The NERC report specifies a level of transcription suitable for large-scale data entry. Speakers are minimally identified and there are simple conventions for unclear passages, hesitations and pauses, speaker overlap, laughter etc. For convenience this is copied in appendix to this paper.