Topic: FB Alpha 0.2.97.43 Bug Reports

[barbudreadmon]

Hi,

it seems there is a color issue in dkong, here is the original pcb :


current mame is close to those colors, i guess it is related to this change :

Code: [Select]

0.141u2: Couriersud fixed a bug in palette_normalize_range which caused color distortion. The luminance normalization now converts r,g,b to y,u,v and normalizes y prior to converting back to r,g,b. This affects e.g. Radar Scope, Donkey Kong and Mario Bros.
fba is a lot more redish.

I'm not 100% sure but i guess it is related to one (or both) of those changes in mame :

Code: [Select]

diff -Naur mame-mame0141u1/src/emu/video/resnet.c mame-mame0141u2/src/emu/video/resnet.c
--- mame-mame0141u1/src/emu/video/resnet.c 2011-01-24 06:09:05.000000000 +0100
+++ mame-mame0141u2/src/emu/video/resnet.c 2011-02-09 16:01:01.000000000 +0100
@@ -564,7 +564,7 @@
  cut = 0.0;
  break;
  case RES_NET_AMP_DARLINGTON:
- minout = 0.9;
+ minout = 0.7;
  cut = 0.0;
  break;
  case RES_NET_AMP_EMITTER:
@@ -685,13 +685,15 @@
  v = vcc - v;
  v = MAX(0, v-0.7);
  v = MIN(v, vcc - 2 * 0.7);
+ v = v / (vcc-1.4);
+ v = v * vcc;
  break;
  case RES_NET_MONITOR_ELECTROHOME_G07:
  /* Nothing */
  break;
  }
 
- return (int) (v *255 / vcc + 0.4);
+ return (int) (v * 255 / vcc + 0.4);
 }
 
 rgb_t *compute_res_net_all(running_machine *machine, const UINT8 *prom, const res_net_decode_info *rdi, const res_net_info *di)
diff -Naur mame-mame0141u1/src/lib/util/palette.c mame-mame0141u2/src/lib/util/palette.c
--- mame-mame0141u1/src/lib/util/palette.c 2011-01-24 06:09:05.000000000 +0100
+++ mame-mame0141u2/src/lib/util/palette.c 2011-02-09 16:01:01.000000000 +0100
@@ -682,8 +682,8 @@
 
 void palette_normalize_range(palette_t *palette, UINT32 start, UINT32 end, int lum_min, int lum_max)
 {
- UINT32 ymin = 1000 * 255, ymax = 0;
- UINT32 tmin, tmax;
+ INT32 ymin = 1000 * 255, ymax = 0;
+ INT32 tmin, tmax;
  UINT32 index;
 
  /* clamp within range */
@@ -707,11 +707,13 @@
  for (index = start; index <= end; index++)
  {
  rgb_t rgb = palette->entry_color[index];
- UINT32 y = 299 * RGB_RED(rgb) + 587 * RGB_GREEN(rgb) + 114 * RGB_BLUE(rgb);
- UINT32 target = tmin + ((y - ymin) * (tmax - tmin + 1)) / (ymax - ymin);
- UINT8 r = (y == 0) ? 0 : rgb_clamp(RGB_RED(rgb) * 1000 * target / y);
- UINT8 g = (y == 0) ? 0 : rgb_clamp(RGB_GREEN(rgb) * 1000 * target / y);
- UINT8 b = (y == 0) ? 0 : rgb_clamp(RGB_BLUE(rgb) * 1000 * target / y);
+ INT32 y = 299 * RGB_RED(rgb) + 587 * RGB_GREEN(rgb) + 114 * RGB_BLUE(rgb);
+ INT32 u = ((INT32)RGB_BLUE(rgb)-y /1000)*492 / 1000;
+ INT32 v = ((INT32)RGB_RED(rgb)-y / 1000)*877 / 1000;
+ INT32 target = tmin + ((y - ymin) * (tmax - tmin + 1)) / (ymax - ymin);
+ UINT8 r = rgb_clamp(target + 1140 * v / 1000);
+ UINT8 g = rgb_clamp(target -  395 * u / 1000 - 581 * v / 1000);
+ UINT8 b = rgb_clamp(target + 2032 * u / 1000);
  palette_entry_set_color(palette, index, MAKE_RGB(r, g, b));
  }
 }
Do you think this change is somehow appliable to fba ?

Best regards .

[gxb]

It seems there's an input issue in Sangokushi III Gaiden: Kakou-On's Revenge DX (a hack of wofch, and its ROM name is wofchdx). This game is unplayable since START button does not work. Is it a known issue?

