Forward Error Correction copes with packet losses, but at the expense of an increase of the end-to-end delay. By failing to take this into account, existing error control schemes for audio often lead to end-to-end delays larger than 150 ms, which has an impact on the perceived audio quality. In this paper, we develop an adaptive error control scheme for audio which is delay aware, i.e. which incorporates the impact on the end-to-end delay in the choice of FEC. To this end, we model the perceived audio quality as a function of the end-to-end delay and of the encoding rate at destination. We develop a joint rate/error/delay control algorithm which optimizes this measure of quality and is TCP-Friendly. We show that our scheme increases utility in a single class best effort network. We evaluate the benefit for audio sources to use the Alternative Best Effort service, which offers applications the choice between lower end-to-end delay and more overall throughput.