Helifire wave effect

[mahoneyt944]

@arcadez2003 There's still a bug with the wave effect which exists in mame2003 and mame2010 too. Current mame looks much smoother. Notice the stepped waves in the sky and stepped water color.

This is where the wave effect is added here.

mame2003-plus-libretro/src/drivers/n8080.c

Line 422 in 62e9e7a

int level = 120 + wave[helifire_mv & 7];

2003+

Current

[arcadez2003]

Looking at the commits i see no major changes in the graphical dept for HeliFire directly, of course there
are tons of corewide improvements since 2010 and anyone of them could have sorted the wave affect
you described above.

However this change might be worth a look as CPU timing quite often affects the gfx

arcadez2003/mame@638032f#diff-d93951d6c145ef9df54cb00f57379c0eca34848a071bcfd1833d60de1912b095

Edit never mind the above would require a major rewrite of the I8085 CPU core which i'd hate to undertake
only to find out it's not those changes we need :)

[mahoneyt944]

Hmm well it might just be how it is here then. I also have an audio issue open in current mame since it's not exactly right either. I'm not going to link it here for obvious reasons, But the issue is the same in both cores.

[arcadez2003]

Aye not a bad idea but they'll know who you are as they like to keep track of who is working on this
horrible nasty nearly 20 year old core :)

[mahoneyt944]

Well it is an issue with their code as well since we are using the same code. Either way it should be known.

[arcadez2003]

Sure i hear ya maybe they'll not say anything to ya, if it was me they'd defo shoot the messenger ha ha
but yeah you've spotted a bug so well worth sharing the info around it.

[mahoneyt944]

Hey @grant2258 I'm about to just hang it up on this issue since I really don't know what would cause such an effect besides something outside of the driver itself such as graphic rendering methods or something. Any ideas or theories on this before just closing it out?

Shadowing effect or blending ? Idk

[ghost]

I haven't even looked at any of the code. Make sure the version of mame your getting the commits from actually works properly.

[mahoneyt944]

It was a graphics anomaly for a long time in mame, seems to be something in how things are drawn in general and not the actual code in the driver. It follows current mame very closely.

[ghost]

when did it get fixed if it was like that a long time.

[mahoneyt944]

I'm not sure exactly, after mame2010.

[ghost]

All I can see off hand is the sound cpu is a different speed. I would test if the commit arcade posed fixes this as well might save you some time. I would need to port this over to my old core to test though will look a little more when I get the time its not really a deal breaker anyway

[ghost]

This is odd i tried mame0139 standalone and it works fine here is a screenshot. No idea why its doing this in mame2010

[mahoneyt944]

Hmm, that is odd

[arcadez2003]

@mahoneyt944 i can port this across to my MAME72 test core it'll take about an hour to set everything up
then at least after that we'll know if this is an issue on the MAME core side or the libretro side, i might as well
ask you now what does the graphical wave effect actually do so i know what to look out for.

[ghost]

@arcadez2003 our core is filling up in gradients. rather than smooth changes. I tested standalone mame095 and is still had the effect we are getting.I The odd thinks is mame0139 standalone is working properly and mame2010 (based off mame 0139 had this oddity as well).

[arcadez2003]

Right no worries so there must be a fix sometime between MAME95 and 139 i'll look into it pinning it down sometime
if you guys dont beat me to it, back to Zelda then :)

[mahoneyt944]

Maybe we can log "level" and "wave" in .139 and mame2010 and compare outputs? I'm not entirely convinced it's the wave because you can comment out the wave effect and you still get the stepped color changes.

The only difference in our video code is the rom loading. For some reason with our old drivers, it prefers the old loading values. The color completely breaks with the new values.

[ghost]

Problems is I dont have windows some mames dont run on wine so cant really back trace what version actually fixed it.

[mahoneyt944]

I've studied the game driver and video driver pretty tightly against a dozen newer mames including current. I don't think the fix is there......Though, when I finally got the color to work in this game the fix was the rom loading values. Our code prefers the old values which differ from newer mames. I'm assuming this is something on the machine level. I suppose this might be a place of interest.

[ghost]

Well the best solution is to see where it works then see whats different to have a reference to see whats changed to everything related to the driver and teh hardware the driver uses.