS=30FPS)
(Mushroom Cup) LR: 0.9-1x, MMF:0.7-0.9x, KTB:0.9-1x, KD: 39.6@0.7x-0.8x
(Flower Cup) TT:57@1.0x, FS:57@0.9-1.0x, CM:57@1.0x, MR:53@0.8x1.0x
(Star Cup) WS: 58@1.0x, SL: 54@0.9-1.0x, RR: 48@08-1.0x, BC: 57@1.0x
(Special Cup) DKJP: 45@0.8-1.0x, YV: 57@1.0x, BB: 51.3@0.-1.0x, RR: 57@1.0x
All tracks on 4P reach 60VIs (30FPS) unless stated otherwise.
I will upload one or two videos showing the differences later, but you can test yourself injecting these codes into the ROM with triclon's Mario Kart 64 Gameshark Code Injector:
https://sites.google.com/view/triclons- ... e-injector
Lagfix
800EA6D2 000D
60FPS Hack (All Modes)
81001890 2419 //60FPS Hack
81001892 0001 //60FPS Hack
81001894 2419 //60FPS Hack
81001896 0001 //60FPS Hack
80000FE3 0000 //Single Player 60FPS
800014CF 0001 //Single Player 60FPS
81001A38 2409 //Two Player 60fps tempo
81001A3A 0001 //Two Player 60fps tempo
81001A3C 2409 //Two Player 60fps tempo
81001A3E 0001 //Two Player 60fps tempo
81001C90 240A //Three and Four Player 60fps tempo
81001C92 0001 //Three and Four Player 60fps tempo
81001C94 240A //Three and Four Player 60fps tempo
81001C96 0001 //Three and Four Player 60fps tempo
80122CBB 001C //Hack
810AC5D4 2401 //Hack
810AC5D6 0006 //Hack
Two player fixed 30fps tempo
81001C90 240A //Three and Four Player 30fps tempo
81001C92 0002 //Three and Four Player 30fps tempo
81001C94 240A //Three and Four Player 30fps tempo
81001C96 0002 //Three and Four Player 30fps tempo
Fix final lap not working broken by lagfix:
D00DC53F 0003 //asks if game is set to battle mode
D00DC52F 0003 //asks if there are 3 or 4 players
D00DC513 0005 //asks if the game has finished
D13B03D2 0200 //asks if the battle BGM is playing
813B03D2 0000 //mutes Big Donut / Skyscraper
D00DC53F 0003 //asks if game is set to battle mode
D00DC52F 0003 //asks if there are 3 or 4 players
D00DC513 0005 //asks if the game has finished
D13B03C4 0002 //asks if the battle BGM is playing
813B03C4 0000 //mutes Block Fort / Double Deck
It must be noted that 30FPS and 60FPS tempo can't go in the same time, they must be injected separately as the game is only able to handle 60FPS if the Single Player 60FPS is injected, by default it will go at 30FPS when not injected, when these codes aren't activated, the game will change the pacing of the race depending on the number of players and the track, more information about this here: http://tasvideos.org/GameResources/N64/ ... tches.html
These results makes me wonder if the clock divider feature and the other updates could further improve this already good build if ported to Wii64 Rice.
Before doing this test I thought the Not64 GFX plugin was the best it could get the Wii to achieve maximum framerate and stability, but I was wrong. I am quite impressed and still can't believe this Rice video plugin can achieve lagless 60FPS Single and Two player GP, and lagless 30FPS 3-4 multiplayer with Mario Kart 64.
Note: LimitVIs = 1 in settings.cfg must be set on Wii64 since it doesn't feature the VI/s frame limiter on the GUI yet.
Note2: Lagfix is MANDATORY for 60FPS, specially with Rice GFX where the 60FPS code glitches the graphics sometimes, Lagfix increases performance overall and prevent the graphic mess on Rice but breaks the final lap music, this seems to happen on all emulators and not only mupen64gc, so I developed a workaround to fix that with more Gameshark codes.
Wii64 (Rice GFX)
2-Player Grand Prix @60FPS
KD: worst 52FPS, average 58-60.
DKJP: worst 52FPS, average 56-60.
All other tracks: 60FPS, with pretty much null frame loss.
