---
pretty_name: real_audio
---
Original Audio Files used to train RVC models and create the Arabic Deepfake Audio Dataset.
Files with the naming scheme: f/m-nn_resampled.wav were not included the fake part of the dataset, due to resource constraints.

Files have been preprocessed as described in the Arabic Deepfake Audio Dataset.