MAMEdev
Replies: 75 comments 135 replies
-
|
Windows 20H2, using Verbose output:
|
Beta Was this translation helpful? Give feedback.
-
|
F310 axis and trigger codes and ranges:
Dual Action axis and trigger codes:
|
Beta Was this translation helpful? Give feedback.
-
|
Windows 20H2, SDL 2.0.14, using Verbose output:
|
Beta Was this translation helpful? Give feedback.
-
|
F310 axis and trigger codes and ranges:
Dual Action axis and trigger codes:
|
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
|
Beta Was this translation helpful? Give feedback.
-
|
Thanks very much.
That’s normal when using edit: Never mind, I see you already tested it. |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
Xbox Guide button is recognized in Mame but not the share button. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks again. It looks like it’s correctly identified the controller type, but for some reason it has no mapping for the share button. The mappings, when reorganised a bit:
|
Beta Was this translation helpful? Give feedback.
-
|
sdlmame -version No visible changes or regression compared to yesterday |
Beta Was this translation helpful? Give feedback.
-
|
Thanks, but you don’t need to re-test all your controllers, only the ones I called out specifically. I don’t want to waste your time. |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
Strange labels Special tokens ZL and ZR are reported as sliders but they instantly go from 0 to -65536 i.e. they act as buttons and are not analogue at all
|
Beta Was this translation helpful? Give feedback.
-
|
As far as I can see on forums all triggers on switch's gamepads (joycon included) are digital. |
Beta Was this translation helpful? Give feedback.
-
|
I'll also test with joycons but probably tomorrow as the battery is completely flat |
Beta Was this translation helpful? Give feedback.
-
No rush, it’s great info so far. |
Beta Was this translation helpful? Give feedback.
-
|
sdlmame -version
Successfully tested disconnection and reconnection |
Beta Was this translation helpful? Give feedback.
-
|
Great! Thanks for re-testing. |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using Verbose output:
Driver is reporting button 14 & 15 but they aren't physically present on the pad. |
Beta Was this translation helpful? Give feedback.
-
|
What are the tokens for the right stick axes? Is it mapping them to Thanks again for testing all these controllers! |
Beta Was this translation helpful? Give feedback.
-
|
It's indeed mapping them to JOYCODE_1_RXAXIS and JOYCODE_1_RYAXIS |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
Tokens The 4 back triggers are acting as replacement buttons for A, B, X, Y. They are working but don't appear as separate output as expected. |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
All buttons are working and correctly labelled |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
Tokens No analog sticks, they have following value: |
Beta Was this translation helpful? Give feedback.
-
|
It's not recognized by SDL Mame, it appears correctly in the windows USB controller properties |
Beta Was this translation helpful? Give feedback.
-
|
Now with joystick fallback introduced through mamedev/mame@e817a3d
All buttons are working albeit not labelled, like for the other version 4 axes are reported but not physically present. |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
Analogue sticks are not working, there is a switch on top to use the arcade stick as left stick, d-pad or right stick but only the d-pad position works. The guide button is functioning. Labels |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
All 3 possible inputs (LS, DP and RS) are working correctly, all buttons are correctly labelled |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
Main Stick is Tokens Some mapping issues: Some mapping misalignment |
Beta Was this translation helpful? Give feedback.
-
|
That one also doesn't appear in SDL Mame, I double checked that other pads are still recognized |
Beta Was this translation helpful? Give feedback.
-
That's normal for now with |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
Tokens Two physical buttons are reported in Mame and Windows but they aren't exposed, probably part of the controller board. |
Beta Was this translation helpful? Give feedback.
-
|
That one also doesn't appear in SDL Mame |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
Wrong labels Tokens
|
Beta Was this translation helpful? Give feedback.
-
|
sdlmame -version
All Labels, axis and buttons assignment are ok. Successfully tested disconnection and reconnection |
Beta Was this translation helpful? Give feedback.
