forums.marpirc.net

If you push F2 and go into the I/O Test, you can see that pausing, holding a button and advancing frame by frame will show that 106 has 1 frame of input lag, but 99 activates on the next frame. The 4 additional frames that are present in gameplay is Capcom's own inherent lag, they do this in a lot of games. I remember this number coming up when some people in the fighting game community were doing tests on how much lag is present in the home ports of Super Street Fighter II Turbo, and those same 3-4 frames of lag were there in the home ports too.

Regardless, thanks for allowing 0.99. One extra frame might not seem like much but if given the choice, why would anybody want to play on a version with any additional input lag whatsoever? I was originally going to ask if 0.99 could be allowed for each game of the tournament when round 1 started, but as it turns out Bubble Memories had tons of slowdown in 0.99 compared to 0.106 so it wouldn't have been wise to use it there. It would be nice if the rest of the games are checked for 0.99, though

wouldn't using an old version constitute cheating then?

you're getting a 1 frame advantage over other players, which probably didn't exist in the actual arcade.

(it's more likely MAME 0.99 had broken sprite buffering, which was later fixed, adding the additional delay you see)

The CPS drivers in general have seen a lot of improvements since .106 even tho...

As I mentioned before, there is no lag in 0.140.

Haze wrote:wouldn't using an old version constitute cheating then?

you're getting a 1 frame advantage over other players, which probably didn't exist in the actual arcade.

(it's more likely MAME 0.99 had broken sprite buffering, which was later fixed, adding the additional delay you see)

The CPS drivers in general have seen a lot of improvements since .106 even tho...

MAME adds input lag

not the other way around

shoesama wrote:MAME adds input lag

not the other way around

You mean sprite or layer buffering? Not an unusual feature in arcade games.

Some knowledgeable people told me that OS overhead (and peripheral latency) add even more lag. With sprite/layer buffering removed, the lag would be more accurate to what the actual PCB has.

Haze wrote:wouldn't using an old version constitute cheating then?

you're getting a 1 frame advantage over other players, which probably didn't exist in the actual arcade.

(it's more likely MAME 0.99 had broken sprite buffering, which was later fixed, adding the additional delay you see)

The CPS drivers in general have seen a lot of improvements since .106 even tho...

People have done tests that show that even on the arcade PCB of SSF2T, the game animates 3 frames after the button is pressed. If you play SSF2T on MAME 0.99 and press a button, the animation also starts 3 frames later. If you play it on 0.106, the animation starts 4 frames later instead. How is it "cheating" to use a version of MAME that matches the original hardware in terms of input lag?

mahlemiut wrote:As I mentioned before, there is no lag in 0.140.

The input lag problem has been fixed for a long time, it was somewhere in the 0.12x's that it was fixed I think. Which is great and all, but doesn't change the fact that 0.106 still has additional input lag, and it's often considered the go-to version of WolfMAME for tournaments (and is the only accepted MAME version by TG). So I really hope that 0.99 will be alright to use for the remainder of the tournament, assuming there's no emulation differences with it like there is with Bubble Memories. At least please consider it, coordinators

Why not using wolf0.140 ?
The problem with wolf99 is that I can't playback inputs (I've got an "access violation" message) so I can't playback for example BBh's first place

Sometimes coins gives you a X2 score when you take them; does anyone say the right way to do it?

It's random or coins must be taken in a specifical way (like chains in Gunbird) ?

mahlemiut wrote:

shoesama wrote:MAME adds input lag

not the other way around

You mean sprite or layer buffering? Not an unusual feature in arcade games.

And OS and monitor latency isn't exactly "MAME"-related ...

shoesama wrote:

Haze wrote:wouldn't using an old version constitute cheating then?

you're getting a 1 frame advantage over other players, which probably didn't exist in the actual arcade.

(it's more likely MAME 0.99 had broken sprite buffering, which was later fixed, adding the additional delay you see)

The CPS drivers in general have seen a lot of improvements since .106 even tho...

MAME adds input lag

not the other way around

MAME takes away lag if you don't emulate things properly, and hack up the video hardware so that sprite buffers don't work as they should; see the abomination that is the 'lag-free shumps build'

When a system is emulated properly MAME has the same amount of lag as the actual arcade system.

BBH wrote:

Haze wrote: At least please consider it, coordinators

From now, and fot the rest of this tournament, is possible choose 0.99 or 0.106 (optional) for all games.

No 140 ? so I've choice between 0.106 with the lag, or 99 which doesn't work with Windows 7

Phil Lamat wrote:No 140 ? so I've choice between 0.106 with the lag, or 99 which doesn't work with Windows 7

Really? I was just about to upgrade and heard there were no problems with MAME on Win7. Did you try XP mode?

The_Pro wrote:

Phil Lamat wrote:No 140 ? so I've choice between 0.106 with the lag, or 99 which doesn't work with Windows 7

Really? I was just about to upgrade and heard there were no problems with MAME on Win7. Did you try XP mode?

XP mode is for business software, not games. It 'emulates' a basic S3 graphic card, the performance on any gaming software is abysmal, and you'll get frameskipping and lag from the XP emulation.

Current versions of MAME run fine on Windows 7. Mame 0.99 is from the middle of 2005, that's over 5 years old and things have changed a lot since then. When using such old software you have no guarantee that it will work as expected anymore.

I strongly recommend that current versions are allowed in these tournaments unless there are significant proven reasons to disallow them (long-term broken drivers etc.) As I said, the emulation of many systems has improved a lot in the last 5 years. You're doing yourself a disservice by not allowing them, even if some users may prefer to use the older ones for performance reasons.