Ok, I did the test described in the first post on Radiant Silvergun. Looks like it jumps two frames.
You're likely seeing the benefit of some hacks to reduce lag in MAMEUIFX. ShmupMAME takes this really far, but the breakdown in emulation accuracy is controversial. Anyhow, just testing vanilla MAME I'm seeing 3 frames of lag with Radiant Silvergun, 5 frames with Cotton Boomerang. GroovyMAME looks worse in the frame-advance test, but due to the way it handles v-sync and input polling it may actually be faster in real-time. Real-time testing is required.
As the OP in that thread pointed out, how do we know this is any different than the original arcade machine?
That's a good point, it's possible that the original hardware has a good bit of lag. That would be a shame though, it shouldn't. Often, in quality shmups on arcade hardware or consoles, the ship/character will move during the very next frame after inputting a direction. Either way, it would be best if we could confirm the real figures.
Nothing in that threads describes doing this test on an actual arcade machine (if its even possible)
It isn't. To test real hardware, you need to wire an LED in series with a button, then film the button being pressed in high speed video (240fps). Advancing that video frame by frame will show you the total time between a button press (LED lights up) and a visible response on-screen. This is an all-encompassing lag measurement which includes every possible source of lag between the physical input of the buttons and the visual output on the screen.
The frame-advance method only shows how many frames of lag come from the game itself and some of the emulators processes. It doesn't encompass any of the lag from the windows host, controller encoder, LCD monitor, etc. It wouldn't change across different PC's. It's just a quick and dirty benchmark that shows some of the variation between games in MAME, and between MAME builds. It doesn't give you all the information though. A high-speed video test is needed to make a proper judgement.
The Saturn has a pretty convoluted I/O setup, wouldn't surprise me if the lag was inherent to all games on the platform to be honest, it's massively over-complicated to program compared to earlier systems (and a clear sign of things to come for later ones!) Namco had similarly complex I/O schemes and many of their games were known to be laggy on real hardware, at the time you tended to not notice because you were blown away by the graphics ;-)
Yeah, I certainly believe it's a possibility. It's just a real shame if it is.
To everyone saying, "no it's fine," just don't say there's "no lag," as that's not true, and input lag affects your gameplay whether you realize it or not. I've been playing a lot of shmups on an actual PC-Engine Duo. After getting used to that, I can feel how unresponsive a few frames of lag is just when trying to move the ship around in these games in MAME. Cotton Boomerang is reeeeeeally bad.
All that aside, back to OP's concern, it is good that you're able to run these in full speed, and yes we'd like to hear about your setup. Please note what MAME build and version you're using too.
Home
Help
Search
Login
Register
Send this topic
Print
Topic: who's playing Sega STV games at full speed on their cab? (Read 9525 times)