-
Yeah, the assistant button is not exposed by the game controller mappings, and I’m making no attempt to sidestep that and add extra buttons for now. Maybe at some point in the future, but no guarantees. |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using
LT is incorrectly mapped to JOYCODE_1_Button12, should be analog axis. Same issue in Windows usb controller utility. Tokens |
Beta Was this translation helpful? Give feedback.
-
|
Windows 11 22H22 using All notes from Logitech controller for XInput apply here... except this controller does not have the Neck Slider. Screenshots also the same. Windows 11 22H22 using All notes from Logitech controller for SDL apply here... except this controller does not have the Neck Slider. Screenshots also the same. |
Beta Was this translation helpful? Give feedback.
-
|
mamedev/mame@701898d improves support for guitar and keyboard controllers. I haven’t had time to do drum kit controllers yet. This only affects XInput (i.e. should be tested with For guitars:
For the Rock Band keyboard:
@darkmos could you please build from latest source and re-test some guitars and see if the controls behave as described? @MrDo-Arcade could you please build from latest source and re-test the Rock Band 3 keyboard and Logitech Guitar Hero Controller? |
Beta Was this translation helpful? Give feedback.
-
|
I don't know if it could be the cause but the Madcatz is a rockband instrument so only B+R+G+Y and the pedal while the Band Hero is a Guitar Hero instrument so one extra button to consider B+R+G+Y+O and the pedal |
Beta Was this translation helpful? Give feedback.
-
I’m still confused by this: I don’t see how it could have been missed the first time you tested as you posted a screenshot that shows LSB off: I guess there’s some chance it changes too fast to see, as the menu updates relatively slowly when it’s displayed over the system selection menu. If you try with a 60Hz game running behind the menu can you see it? (You can get to the input devices menu while in-game by bringing up the main menu, choosing Input Settings, then choosing Input Devices.) If the buttons are all behaving as expected with SDL, that’s even more confusing. It must be reporting the correct values, but they’re somehow getting mangled before they get to MAME. I’m really struggling to understand how it’s happening. Does it show the expected button values in the DirectInput control panel? (On English Windows 10, this is called “Set Up USB Game Controllers” – I don’t know what it’s called in other languages.) Should I give you a replacement input_xinput.cpp that spews noisy logging to try and see where the bad value is coming in? |
Beta Was this translation helpful? Give feedback.
-
|
I see it as button 5 when I press the pedal in control panel, button 9 is always on (I guess it's the LSB button) I downloaded Vanilla Mame 0.251 binaries from mamedev The pedal registers as LB + negative value on RSY and LS is always on |
Beta Was this translation helpful? Give feedback.
-
Sure we can give it a try |
Beta Was this translation helpful? Give feedback.
-
So there’s something weird going on, as the left stick button is now stuck on from Windows’ point of view. I wonder what would cause that?
Well, we know now that there’s something weird happening in the XInput driver. SDL isn’t having the issue because it bypasses that driver and reads HID reports directly. I just don’t know why it’s now having this issue when it worked a couple of days ago. |
Beta Was this translation helpful? Give feedback.
-
|
That's cool... Rock Band keyboard works as expected.
Went back and downloaded the manual... there are three jacks on the right of the keyboard, next to the pairing button.
Researched, and I can just use my drum pedal in the Overdrive Jack. In the screenshot from earlier, RSY is this input. Grabbed an earlier MAME version... at rest is -65536; activate the input, it is 0. In Windows game controller properties, this is defined as "Y Rotation." Far as I know, the actual Overdrive Pedal that plugs in here is a digital input, like the kick pedal. Not sure what would happen if an analog pedal was plugged in here... snip from the MIDI manual for the keyboard:
And... as I'm typing this... you've got another update... fun. Two more questions then:
So now that it's easy to update... everything above still applies... I see that LB/RB have been added... but still not sure what they are for. I do know that the Touch Strip and Sustain button are not working... but don't work in earlier builds either... and also don't register in Windows. |
Beta Was this translation helpful? Give feedback.
