03759: gradius2, gradius2a, gradius2b, vulcan, vulcana, vulcanb: Title screen demo plays the stages in the wrong order

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
03759	Gameplay	Trivial	Always	Feb 28, 2010, 19:34	Mar 3, 2010, 17:20

Tester	gamestone	View Status	Public	Platform	MAME (Official Binary)
Assigned To		Resolution	Open	OS	Windows XP (32-bit)
Status [?]	Acknowledged			Driver	konami/twin16.cpp
Version	0.136	Fixed in Version		Build	Normal
		Fixed in Git Commit		Github Pull Request #

Summary	03759: gradius2, gradius2a, gradius2b, vulcan, vulcana, vulcanb: Title screen demo plays the stages in the wrong order
Description	The attract mode title loop shows three stages with Heros Screen and History Intro in between each one. These are all taken from the original Japanese PCB Gradius II GOFER no Yabou on first booting the board, I tested this three times to make sure it repeats the same loop each time which it does. Stage 1 Burning Heat - 1 - Player gets to the to the end of stage boss (Phoenix) and crashes into the blue flame bullet. Heros Screen History Intro Stage 3 Crystal - 1 - Player crashes into a crystal chunk. Heros Screen History Intro Stage 6 Maximum Speed - 1 - Player gets to end of stage boss (Big Core MkII) and then crashes into a laser. Heros Screen History Intro Stage 1 Burning Heat - 2 - Player crashes into the flaming snake. Heros Screen History Intro Stage 3 Crystal - 2 - Player gets to end of stage boss (Crystal Core) and then crashes into a laser. Heros Screen History Intro Stage 6 Maximum Speed - 2 - Always the same as first loop. Summary... Stage 1 - Demo loop one always reaches the end of stage boss then on demo loop 2 crashes into the flaming snake, it then alternates between the two. Stage 3 - The first demo loop crashes into crystal chunk, second demo loop gets to the stage boss, then every loop after that crashes into various crystal chunks and never gets to the stage boss ever again. I can't see any pattern in the 28 loops I recorded. Stage 6 - This is always the same demo, gets to end of stage boss and crashes into laser. From what I can see by comparing the waveforms, MAME missing the first loop of the three demo stages and then the order of the demo stages is wrong to the arcade original.
Steps To Reproduce	Just watch the title screen demo play.
Additional Information	I've recorded the audio for 28 demo loops (arcade) so you can compare it to the MAME 17 demo loops. Sound recording of the title screen demo loop from the original Gradius II GOFER no Yabou PCB. 122meg Stereo mp3 http://www.gamestone.co.uk/mame/gradius_ii_gofer_no_yabou_attract_mode.mp3 Sound recording from MAME 0.136 using the "mame gradius2 -wavwrite filename" command line switch. 46meg Stereo mp3 http://www.gamestone.co.uk/mame/gradius_ii_gofer_no_yabou_attract_mode_mame.mp3
Github Commit

Flags	Verified with Original
Regression Version
Affected Sets / Systems	gradius2, gradius2a, gradius2b, vulcan, vulcana, vulcanb

Attached Files

No.05801 Haze Senior Tester Feb 28, 2010, 20:13	does the set on your PCB match the one you were testing with in MAME. This type of thing is controlled by the program code (although if it's using an internal random number generator any slight timing error could cause a different sequence of stages, but you'd see the same on the real PCB if any of your clocks started to drift too)
No.05803 gamestone Tester Feb 28, 2010, 20:54	Checking the rom numbers on the board, they correspond to the gradius2 set gradius2, gradius2a, gradius2b, vulcan, vulcana, vulcanb all play the same, which is not like the original arcade board.
No.05805 Haze Senior Tester Feb 28, 2010, 21:34	Acknowledged, without looking at the code which decides the order of the stages it's still possible that it's something that could differ between original boards tho, chances are it will need a more in depth analysis of the game code to determine why the behavior you see in MAME differs from your PCB.
No.05808 Tafoid Administrator Feb 28, 2010, 22:31	Ultimately, it would make sense that your board be verified to be 100% the same code as currently residing in MAME. As Haze mentioned, it's not beyond the realm the possibility the code on your board might be different, even if the rom labels match. How possible is it that you can get your board's chips read/dumped and compared to MAME's?
No.05811 Haze Senior Tester Mar 1, 2010, 19:26 edited on: Mar 1, 2010, 19:26	given that all the current sets behave in the same way I think it's fairly likely that it's a MAME vs PCB difference even if it isn't known for sure that the set is the same. The main question is what determines the stage order, and if it could potentially differ even on different PCBs running the same roms. It would definitely be worth checking that none of the dipswitches affect it, but if it's just highly timing sensitive then manufacturer tolerances could mean that the behavior will differ even from one original PCB to another, or, if it'ssensitive but not THAT sensitive, it might be yet another indication that MAME needs better waitstates etc. but measuring all that is a task on it's own.
No.05823 gamestone Tester Mar 3, 2010, 17:20	I've managed to find some help from superdeadite on YouTube, he's recorded three loops of the title screen of his Gradius II GOFER no Yabou board and it's exactly the same as my board, so I would say either we both have boards that are not dumped yet or MAME has it wrong in some way. I don't have access to a rom programmer for reading the chips. Part 1 Part 2 Plus a video introducing the board The dip switch setting have no effect on the stage demo's.