Wii64 (Rice GFX)
4-Player Versus @30FPS
KD: worst 26FPS, average 29-30. (1.0x)
DKJP: worst 26FPS, average 28-30. (1.0x)
All other tracks: 30FPS (1.0x), with pretty much null frame loss.
Bug encountered: When loading the game for the first time everything works correctly, however if you restart game or reload the same game again some textures like the skybox and the item boxes won't load properly. It happens with Mario Kart 64 only on Rice GFX plugin.
https://prnt.sc/tcui1w
Wii64 (glN64 GFX) and Not64
(Mushroom Cup) LR@50+++, MMF@39+, KTB@52+, KD@40+
(Flower Cup) TT@54++++, FS@48+++, CM@60+++++, MR@50+++
(Star Cup) WS@57++++, SL@57++++, RRy@52+++, BC@54++++
(Special Cup) DKJP@40+, YV @52+++, BB@50+++, RRd@57++++
All tracks on 2P Versus reach 60FPS unless stated otherwise.
4-player Versus@30FPS
Instead, I use the Speed Multiplier next to the VI count, on some tracks I include them.
(Note: 60VIS=30FPS)
(Mushroom Cup) LR: 0.9-1x, MMF:0.7-0.9x, KTB:0.9-1x, KD: 39.6@0.7x-0.8x
(Flower Cup) TT:57@1.0x, FS:57@0.9-1.0x, CM:57@1.0x, MR:53@0.8x1.0x
(Star Cup) WS: 58@1.0x, SL: 54@0.9-1.0x, RR: 48@08-1.0x, BC: 57@1.0x
(Special Cup) DKJP: 45@0.8-1.0x, YV: 57@1.0x, BB: 51.3@0.-1.0x, RR: 57@1.0x
All tracks on 4P reach 60VIs (30FPS) unless stated otherwise.
I will upload one or two videos showing the differences later, but you can test yourself injecting these codes into the ROM with triclon's Mario Kart 64 Gameshark Code Injector:
https://sites.google.com/view/triclons- ... e-injector
Lagfix
800EA6D2 000D
60FPS Hack (All Modes)
81001890 2419 //60FPS Hack
81001892 0001 //60FPS Hack
81001894 2419 //60FPS Hack
81001896 0001 //60FPS Hack
80000FE3 0000 //Single Player 60FPS
800014CF 0001 //Single Player 60FPS
81001A38 2409 //Two Player 60fps tempo
81001A3A 0001 //Two Player 60fps tempo
81001A3C 2409 //Two Player 60fps tempo
81001A3E 0001 //Two Player 60fps tempo
81001C90 240A //Three and Four Player 60fps tempo
81001C92 0001 //Three and Four Player 60fps tempo
81001C94 240A //Three and Four Player 60fps tempo
81001C96 0001 //Three and Four Player 60fps tempo
80122CBB 001C //Hack
810AC5D4 2401 //Hack
810AC5D6 0006 //Hack
Two player fixed 30fps tempo
81001C90 240A //Three and Four Player 30fps tempo
81001C92 0002 //Three and Four Player 30fps tempo
81001C94 240A //Three and Four Player 30fps tempo
81001C96 0002 //Three and Four Player 30fps tempo
Fix final lap not working broken by lagfix:
D00DC53F 0003 //asks if game is set to battle mode
D00DC52F 0003 //asks if there are 3 or 4 players
D00DC513 0005 //asks if the game has finished
D13B03D2 0200 //asks if the battle BGM is playing
813B03D2 0000 //mutes Big Donut / Skyscraper
D00DC53F 0003 //asks if game is set to battle mode
D00DC52F 0003 //asks if there are 3 or 4 players
D00DC513 0005 //asks if the game has finished
D13B03C4 0002 //asks if the battle BGM is playing
813B03C4 0000 //mutes Block Fort / Double Deck
It must be noted that 30FPS and 60FPS tempo can't go in the same time, they must be injected separately as the game is only able to handle 60FPS if the Single Player 60FPS is injected, by default it will go at 30FPS when not injected, when these codes aren't activated, the game will change the pacing of the race depending on the number of players and the track, more information about this here: http://tasvideos.org/GameResources/N64/ ... tches.html
These results makes me wonder if the clock divider feature and the other updates could further improve this already good build if ported to Wii64 Rice.