08912: sgx [daimakai]: Clipping (Window in huc6202) in game ending is broken

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
08912	Graphics	Minor	Always	Sep 5, 2024, 05:10	Oct 2, 2024, 13:26

Tester	dink	View Status	Public	Platform	MAME (Official Binary)
Assigned To		Resolution	Open	OS	Windows Vista/7/8 (64-bit)
Status [?]	Confirmed			Driver	nec/pce.cpp
Version	0.269	Fixed in Version		Build	64-bit
		Fixed in Git Commit		Github Pull Request #

Summary	08912: sgx [daimakai]: Clipping (Window in huc6202) in game ending is broken
Description	The ending squeezes the sides of the screen, but, something went wrong? Note; each side is supposed to be squeezed by the same amount of pixels, so that the view is in the center of the screen.
Steps To Reproduce	Play 2-all, watch ending
Additional Information	seriously, just play the game; it's really easy, and fun. Then watch the ending after beating king lucifer. ... or just download attached state file, 0.269-compatible
Github Commit

Flags
Regression Version
Affected Sets / Systems	sgx [daimakai]

Attached Files	dmk_window.png (81,666 bytes) Sep 5, 2024, 05:10 Uploaded by dink
	gab75_daimakai_end.zip (59,039 bytes) Sep 5, 2024, 05:14 Uploaded by dink ending state for MAME 0.269, c/o my good friend Gab75
	dmk_probably_ok.png (13,051 bytes) Sep 5, 2024, 13:24 Uploaded by dink assumed this is how it's supposed to look, (a few frames later than previous screenshot - w/text)

No.22305 Robbbert Senior Tester Sep 5, 2024, 09:17	Used the save state. Also noticed that the text on the left side is truncated when the screen narrows.
No.22352 dink Tester Oct 2, 2024, 13:26	in devices/video/huc6202.cpp, function: next_pixel() It's setting up a priority bitmap based on the window arrangements 0..511, but, huc6270 was updated to use 1024 (or 1024+64? +64?) instead of 512. The window value, which is 0-512 needs to be scaled to the new pixel size, this is how I do it on my side: int win1_adj = m_window1 * 1024 / 512; int win2_adj = m_window2 * 1024 / 512; if ( m_window1 < 0x40 \|\| i > win1_adj ) { m_prio_map [ i ] \|= 1; } if ( m_window2 < 0x40 \|\| i > win2_adj ) { m_prio_map[ i ] \|= 2; } note: my window size is 1024, but, I think MAME is using 1024+64 or 64+1024+64, (look at the set_raw in mame/nec/pce.cpp) note2: 0x40 is a special value, so m_window1, m_window2 needs to be tested for this, not the new adjusted value. best regards, - dink

No.22305

Robbbert

Senior Tester

Sep 5, 2024, 09:17

Used the save state.

Also noticed that the text on the left side is truncated when the screen narrows.

No.22352

dink

Tester

Oct 2, 2024, 13:26

in devices/video/huc6202.cpp, function: next_pixel()

It's setting up a priority bitmap based on the window arrangements 0..511, but, huc6270 was updated to use 1024 (or 1024+64? +64?) instead of 512.
The window value, which is 0-512 needs to be scaled to the new pixel size, this is how I do it on my side:

int win1_adj = m_window1 * 1024 / 512;
int win2_adj = m_window2 * 1024 / 512;

if ( m_window1 < 0x40 || i > win1_adj )
{
m_prio_map [ i ] |= 1;
}
if ( m_window2 < 0x40 || i > win2_adj )
{
m_prio_map[ i ] |= 2;
}

note: my window size is 1024, but, I think MAME is using 1024+64 or 64+1024+64, (look at the set_raw in mame/nec/pce.cpp)
note2: 0x40 is a special value, so m_window1, m_window2 needs to be tested for this, not the new adjusted value.

best regards,
- dink