Calibration issue: cameras are not in the right places in SteamVR

[keltrivera]

After a lot of calibrations (I've tried 2-4 PS Eye cameras and 1-3 PS Move Controllers) I can't get psmoveservice to work correctly with SteamVR. For some reason, cameras are in really weird places when I look at them in headset from SteamVR. They have incorrect orientation (one o them could be below the floor in VR while it is pretty high in real life. It looks like camera positions were rotated and scaled somehow though their relative position to each other seems ok-ish. I've made sure that callibration procedure is right and tried different settings, but nothing helps. Here is an example of how does it look in SteamVR compared to real life:

[Cretkensigh]

This has happened to me a few times. Do the camerea appear in the supposed spots when you test tracking pose on ps move?

Check each controller individually and take note where the green cone is coming off the cameras. If its not in the right spots with its fov (green cone shape) you may need to do the calibrate pose again with that cameraz thats out.

When mine was out it was tracker 2 (id tracker 1) just kept doing it when all controller and tracker until they are in the right spots

[Cretkensigh]

And when you do the start and select reset orientation it seems to move the tracker relitive to your facings so always make sure your facing them. Also put move controllers against your hmd. Right up against it with triggers facing forward. Seems to give the best correction

[keltrivera]

I tried calibration many times, but it didn't help. So I've decided to check older releases, downloaded 1.1 and it worked! Now any calibration works as intended. May be it is an issue in PSMoveService itself? Not sure about other versions (didn't try them) but 1.1 works for me like a charm, but latest 1.4.2 don't. Also when I test calibration poses, only a small domain is shown, so it is hard to guess, where are all the cameras and controllers. But it says OK on all the cameras most of time so I guess they see the controllers.

And here's a stuff may be not directly connected to SteamVR bridge, but calibration mat is in only available in Letter format, which is used only in U.S. and Canada, so you won't find any paper of that size in other places. I had to combine several sheets of paper to do it. If it is possible (and appropriate in this thread) I would like to request an A4 - formatted calibration mat as an option

[Cretkensigh]

Hmm maybe downgrading might sort my one controller issue too, might have to try it. In the test calibration pose where you see that small space try using the mouse wheel to zoom out and click drag to rotate view shpuld be easier to see.

Calibration mat doesnt seem to be needed i use a huge book cover thats roughly square shape and just eye place them in the corners and middle when needed. Seems to work for me lol

[stephenohair]

I had a similar issue with Trinus + PSVR and PSMoveService using this SteamVR driver (PSMoveSteamVRBridge) and PSMoveFreePieBridge. I did 2 things to fix it.

1. I made sure that when setting up room scale I made sure to use advanced and edit the orientation of the green play area so the arrow points to the front. I physically placed the tracked headset near the front of the play area. Then in the room scale setup I edited the play area and used the corner arrows to rotate the square facing the same direction as the headset was pointing.

2. I then had to recompile the FreepiePSMoveBridge and invert the X and Z axis that are written to FreePieIO. I'm pretty sure that there's some funky workaround in Trinus 0.9.2 that isn't reading the FreePieIO correctly since I don't need to do this in Riftcat. The last thing I changed in PSMoveFreePieBridgewas I took 100cm off the Y axis (up/down). I did this since my HMD pose appeared a good meter above where it was supposed to be.

This is one weird and wacky suite of software, of which I'm very grateful for, that needs to be unified at some point. A standard plugin approach might have been better.