I am currently developing an HEVC player using VideoToolbox on an iOS device. I have successfully created an HEVC decoder that receives HEVC streams from our custom image capture and encoding device, and it can decode and display images properly.
However, when my image capture and encoding device configures the encoder to output HEVC streams with fragmented NALUs (i.e., an I-frame or P-frame is split and stored across multiple slice NALUs), the iOS decoder can be initialized successfully but fails to decode and output images.
Can VideoToolbox properly decode HEVC bitstreams when a single frame is split into multiple slice NALUs?
Key Observations: 1. Single-NALU frames work fine. 2. Multi-NALU frames (sliced I/P-frames) cause decoding failure. 3. The decoder session is created successfully (VTDecompressionSessionCreate returns no error).