This dataset was created for the third Clarity Enhancement Challenge (CEC3), which asked participants to improve the intelligibility of speech-in-noise for hearing-impaired listeners. CEC3 consisted of three tasks, each extending the simulated data used in the second Clarity Enhancement Challenge (CEC2) in a different way: Task 1: Simulation using real ambisonic impulse respond recordings. Task 2: Recording made over real hearing aid devices worn by a listener and with real head motion Task 3: Simulation with target speech mixed complex dynamic scenes recorded with an ambisonic microphone. This dataset contains the material for Task 3. The data for Task 2 is provided in three parts: clarity_CEC3_data.task3.metadata_only.v1_0.tar.gz — training and development metadata, including a checksum manifest (cryptographic hash values) for all training signals, to verify data integrity and ensure reproducibility. clarity_CEC3_data.task3.v1_0.tar.gz — training and development signals (due to file size, this archive is hosted separately at myairbridge link, 111 GB). clarity_CEC3_data_eval.task3.v1_0.tar.gz — evaluation signals and metadata. Further details about the challenge—including background, task definitions, and baseline systems—are available in the CEC3 documentation.