Main > Driving & Racing Cabinets

Saga Rally M2 FFB (bug identified)

(1/5) > >>

GameOverYeah:
Hey all,
Ive been busy building a cab over the last few months and have done some intensive testing on the FFB coming out of M2 and into my where base!

After a few phone calls to a store that has an old (working) OG SRC unit and many sleepless / late nights testing a myriad of settings, I'm pretty confident I've identified why the SR FFB in M2 (potentially Daytona etc to) is so poor on the Model 2M@ emulator.

All my testing has lead me to identify that the FFB coming into the base from the emulator (particularly in SRC) is inverted/reversed on 1 channel!

You can test this yourself by doing the following:
Set all FFB settings to the M2 defaults (for consistency but feel free to use your own)
Open the test menu and turn FFB to max on the ROM menu. (again for consistency but feel free to leave this at default)
Turn the FFB to max on your wheel/wheel software. (for consistancy, but feel free to leave it as-is)

Start the game.
Do not move the vehicle!
Turn the wheel left and right and you will notice that the wheel rumbles, jerks, judders as if you are bouncing over a rough surface (violently)

NOW!
begin to move the vehicle.
As soon as you fo, you will notice that as soon as you move, the wheel gets lighter and the FFB stops!
 
ALL FFB / RUMBLE / JERK / JOLTS DISAPPEAR as soon as the vehicle begins to MOVE but is present when the vehicle is stationary!

I have called the store/arcade with the OG SRC Cabinet and asked if they would test this for me, and i can confirm that the the CAB works in the exact opposite manner to how I described!
The CAB has a light feedback when the vehicle is stationary and then rumbles and shakes etc when you begin to move the vehicle!

Therefore it would appear that the game/emulator is giving users 100 max FFB rumble/gravel effect when the vehcile is actually stationary, and gradually reducing this effect as soon as you start moving!

Which is the polar opposite to what it actually should be doing!

Now I can't seem to correct this using the settings provided in the .ini and i have tried numerous settings from here , and have tried a myriad of my own settings from every number and figure conceivable within the parameters (including negative numbers)

Any questions just ask!
I've tested with a myriad of builds, settings, and my cabs are using a mixture of OSW, CSWv2.5 and CSL Elite bases.
I've tried a combination of FFB settings ranging from defaults to virtually every config you can imagine, including the ones posted here and in other forums!

However, its NOT possible to fix the using the emulator or our FFB profiler etc.

Hopefully, now the issue has been identified we can perhaps get a few minds together and see if we can fix it!


GameOverYeah:
Quick Update!

I have been to the north east today and (about 1hr away) and tested SRC Cabinet in the arcade myself :)
I can confirm that the actual cabinet works in the opposite way in which M2 deleivers the rumble effect etc which is also affecting other FFB effects.

I sopke in depth to the arcade owner and he too has had issues with cabinets he as built for personal use.

I also tested a huge array of settings listed here, but noticed that many of them are flawed, broken or unusable.
For example:
The settings found here (http://forum.arcadecontrols.com/index.php/topic,130376) posted by user Jollywest have a serious anomaly where upon the L/R forces have a disparity in the negative range:
FE_LEFT_Gain=0.5      ;Global gain
FE_RIGHT_Gain=-0.5      ;Global gain

Not only does this go against the direct instructions and user guide posted by M2 it also goes directly against the recommendations of the HTML guide posted here (http://forum.arcadecontrols.com/index.php?topic=131527.0)

After trying these my OSW / CSL / CSWv2.5 all went absolutely crazy and began banging/pulling themselves violently against one side! I certainly don't recommend these settings for home use!

