XInput Wrapper for DS3 and Play.com USB Dual DS2 Controller

[Scarlet.Crush]

The Game Controller settings app on Win 8 is kind of broken, if you want a laugh try calibrating your DS3/X360 with it.

Z-Axis (Triggers) is displayed correctly during calibration, but not on the Test page after calibration.

[Gevaerd]

@Scarlet:
I did your tests with ScpServer-0.4.1.85 and XInput-Wrapper-SCP-2.2.1.88 with both Win32 and Win64 ScpUser.exe, and the results were the same:
- with XInput1_3.dll present: problem reproduced in ScpServer-0.4.1.85, but not in 2.2.1.88 (same behaviors I noted before).
- with XInput1_3.dll renamed to _XInput1_3.dll: ScpUser just says that controller is disconnected, even though it works in windows game controller test dialog at the same time (where it reproduces the problem in ScpServer-0.4.1.85, but not in 2.2.1.88). Tried connecting the cable before opening ScpUser, connection after opening it, disconnect and reconnect, but ScpUser does not recognizes it as connected.

I used an official PS3 USB cable, but the behavior is still the same as when using my camera USB cable, so I guess the cable is fine.

The same behavior reproduces in the Calibration Dialog. And it doesn't help anything in solving it.

My OS: Windows 7 SP1 x64

What dazzles me is that only 2.2.1.88 (did not try older 2.* though) does not show that problem. Does it have anything that could hide the problem? Such as smoothing movement value or slower value read refresh rate (the value refresh seems fast, like the other). Or could it be more compatible by reading the data someway different?


I discovered that in ScpUser if I disable deadzone, after moving the stick beyond the position where the axis gets its maximum value (where that strange variation occurs for an instant), the other axis keeps varying between -1296 and -1806, with small pauses in -1548 (negative when moved the X axis and variation is in the Y axis; positive when moved the Y axis and variation is in the X axis). In center position currently the X axis is -258 and Y axis is 774. But those variation values are the same for both axis, so it is not relative to the center position, it is somehow something absolute. This variation seems rhythmic. I'm sure its not me moving it, because I can easily move an axis keeping the other at a fixed value before moving into the area that misbehave.


I'm also attaching image of the SCP DS3 Server output when I try to connect to my bluetooth, in case you can tell from it why it doesn't connect. My bluetooth dongle is very cheap, so I wouldn't be surprised if it cannot work with DS3, but it says it is 2.0+EDR so I thought it might work. FYI, this bluetooth also did not work with MotionInJoy.

[Scarlet.Crush]

@Gevaerd: the only thing the filter driver (2.2.1.88) does is send a "Start command" to the DS3 when it is connected. No smoothing or filtering of the report stream.
The XInput1_3.dll when used with the filter driver (2.2.1.88) reads HID reports from the "HID-compliant game controller" device, when used with the server (0.4.1.85) reads reports via UDP from the Service/Server which reads them directly from the USB port.
The content of the reports is exactly the same, the dll then formats them into XInput format using the same code. So you shouldn't see any difference.
The direct USB read is a lot faster than an HID read, so pushes the DS3 harder.

Was there no more output after the last line of the screenshot?
The last line shows a message being sent to the DS3 requesting a PSM Service Channel connection, the DS3 is not responding to it - so I don't think it is an issue with the dongle.

Are you sure it is a genuine Sony DualShock 3? Read here. It doesn't sound like it is behaving properly. Where did you buy it? CECHZC2U and CECHZC2H are for 2 different regions.

[Gevaerd]

@Scarlet:
There is no output after that last line. The controller lights still blinks for some time after the last line is shown, then it turns off. And that is always the last line in the output.

I saw in Wikipedia that Sony has quietly released a newer model of the DualShock 3, one that is 500 mA and 5.0 V. Mine says that it is 3.7V and 300mA, that would be the older DS3 model. Anyone there has a 3.7V DS3 and experienced or not these issues in the sticks?

Examining the controller it seems genuine, but I cannot be absolute sure. The package does not have the sticker the site you pointed says... but I see in Here that the sticker program started after June 2009 (doesn't say exactly when after). Could be my controller from before that program? More than 3 years ago seems a lot.
I bought it in Brazil, but it is imported and was not produced specifically for here. Because if it were, the packaging would need to have some writings in our language. It says Made in China.

I don't think USB read vs HID read has something to do with the differences in behavior, because MotionInJoy and Tamamy also recognize the device as HID and the problem reproduces on them.