-
|
If you connect it via MIDI, you can use it as a MIDI source for emulated systems with MIDI ports (e.g. synths). It won't go via the input system if you do it that way. It does support an analog pedal - in MIDI mode you can assign it to expression, foot controller or channel volume. You should be able to use an expression pedal for a Yamaha digital piano, or any compatible pedal. Looks like there's no way to determine whether a pedal is connected. I'll patch it up to work. |
Beta Was this translation helpful? Give feedback.
-
|
BTW it's pretty easy to find the preliminary MIDI manual for it, e.g. here: https://archive.org/details/rb3keyboardmidimanualv2/page/n2/mode/1up |
Beta Was this translation helpful? Give feedback.
-
The controller reports the presence of LB and RB. I’d prefer to err on the side of showing controls that are reported as present even when they don’t physically exist rather than making assumptions about controls not being present and making them unusable if the do happen to be present on some other variant of a controller. RB on guitar controllers is another case of this. I don’t know whether it’s possible to read the touch strip or the button beside it via XInput or if it’s only for generating MIDI modulation/pitch control events. It isn’t straightforward at any rate. |
Beta Was this translation helpful? Give feedback.
-
|
Update on Mad Catz Wooden Rock Band 2 Guitar: Most everything works as expected, from the newest update. Exception - Pickup Controller:
Current build... Pickup Selector still doesn't work. |
Beta Was this translation helpful? Give feedback.
-
|
Update on Logitech Guitar Hero controller: Everything works as expected. As mentioned before, RB doesn't seem to do anything on this controller (I haven't read through darkmos notes yet... so not sure if that is why it is still there). For @rb6502 ... after reversing both axis for the tilt and whammy bar... turning down sensitivity levels a bit... and a few rounds of practice, I was able to get to the first checkpoint in Outrun. 😎 And that was a workout... I might have to install some streaming software and post a video after all. |
Beta Was this translation helpful? Give feedback.
-
What scheme were you using? That game might work better on a guitar with a neck slider strip. Some I thought of for those guitars:
|
Beta Was this translation helpful? Give feedback.
-
|
mamedev/mame@8b224a5 should hopefully finish of rhythm game controller support so I can move on to addressing mamedev/mame@10820. Support for the DJ Hero turntable has been added:
The Rock Band keyboard pedal should work as a second slider axis – zero at rest, negative when depressed. The Logitech Guitar Hero Controller pickup controller should be fixed – dumb bug on my part that I didn’t notice because I’m not testing with real controllers. Note that MAME’s large default joystick dead zone swallows one of the positions. If you want to see all five positions as distinct values, use I still can’t see why the drum kit controller bass drum pedals aren’t working properly. If there’s still an issue, I’ll need another set of eyes to help debug it. @darkmos could you please test the turntable controller when you have a chance to update again? @MrDo-Arcade could you please test the Rock Band keyboard pedal and Logitech Guitar Hero Controller pickup selector when you have a chance to update again? |
Beta Was this translation helpful? Give feedback.
-
|
Functionaly everything works as expected for the Dj Hero controller. Scaling-wise the turntable goes to -+30000 when it's rotated very quickly so it's ok but the effect goes to 7 or 8 digit values in each direction as soon as you turn or even touch the potentiometer, there I'd say it's way too sensitive. |
Beta Was this translation helpful? Give feedback.
