shmups.system11.org

*Stands up for running at 100%*

We've already agreed that if the game is not running at 100%, then the score should not be allowed. As stated, the lack of slowdown shouldn't matter because it actually makes scoring as well as the game itself more difficult.

It's easy to assume that everything before SH3 is running at 100%, but that's always the case, and it's not always because of different PC specs. It could also be because people are using different builds of MAME.

Game has to be running at 100%, if not, score doesn't count. What's the problem?

Well, I get your point. I just don't think it would be that much of a problem if only scores from people running the game at 100% were allowed (We pretty much all agree on the point that anything <99% is not acceptable). I guess it's not that important an issue though to start any drama about it. Just beating my own Hi Scores, as even if the emulation was perfect and Hi Scores were considered legit, I probably wouldn't make it to the tables. ;p

I'll tell you this right now

Emulation of the slowdown will never be 100% accurate to the PCB, not in MAME, not in the ports, there are too many factors involved on anything past your bog standard 68k based titles.

Calling the emulation 'shit' just because the slowdowns aren't right is quite frankly insulting tho, and makes me wish I'd never bothered.

It is universally acknowledged on this forum that SH3 emulation is a good and exciting thing and I'm sure we're all very grateful for all the efforts that have been put in to make it possible as well as glad to be able to play these games in MAME.
The arguments in this thread simply revolve around whether the emulation on differing PC hardwares is consistent enough to warrant hi score boards.
This is certainly not a critiscism of your work or the work of others involved.
There's no lack of appreciation for the recent breakthroughs here

I can see the logic of both sides here but frankly I'm more inclined to think that when it comes to Cave SH3 scoreboards here the hosts should either:
A) accept Demul 0.5.7 scores when its released, as using the dynarec will give you way more consistency in 100% speeds on computers, while rejecting MAME stuff due to performance
B) reject the lot outright, as the inconsistency of the emulation setup is a headache (or you're an elitist fuck who thinks PCBs are the only way to play )
C) accept Slowpoke MAME scores for games which have their fillrates close to exact.

For example, I see nothing wrong with submitting a Deathsmiles MAME score at fillrate set to 63% because as far as I can tell it behaves exactly as the port which regardless of whether the port is accurate is accepted in score threads. But since no one seems to have an idea of what's exact for the other games (and no one seems to be able to agree on an exact fillrate for Futari let alone anything else) this may be more trouble than it's worth, so it'd be understandable for a score thread to just not accept MAME scores outright. I think if Demul got a release the issue of performance on people's machines would be nullified and then the only issue left would be the slowdown which has already been addressed in this thread - agree on a fillrate for a game or leave fillrate default.

No ones calling the emulation shit? i did in an example to make the argument work its not literal though.

Wow just wow , lets just leave it as you see it your way i see it mine especially as ive addressed many points you just looked past .Do you actually think for the people with powerful enough pc,s that the emulation is working differently ? lol ok then.How about the people who overclock in mame to remove all slowdown from games anyway? and you ignored my garegga example some people play for score in FBA which removes all slowdown it never comes under scrutiny its all just bogged together with mame players.Hwl is spot on people use different mames different machines different controls etcetc so being obsessive over some slowdown is a joke at best especially as it makes it harder to do well anyway .

(double post removed /nZ)

Nah, no poll !
Performance discrepencies between players should be enough to void everything in favor of MAME SH3 scoreboards. (not to mention that slowpoke build which lets everyone slow down the game if they want !)

Slowdown is available in many builds of mame for many years via underclocking, and you are being totally ignorant to the purpose of the blitter rate control on slowpoke mame .

Yes it's clear that people's opininons are polarised here and that no amount of arguing is going to change anyone's minds either way.
I'm sure we're all having fun playing these games and some of us are having fun arguing about them too but, as history often shows, those with the loudest voices aren't always right
At least a poll would put this shit to bed...

I dont care for a poll but agreed on the rest.

I'm not being ignorant, not at all, much to the contrary I'm aware that people can tweak the slowpoke build to whatever needs they have, for ANOTHER WAY to slow down the game. Even if it's not the intended purpose, people WILL still do it whether you like it or not.

