I’m trying to understand the behavior we’re seeing on two comparable Lenovo laptops. One is 21KWS7FD (with A1000), the other is 21FWS2JN00 (with RTX 4050)
We run the same code, initializing DesktopDuplication API, and then encoder on the adapter servicing the monitor being captured.
The difference manifests when capturing display connected to USB-C port. In case of the first laptop, the USB-C port is always routed to iGPU, the BIOS has no settings to alter GPU settings. The initialization always succeeds.
In case of the other laptop, the USB-C port is routed to dGPU. Data duplication initializes fine using that output. However, the encoder, when trying to activate the MFT, with default BIOS settings, always returns Catastrophic failure [0x8000FFFF].
However, that laptop does have ability to select iGPU vs dGPU (default) in BIOS. When changing the setting to iGPU, the pipeline initialization succeeds – however the adaptor being used for USB-C output (and thus for both DDA and encoder MFT) is still the nVidia adapter.
I’m trying to understand the logic here, and if there is a way to make this work regardless of what the BIOS setting is set to.