[dink]

Hi,
wofch and wofchdx will play fine: but its a weird game, you have to wait until the game starts demo play before you can start.  I think because its a CPS Changer game, so its on FreePlay mode. 

best regards,
- dink

[barbudreadmon]

Hi,

I found 3 new bugs (tested all of them in fba standalone) :

1. P2 can't join in monster maulers and its clone

2. Most games in Taito "The NewZealand Story" driver are broken (fba crash). Actually i noticed something really weird : "kabukiz" is not present in the fba arcade dat file while the game is present and marked as working in the driver.

3. In megadrive's general chaos, it seems multitap is detected, but either i don't understand how to move P2/P3/P4, or their controls are broken

[dink]

Hi,

I found 3 new bugs (tested all of them in fba standalone) :

1. P2 can't join in monster maulers and its clone

2. Most games in Taito "The NewZealand Story" driver are broken (fba crash). Actually i noticed something really weird : "kabukiz" is not present in the fba arcade dat file while the game is present and marked as working in the driver.

3. In megadrive's general chaos, it seems multitap is detected, but either i don't understand how to move P2/P3/P4, or their controls are broken

Thanks for the reports!  I fixed up tnzs a bit, but need to do some quick testing tonight when I get home. 

General Chaos, select 4 players from the title screen, then theres a tricky part about squad selection, I forget how to get past that other than random button presses    But once the game finally starts I can control all 4 players.

best regards,
- dink

[Barry Harris]

Actually i noticed something really weird : "kabukiz" is not present in the fba arcade dat file while the game is present and marked as working in the driver.

kabukiz is only in debug builds (both for the game list and the dat).

[barbudreadmon]

kabukiz is only in debug builds (both for the game list and the dat).

Thanks for the information, i didn't think about checking gamelist.txt, is there any way to know this when looking at

Code: [Select]

struct BurnDriverD BurnDrvKabukiz = {
"kabukiz", NULL, NULL, NULL, "1988",
"Kabuki-Z (World)\0", NULL, "Taito Corporation Japan", "Miscellaneous",
NULL, NULL, NULL, NULL,
BDF_GAME_WORKING | BDF_HISCORE_SUPPORTED, 2, HARDWARE_TAITO_MISC, GBF_SCRFIGHT, 0,
NULL, kabukizRomInfo, kabukizRomName, NULL, NULL, CommonInputInfo, KabukizDIPInfo,
Type2Init, DrvExit, DrvFrame, DrvDraw, DrvScan, &DrvRecalc, 0x200,
256, 224, 4, 3
};?

[iq_132]

BurnDriver - Normal
BurnDriverD - Debug only
BurnDriverX - Don't show ever

[dink]

Wow, I'm a dummy!  I fixed / made Kabuki-Z playable waaay back in .37.  And it's been BurndriverD'd this whole time.  DERP!
I remember being super happy about finally getting that game working back then, too.. hehe.. oh well - sorry guys!

best regards,
- dink

[iq_132]

We should run a burndriver check before releases. This isn't the first time that this has happened

[dink]

We should run a burndriver check before releases. This isn't the first time that this has happened

As long as I get credit for the last time it happened!

[dink]

Regarding VS Block Breaker issue with the hamster-paddle tile covering up part of the titlescreen:
This was finally fixed tonight.  It was quite perplexing, and I'm happy to finally fix it.  On a side note,
another bug got fixed that "covered up" the flashing red or yellow that happens when the hamster
gets struck with the paddle in the intro.  yay!

best regards,
- dink

This one almost drove me crazy as i wasn't able to reproduce it and was starting to wonder if i dreamed it, but i was finally able to reproduce it.

[IPLA]

Hi,

Contra color issues

[Neildark]

Hello, first of all thanks for your work; has found a couple of errors: in the Wizard Fire (Over Sea v2.1) in the second and last level (attached screenshot) the clouds prevent to see the stage (In MAME there are transparencies)
In Shadow Dancer (set 3, US) in phase 2-4 the Final Burn Alpha has an error and closes
I copy what the log tells me
Thanks for everything. A greeting!
Exception 0xC0000005 (EXCEPTION_ACCESS_VIOLATION) thrown.
EIP: 0x00C8
 (attempting to write address 0x77CCF456)

EAX: 0x00000000, EBX: 0x00000000, ECX: 0x00000000, EDX: 0x77CCF456
ESI: 0xFFFF8000, EDI: 0x004EC8F0, ESP: 0x001BDA30, EBP: 0x7A120000

[dink]

Neildark, hi, thanks for the bug reports!
Do you have savestates for Wizard Fire cloud level and Shadow Dancer from just before the crash?

best regards,
- dink