The point is that people can tweak any build for any system. It may take slightly more effort than cheating with the current slowpoke build, but it's still not too hard. Also, any provided .inp files with a blitter rate change are rather unlikely to sync; forcing submission of a .inp would help preventing the problem of people underclocking the blitter.

Having that said, I'll just sidestep the current issue entirely and say that it'd be a good idea to respect Cave's decision to get the drivers removed for now. Leave the games playable by those who care to find them (and I'll be honest, that includes me as much as everyone else) but don't encourage people to play them at this point.

Have you guys tried GroovyMAME?, then you'll get true 100% speeds

Link please .

Oh, btw: I got my 360+Futari now, and I'll tell you this: It's about 10 times easier on 360 than emulated on slowpoke build with 100% speed. There's a LOT of slowdown missing in MAME; especially stage 3 is way easier on Xbox.

I know right , and 1.5 easy enough as is without any slowdown I really dont think that game should have any . Im loving 1.0 original imo they dumbed down 1.5 way to much in stages 1 2 and 3 .

http://mario.groovy.org/GroovyMame/

Someone's going to need to explain this to me. It looks like it needs a disc image mounted at the same time to run correctly? "Building switchres" into it? Installing drivers? Might be Linux only?

GroovyMAME won't help at all. It's solely for removing tearing and jitter, not poorly emulated blitters.

Yeah, 1.0 Original is a great game and much more fun survival challenge than 1.5 Original. Sucks that not many people have played it.

I don't think the emulation is "shit" by any means. It's quite good even if the slowdowns may differ. I'm pleased with the progress of the MAME driver so far, and this comes from one who owns an actual Ibara PCB. Also... people who complain need to remember they paid nothing for it. Those who call the emulation "shit" need to STFU or take their own shot at writing a CV-1000 emulator.

No one did call it shit it was a misunderstanding .

See here for the windows version, http://mame.groovy.org/WindowsATIDrivers/. Ideally you want to be using an older ATI card below 5 series, custom drivers for said ATI card and CRT display...but i have been using it on my LCD TV with an Nvidia card combined with powerstrip (a new feature added to GroovyMAME) and get suber results.

Look at the Groovymame forum over at BYOAC forum for more info on how to set it up........http://forum.arcadecontrols.com/index.php?board=52.0

Blurb on GroovyMAME......


* Generate custom modelines and use them as game calls for them
* In Windows with ATI cards we can alter the refresh rate of existing modelines for game requirements
* Resolution change capability with modeline switching in Windows and Linux, PSX games and others
* Multithreaded mode and waitvsync work together in Windows without throttle
* MKChamp hi score patch compatible/ Works with Linux too (hiscore.dat goes in the \hi\ directory)
* Froger/Galaxian resolution fixes for Windows and Linux (so they look normal for arcade resolutions)
* Sound sync for Windows (not in Linux) triplebuffer, capable of being turned off (default)
* Clean stretch both Windows and Linux
* Redraw frames so 30Hz games run at 60Hz like Tron in Windows and Linux
* Most settings and features are automatically set as needed depending on the resolution used,
like if throttling is necessary, or can use vsync instead, or fall back to triplebuffer.
* In Windows can use ArcadeVGA 3000 cards or others without any .ini files, picks best resolution automatically

Notes:
Always start with a fresh mame.ini file generated from groovymame, the defaults are
the best for modeline generation and different from normal mame or cabmame or any other
mame.

ATI cards, mostly 9200/9250 and HD2xxx and above cards should be used. In Linux
anything besides the X8xx series should work and in Windows your limited only by
ATI cards that work with Soft15khz (Since we use the same registry custom modelines).

In Linux possibly other cards work, it just depends on if xrandr can setup custom
modelines and the card can handle vsync interrupts properly. Any testing results of
stray cards are welcome, reports are helpful in getting more cards working in the future.

Calamity has custom ATI drivers, 32 and 64 bit, which contain preset custom modelines to
work best with groovymame. That way you don't need Soft15khz unless you want to add more
custom modelines, his drivers have the ability to store close to 120 modelines and that
is the limit (normally only 60 on regular catalyst drivers).