Question ARC Raiders — Invalid CameraManager Coordinates & POV Struct Issues

[byte_corvus]

Boys, has anyone been digging into the current camera state for this title? I am having a headache with the CameraManager offsets. I have verified the object via the PCOwner pointer (0x3A8), but reading the struct values for rotation and location is returning pure garbage.

The Technical Breakdown:

1. Environment: Standard External approach using C# with SharpDX. Kernel-level read via custom IPC.
2. The Issue: FOV reads perfectly at 80, but XYZ location and rotation are returning invalid float values (e.g., -1.25e-14, 2.0e+09).
3. Observations: I suspect the CameraManager struct has changed or the data is being obfuscated/encrypted in recent builds. I have already scanned the memory around the object for a valid POV struct, but found nothing consistent.

// Current read attempts
camLoc = _ipc.Read<Vector3>(_pid, cameraManager + 0x400);
camRot = _ipc.Read<Vector3>(_pid, cameraManager + 0x428);
camFov = _ipc.Read<float>(_pid, cameraManager + 0x450);

Things I have considered:

1. Pointer Mismatch: The PCOwner is definitely correct, so the CameraManager address isn't the problem.
2. POV Struct: Is it possible the engine is using a different struct layout for the ViewTarget now? It feels like the memory layout might have shifted or is being actively protected.
3. Encryption: I noticed the game uses SIMD for some pointer decryption. Do you think the camera data itself is being processed through a similar XOR or shift/rotate routine?

I am seeing some weird garbage in my logs, which suggests I am hitting the right address but the data representation is non-standard. If anyone has mapped these offsets or dealt with the recent camera updates, drop a hint. Has anyone successfully managed to get a stable viewmatrix to draw ESP yet? My current W2S implementation is failing because the cam data is clearly corrupt/wrong.

Let me know if you guys think I should pivot to a different hook or if there is a known shift in the CameraManager structure I missed. If someone has a cleaner way to grab the view matrix directly without relying on this specific struct, I am all ears.