| ID | Category [?] | Severity [?] | Reproducibility | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 00199 | Color/Palette | Minor | Have not tried | Jan 26, 2008, 19:54 | Apr 26, 2008, 20:56 |
| Tester | Stefan Lindberg | View Status | Public | Platform | |
| Assigned To | Resolution | Reopened | OS | ||
| Status [?] | Confirmed | Driver | |||
| Version | 0.116u3 | Fixed in Version | 0.116u4 | Build | |
| Fixed in Git Commit | Github Pull Request # | ||||
| Summary | 00199: spelunk2: On my Spelunker 2 PCB is the wall you can pass thru darker that the rest, in MAME that wall is not dark. | ||||
| Description |
On my Spelunker 2 PCB is the wall you can pass thru darker that the rest, in MAME that wall is not dark... see pic. Phil Bennett says : The sprites and characters use separate color PROMs and resistor networks. The sprite network has a 470 ohm pull-down resistor. I'm referring to pages 26 and 30 of the Kid Niki schematics (an identical/very similar boardset):http://arcarc.xmission.com/PDF_Arcade_Manuals_and_Schematics/Kid_Niki_Radical_Ninja_Conversion_Kit_(780-0052-00).pdf Stefan pointed out a while ago that Kid Niki's hair should be visible over the dark background of the title screen. In the current implementation there is no contrast. |
||||
| Steps To Reproduce | |||||
| Additional Information | |||||
| Github Commit | |||||
| Flags | |||||
| Regression Version | |||||
| Affected Sets / Systems | spelunk2 | ||||
|
Attached Files
|
 spelunk2_0116u3gre.jpg (154,430 bytes) Feb 13, 2008, 21:51 | ||||
 No.00760
nicolasa Developer
Apr 22, 2008, 09:03
|
This is still happening. |
|---|---|
|
No.00768
couriersud Developer
Apr 22, 2008, 20:37
|
From the old mametesters board: The Kid Niki screenshots are great. The borders around Niki's hair are roughly 1/2 pixel wide. Furthermore the sprites do not show any borders. Here is an example: 0: transparent "black" 1: "black" - actually a light grey 0123456789ABCDEF ---------------- 0|0000000011111111 1|0000000111111111 2|0000011111111111 3|0000111111111111 4|0001111111111111 5|0000011111111111 To display a border, it should look something like: Code: 0123456789ABCDEF ---------------- 0|0000000010000000 1|0000000100000000 2|0000010000000000 3|0000100000000000 4|0001000000000000 5|0000010000000000 I do assume that due to timing issues in the priority system the borders are created. This would happen if both guns are active for the period of half a pixel while the logic is disabling the tile prom and enabling the sprite prom: Code: Tile___CS:_1111111111000000000000001111111111111111 Sprite_CS:_0000000001111111111111111000000000000000 Effect:____1111111112111111111111112111111111111111 ____________________*______________* Unfortunately that's it for the moment. The only remedy I see would be to blow the resolution by 2 and model the effect. But that is in my opinion a bit of an overkill. |
|
No.00769
couriersud Developer
Apr 22, 2008, 20:41
|
Regarding spelunkr: If I remember it correctly, a sprite is drawn over the tilemap. a) This may be a mixing issue. Mixing is not really understood until we are able to decode the pals. Stefan made a dump, but I was not able to decode it. Eventually it is a protected PAl. b) Resistor values may slightly deviate. Resistor are allowed to deviate about 5%. Consequently there may be slight variations in color. |
|
No.00770
Haze Senior Tester
Apr 22, 2008, 20:42
|
"Unfortunately that's it for the moment. The only remedy I see would be to blow the resolution by 2 and model the effect. But that is in my opinion a bit of an overkill." Similar was done for the Galaxian stars, which are technically half a pixel in size.. It's an acceptable solution. |
|
No.00815
nicolasa Developer
Apr 25, 2008, 23:19
|
"b) Resistor values may slightly deviate. Resistor are allowed to deviate about 5%. Consequently there may be slight variations in color. " I agree with that. Since sprite and tile colors come from different resistors, the normal differences within tolerance will cause slight differences. By the looks of it, it would seem that the game designer's intention was for the secret passage to NOT be visible. |
 No.00816
Stefan Lindberg Senior Tester
Apr 25, 2008, 23:42
edited on: May 17, 2009, 19:59
|
I guess this is the cause of the "lines" seen in the third pic at the following report: 01468 I have noticed that my LDRUN3jp PCB does not have any kind of these colour differences... actually all other with it (that i noticed are spelunker 2, ldrun4/2, kidniki, kungfu master) are using a "rev C" B-board and Lode Runner 3 uses a "rev E" so i might test this newer rev on spelunker 2 and see if there is any difference. |
|
No.00826
Stefan Lindberg Senior Tester
Apr 26, 2008, 20:56
|
oops... i was too tired last night my ldrun3jp PCB do use a "rev C" B-board and so does all my other m62 games, but i did try it on Spelunker 2 anyway but there was no difference (the wall was still slightly darker). |