-
OK, are you comfortable hacking the source a bit for testing? The code in question is in the function For testing, could you try changing line 1394 to change the multiplier to // translate axes
m_axes[AXIS_TURNTABLE] = thumb_left_y() * 1'024;
m_axes[AXIS_CROSSFADE] = normalize_absolute_axis(thumb_right_y(), XINPUT_AXIS_MINVALUE, XINPUT_AXIS_MAXVALUE);
// convert effect dial value to relative displacement
s32 effect_delta = s32(u32(u16(thumb_right_x()))) - s32(u32(m_prev_effect));
if (!was_reset)
{
if (0x8000 < effect_delta)
effect_delta -= 0x1'0000;
else if (-0x8000 > effect_delta)
effect_delta += 0x1'0000;
m_axes[AXIS_EFFECT] = effect_delta;
}
m_prev_effect = u16(thumb_right_x());Then build again and see if the response looks better. You want it to be around ±60000 when you spin it very fast. You can adjust the multiplier on line 1405 more if you need to. |
Beta Was this translation helpful? Give feedback.
-
|
After recompiling with change in line 1394 the Turntable is spot on, you really need to make it spin quickly to reach 60000 values. |
Beta Was this translation helpful? Give feedback.
-
That’s OK, I can put in reasonable values from those numbers – pushed mamedev/mame@bad01be. You’ll need to revert that file before you can pull from git again, e.g. |
Beta Was this translation helpful? Give feedback.
-
Recompiled and both inputs pass the Arkanoid test :) |
Beta Was this translation helpful? Give feedback.
-
|
Update on Mad Catz Wooden Rock Band 2 Guitar: It's actually this controller with the Pickup Selector. Works now. And with the alternate settings provided, the second input is also recognized. |
Beta Was this translation helpful? Give feedback.
-
|
Update on Rock Band keyboard: Well that's interesting:
Going back to controller properties in Windows:
|
Beta Was this translation helpful? Give feedback.
-
|
Well, I think this is going to be all for this phase. Everything seems to be working as well as one might expect, and definitely better than it was before. @MrDo-Arcade I'm not going to be able to make the keyboard work any better without actually having a device to probe. @darkmos given the stuck phantom input on the wireless drum kit is apparently creeping in between the HID driver and the XInput driver, I can't really give any troubleshooting advice, besides the usual Windows recommendations - try deleting the device in Device Manager and letting Windows reinstall it, try rebooting, reset defaults in the DirectInput control panel to clear corrupt calibration, etc. Thanks for all the testing! I'll start a new discussion here (and probably link from reddit) when the next phase is ready for testing. |
Beta Was this translation helpful? Give feedback.
-
|
You're welcome, I was happy to contribute and make use of all those peripherals I'm ever touching :p |
Beta Was this translation helpful? Give feedback.
-
|
Some other notes I probably should have mentioned sooner, since I don't see them mentioned above: Rock Band Drum Kit
Rock Band 4 Instruments for XBox:
|
Beta Was this translation helpful? Give feedback.
-
|
Running Windows 10 Pro 21H2. I'm running this with the standard MAME version available from MAMEDev (since my build settings got hosed and need to redo everything.) This is a Hori Fighting Stick EX2 for the Xbox 360, I also have an Xbox 360 pad connected (my main input device.)
Skipping DirectInput for XInput compatible joystick 0xfded0f8a98. No other lines related to the stick showed up in the console.
The stick shows as a POV hat and is recognized by MAME as "XInput Player 2." The stick is recognized as the DPAD, and the buttons are all detected as normal, except for the left / right sticks, which aren't present.
All buttons and the stick show 0 when not pressed. LT and RT show -65536 when pressed, A / B / X / Y / LB / RB / Start / Back / DPAD show 1 when pressed. The Xbox button brings up the Xbox App on Windows and isn't shown as a button in the Fighting Stick EX2 properties window. I'll try to build a current source version of the emulator in the next few days, just lemme know if there's anything else I need to do to get more info for you. |
Beta Was this translation helpful? Give feedback.
-
The whole point was to use a current build because the code was overhauled… Also, GitHub CI is a thing – you can grab a binary of GitHub and drop it over your checked-out source if you can’t build your own. |
Beta Was this translation helpful? Give feedback.