There are also settings here and all over the forums advising users to run settings higher than the EMU can actually present? This is also incredibly bad, dangerous and pointless.
you can see examples of this here (http://forum.arcadecontrols.com/index.php?topic=134688.0)
FE_UNCENTERING_Gain=10.0      ;Global gain


And here where someone suggests running gain at 10000 (TEN THOUSAND) :-0 ???
The MAX the emulator actually handles (and recognizes) is a figure of 5.0 and even then is solid clipping!
As you can see here (http://forum.arcadecontrols.com/index.php?topic=130376.0) this advice is not helping people identify and rectify these FFB issues.
FE_CENTERING_Gain=10000      ;Global gain

After a lot of testing (months trying to figure this out) and finally today getting to use an actual SRC Cabinet i can positively say that the FFB posted here are all incorrect, have invalid numbers, incorrect settings and are all very very very bad advice!

I do have a set of figures that i believe mitigate the issue, but there is nothing we can do to get the correct center/rumble/ffb shake when driving on straights as this is a fundamental issue with the emulator! 

Either the emulator is sending the FFB inverted, or the dip switches are required to invert the signal... OR because we are inverting our accelerator and brake inputs etc this could be affecting the FFB. for example, is entirely possible that this effect is purely canned (i think it is from testing) and was/is activated when the acc pedal is depressed (but confusion arises when the pedal is inverted perhaps? im not sure!) but overall that signal in indeed inverted and we need to reverse how it works!

I am currently testing a middle man FFB software that may resolve the issue. If anyone else is interested in sorting the FFB for all users and providing a solution feel free to join in :)

Boomslang:
Hm i can properly look at Sega rally on m2 today and make an alternative ffb plugin to try instead of using m2 ffb

Is it just Sega rally or other games too?

Boomslang:
Try this out and see how it compares


One thing i found wierd was hitting walls rotates the wheel opposite to most games, but I noticed the original m2 ffb reacts the same so I guess it's normal


Just copy files into m2 emulator folder (version 1.1a) and turn off EnableFF in EMULATOR.ini to disable the M2 FFB

load emulator_multicpu.exe and then close after a few seconds and check log.txt for your device GUID and copy to FFBPlugin.ini

Turn off Logging in FFBPlugin.ini now and load Sega Rally Championship and try it out

(ive set Min Force=20 in ini atm, change this to modify the strength of lower values. Generally this would be set to 0 but I raised it as strength felt a little low)

GameOverYeah:

--- Quote from: Boomslang on August 18, 2018, 04:51:43 pm ---Hm i can properly look at Sega rally on m2 today and make an alternative ffb plugin to try instead of using m2 ffb

Is it just Sega rally or other games too?

--- End quote ---

Seat Rally is the title I have been focusing in primarily as I'm building a replica cabinet (art work et all) however it's fair to say that the FFB in Daytona etc is also poor.

It's perfectly clear (and very easy to reproduce) to see where the FFB goes wrong on the M2 version of Sega Rally, my issue is I don't know how it can be fixed easily, because as soon as the user accelerates, the FFB output on that rumble Channel is cut... entirely!

It's very easy to test, reproduce, replicate and verify this issue, and it's clear to see the FFB is indeed functioning incorrectly on that specific channel.

The rumble, road texture effect is sublime when the vehicle is stationary, it's perfect (if you rotate the wheel) but as soon as you move the vehicle forward with the accelerator, you loose all road texture and rumble! The effect is CUT entirely from the game. (it should be OFF when stationary and ON when moving like the actual cabinet)

The FFB signal is active when stationary BUT inactive/deactivated when moving

So in short, if you are able to correct Sega Rallys FFB, I'm sure the community would love that, I know I would!

I've been planning this build for a few years now, and was incredibly disappointed to see the cabinet would have to be built with junk ffb due to the error in M2's FFB code, or an error in the dump.

It's a shame there is no other Model 2 emus, 1 line of code in the emulator that allows users to add a line to invert the on/off function to the config ini could solve this entirely!

if we could just invert / reverse the WAY in which the wave/rumble effect works (not its polarity/phase etc or even its effects etc) as It's Not phase related, but just reverse HOW/WHY/WHEN the effect is actually enabled / disabled as currently is backwards.

Navigation

[0] Message Index

[#] Next page

Go to full version