Assessment reforms in higher education recognize the need for effective feedback that is timely, informative, and encourages positive attitudes towards future learning amongst students1. Although it is well established that feedback to students is one of the educational practices with large positive impacts on student learning2, evaluating the actual extent to which students engage with and utilize feedback is difficult. Studies show that when students do receive feedback it is often too brief, too broadly stated, and is often misinterpreted by students3. Recent studies have suggested that students find audio feedback more engaging and informative than traditional written feedback4, and prefer either audio or a combination of text and audio5 to text alone. During this study, we have developed a ‘rich media’ marking system, to help markers provide students with timely, detailed and situated feedback, and to investigate the potential of technology to facilitate new methods of feedback provision, such as in situ audio comments. The system can support large cohorts of students, large teams of markers, moderation of markers, and provides learning analytics during the feedback process. It consists of three components: 1) Administration interface that processes submissions for marking and moderation, 2) iPad application that allows markers to provide feedback as audio, handwritten and typed annotations, and to grade using criteria rubrics, and 3) Feedback viewer allowing students to view their feedback on any web browser. In addition, the in-built learning analytics allows for extensive evaluation of the provision and use of feedback, providing empirical evidence of both marker and student behaviors. To date, the system has been trialled on 852 student assignments in two separate trials, the first based on a final student laboratory report, and the second on a draft of a scientific literature review. Preliminary analysis of the learning analytics data indicates extensive provision of feedback by markers and use of the in-situ feedback by students. However, it also revealed differences in the ways in which markers use the annotation modalities, with almost half avoiding use of audio annotations whilst the remainder used audio annotations extensively. Differences were also found in student use of feedback, with students in the second trial viewing their feedback for considerably longer time periods than the first, suggesting there was an effect of timeliness of feedback provision. For audio feedback specifically, students reported in interviews that they (i) favoured audio feedback for its greater detail and personalized nature; (ii) valued contextualized, in-situ feedback; and (iii) were concerned if audio comments were too long or lacked meaning. These findings are currently being used to target professional development for markers and to enhance the provision of quality feedback.