This article examines the meeting point of Olivier Messiaen, Australia and birdsong, particularly as it relates to the transcription of pied butcherbird (Cracticus nigrogularis) vocalizations. It draws upon correspondence from Messiaen to the Australian ornithologist Sydney Curtis, printed here for the first time, as well as two recordings not previously available to musicologists, from which Messiaen transcribed. Both the recorded birdsong models and Messiaen's transcription of them in his cahiers are subjected to sonographic and waveform analysis. In analytical scrutiny of eight of these transcriptions, I demonstrate that Messiaen's pied butcherbird transcriptions conform to their models in a partial and highly personal way. I propose a provisional template for Messiaen's approach to birdsong transcription, in order to answer Alexander Goehr's question: 'Why do birds sound like birds, but Messiaen's birds sound like Messiaen?'