-
|
Sorry about that, I wasn't aware that I could just grab a prebuilt exe that way. Here's the output of a proper build with the same Hori Fighting Stick EX2:
DPAD is now recognized as Joystick, and LT/RT now output 1 when pressed. All buttons except for Xbox Guide work fine, Guide still brings up the Xbox Windows app overlay and can't be configured. I have a couple of other sticks I'll be trying out, one of them is Bluetooth, would that cause any issues? |
Beta Was this translation helpful? Give feedback.
-
8bitdo SN30Pro v1 in Switch mode
8bitdo SN30Pro v1 in Apple modeAll buttons map up except for touchpad, which has no mapping. Slight bit of drift in the analogue as noted in Switch mode, but it wasn't moving at all in that space. It stayed steady at the point in the screenshot.
8bitdo SN30Pro v1 in Android modeHere we hit a problem. Left analogue maps correctly. Right analogue is mapped to RSX for both X and Y axis movement. Buttons map correctly.
|
Beta Was this translation helpful? Give feedback.
-
I think it’s just a bad built-in mapping string. For some reason it’s got this: |
Beta Was this translation helpful? Give feedback.
-
|
8bitdo Arcade Stick Linux, multiple modes (switch/xinput) on these again, so i've put the data here Runs OK in each mode, as per normal 8bitdo approach any wired'eque mode in xinput reports as an x360 controller, while BT reports as Xbox One wireless controller. The physical P1/P2 buttons do not output data at the OS level in any mode, not sure why yet. |
Beta Was this translation helpful? Give feedback.
-
|
See the image of the controller below, Button 11 and Button 12 are missing in MAME. Using Stable Controller IDs:
The Xbox Wireless Adapter for Windows 10 can connect up to 8 controllers. I only had 2 to test with and it seems that the Instance Guid of the controller is linked to the JoystickId, it is not a unique identifier for the controller. For example, if I turn controller 2 on first it will be JoystickId 0 and Instance Guid: {170c5400-59e9-11ee-8001-444553540000}. The Device Instance Path is also not a unique identifier as the values in <> change.
In the video I didn't press Button B because it would act as the back button. mame_test_controller.mp4Environment information |
Beta Was this translation helpful? Give feedback.
-
|
The following are the default values when first opening MAME and before turning the wheel or pressing either the accelerator, brake or clutch. After turning the wheel or pressing either the accelerator, brake or clutch, the values return to:
Screenshots taken from Logitech G HUB:
In the video I didn't press Button 1 because it would act as the back button. mame_test.mp4Environment information |
Beta Was this translation helpful? Give feedback.
-
|
Does MAME identify all the axes for this controller properly, and does it give sane default input assignments? I didn’t have any driving controllers to test with. |
Beta Was this translation helpful? Give feedback.
-
|
I updated both posts with extra information including images and videos, main text updates: Controller: Using Stable Controller IDs:
Driving Controller: After turning the wheel or pressing either the accelerator, brake or clutch, the values return to: |
Beta Was this translation helpful? Give feedback.
-
Does anyone have Xbox-compatible flight sticks, driving controllers or dance mats and a Windows system they can do some testing with to help me out? In particular, there’s some stuff I’d like tested with Xbox-compatible sticks like the Thrustmaster T.Flight HOTAS One or Hori HOTAS Flightstick. I’d appreciate testing with the
winhybrid(Windows) andsdlgame(Windows/macOS/Linux) input modules.If you you can build
OSD=windows(default on Windows), please set your controller(s) to XInput mode if possible and post the following:mame -joystickprovider winhybrid -verbose -joystick_deadzone 0– filter for lines containing “XInput”If you you can build
OSD=sdlfor Windows, macOS or Linux please post the following:mame -joystickprovider sdlgame -verbose -joystick_deadzone 0– filter for lines starting with “Game Controller:”Windows DirectInput controller handling hasn’t been changed for now.
Beta Was this translation helpful? Give feedback.
All reactions