Ability to delete the audio by fragments (each time you hit pause when recording would be the limit of a fragment). Currently, you can either discard or save the whole recording. It would help a lot the recording process if the user could gradually save the fragments of audio that are good and iterate on the next fragment individually until it's good enough