At least the filter driver works well for me, so I will stay using that one. Ignoring the smaller movement length for the sticks to reach maximum value, the controller works perfect with the filter driver.

[Scarlet.Crush]

@Gevaerd: I have 2 SixAxis (3.7v 30ma) from when I bought my release day fat 60GB PS3, 2 DS3s (3.7v 300ma) from when they were released in the UK and a Rev A1 DS3 (5.0v 500ma) I bought in the last week or so.
Rev B1 was released recently. Rev A1 has been around for quite a while.

All my controllers accept the PSM Service Channel connection, a few users have reported they are using rev B1 controllers without issue.

If you bought the controller recently it should be at least a Rev A1 model if it's genuine Sony, nowhere would hold stock for 3+ years.

[unvaluablespace]

Just want to point out too that while googling around about the newer revision controllers, other improvements have been made it seems, like better force feedback, higher quality controller casing (some have said the texture is slightly different, more durable), and the analog sticks have received small improvements as well. Can't say for sure if these are accurate. Just stuff I found during research, from various sites and forums.

[Scarlet.Crush]

Bluetooth Alpha 1

1. Virtual Bus - Added support for XP and Vista.
2. Virtual Bus - removed all Toaster specific code and changed all Guids. (No conflicts with Toaster or UsbIP)
3. Added Suspend/Resume support. (There is a 10sec delay built in after Resume before devices are reconnected, this is to allow Windows time to recover - may be able to reduce this)
4. Included custom Lilipad binary.
5. Included custom Lilypad source changes.
6. Included sources for ScpUser and XInput1_3.

Important - make sure your Bluetooth dongle and DS3(s) are plugged in via USB before running ScpDriver.exe. Make sure you have the latest DirectX redist installed and the latest X360 drivers available.

New Install
1. Connect your Dongle and DS3 via USB and let Windows install it's default drivers.
2. Unpack the archive.
3. Create a directory where you want the Service/Server to run from [I used C:\Program Files\Scarlet.Crush Productions - create from an elevated command prompt]
4. Copy the contents of the ScpServer\bin to the directory you created.
5. Run ScpDriver.exe as Administrator.
6. Click Install [if you don't want the Service installed uncheck Configure Service first]
7. Click Exit.

Upgrade from Prototype 4.1
1. Unpack the archive.
2. Stop the Service [net stop "SCP DS3 Service" - from an elevated command prompt, or use SCM]
3. Close ScpMonitor.
4. Copy the contents of the ScpServer\bin to the directory you created.
5. Run ScpDriver.exe as Administrator.
6. Uncheck Configure Service, then click Install.
7. Start the Service [net Start "SCP DS3 Service" - from an elevated command prompt, or use SCM]

Note : I bought a USB 3.0 hub today and decided to do some compatibility testing with it. I also have a keyboard with a USB 1.1 hub which I also tested.

1. PC Trading Dongle (CSR based) BT 4.0 - compatible with USB 1.1, 2.0, 3.0
2. Jsg Accessories Dongle (CSR based) BT 2.0+EDR - compatible with USB 1.1, 2.0, 3.0
3. Belkin Dongle (Broadcom based) BT 2.1+EDR - compatible with USB 2.0, 3.0 - not recognised at all by OS in USB 1.1
4. Asus Dongle (Atheros based) BT 3.0+HS - comapitible with USB 2.0 - reports different hardware id and Windows fails to find a driver for it in USB 1.1, 3.0 *

* ScpServer / ScpMonitor show 00:00:00:00:00:00 as the Host Address. If you see this make sure your dongle is plugged in to a USB 2.0 port.

@Gevaerd: I have created a custom build of ScpControl.dll for you for Bluetooth support. Install as above, stop the Service then replace ScpControl.dll with the one from the archive below, start the service.
The PSM Service Channel code is bypassed, it works on my SixAxis and original DS3s, let me know if it fixes your issues when connecting and using via Bluetooth.

[Bositman]

May I suggest updating the first post so people can be easily directed? This will get buried pretty quick Awesome job BTW

[Scarlet.Crush]

@Bositman: I've been wanting to limit it's exposure to people who are technically aware, and willing to test and provide feedback until it has reached Beta. Hence not posting in #1 and limiting the hardware ids to specific vid/pid/revs.

Shouldn't be too long before getting to Beta. It's been a lot of fun so far

[Bositman]

OK your choice, although you could just mark it with red letters as WIP or something Anyway keep it up