The Clausula Archive of the Notre Dame Repertory (CANDR) is an in-progress PhD project with the aim of cataloguing, transcribing and analysing digital facsimiles of the thirteenth-century repertory commonly termed Notre Dame polyphony, and a secondary aim of providing new datasets and analytical tools for studying medieval polyphony. This poster highlights the use in the project of (a) a new methodology for de-skewing facsimile images, and (b) average symbol masks in an OMR–enhanced workflow with an emphasis on creating an OMR workflow that is ‘good enough’ to accelerate the annotation of an image dataset of particularly transitional notation.