Viewing Issue Advanced Details
ID Category [?] Severity [?] Reproducibility Date Submitted Last Update
04891 Color/Palette Minor Always Jun 9, 2012, 09:09 Nov 9, 2016, 04:55
Tester Kold666 View Status Public Platform MAME (Official Binary)
Assigned To Resolution Open OS
Status [?] Confirmed Driver
Version 0.146 Fixed in Version Build Normal
Fixed in Git Commit Github Pull Request #
Summary 04891: bankp: Palette problems
Description On the original pcb the palette is different from MAME.

The original flyers confirm this: http://flyers.arcade-museum.com/?page=flyer&db=videodb&id=1896&image=1

I have uploaded a video

Also the music is noticeably faster than the emulated one.
I checked the clocks and they are right.
Steps To Reproduce
Additional Information Converted original attachment video:
Github Commit
Flags Verified with Original
Regression Version
Affected Sets / Systems bankp
Attached Files
 
Relationships
There are no relationship linked to this issue.
Notes
8
User avatar
No.08726
Fujix
Administrator
Jun 9, 2012, 11:18
The music speed sounds the same to me.
User avatar
No.08727
Kold666
Developer
Jun 9, 2012, 14:07
sorry it's not noticeably, but a little bit faster.

If you have the chance to run the video in sync with mame you will notice it.
User avatar
No.08728
hap
Developer
Jun 9, 2012, 14:20
"I checked the clocks and they are right."
You'd need to measure video h/v timing for the music speed issue.

MAME has:
MCFG_SCREEN_REFRESH_RATE(60)
Exactly 60Hz, that's not the right freq.
User avatar
No.08730
Kold666
Developer
Jun 9, 2012, 18:46
it's 61hz
User avatar
No.08731
hap
Developer
Jun 9, 2012, 20:04
exact 61? :p .. and what about horizontal timing?
If we have accurate measurements of h/v timing, we can usually determine video clock, number of lines, and number of clocks(pixels) per line.
User avatar
No.08733
Kold666
Developer
Jun 10, 2012, 08:07
H= 15.61khz
V= 60,99hz

My frequency counter has an error of 0,01 hz
User avatar
No.08735
hap
Developer
Jun 10, 2012, 18:25
Thanks.
As for the colors, I think most of the difference is due to screen contrast and saturation. But it wouldn't hurt to confirm the resistors near the color prom. The ones in video/bankp.c are currently based on assumption:

---------------------------

  Bank Panic has a 32x8 palette PROM (I'm not sure whether the second 16
  bytes are used - they contain the same colors as the first 16 with only
  one different) and two 256x4 lookup table PROMs (one for charset #1, one
  for charset #2 - only the first 128 nibbles seem to be used).

  I don't know for sure how the palette PROM is connected to the RGB output,
  but it's probably the usual:

  bit 7 -- 220 ohm resistor -- BLUE
        -- 470 ohm resistor -- BLUE
        -- 220 ohm resistor -- GREEN
        -- 470 ohm resistor -- GREEN
        -- 1 kohm resistor -- GREEN
        -- 220 ohm resistor -- RED
        -- 470 ohm resistor -- RED
  bit 0 -- 1 kohm resistor -- RED
User avatar
No.08754
Kold666
Developer
Jun 11, 2012, 20:58
the only 8 resistances has the same values written in video/bankp.c

One pin of each resistance is tied to +5V, the other pins go to a 74ls07, then to a 74ls273 and to the colour prom.