Aphasia is a speech disorder usually caused by stroke or head injury and may involve a variety of communication difficulties. As 30% of stroke sufferers have a persisting speech and language disorder and therapy resources are low, there is clear scope for the development of technology to support patients between therapy sessions. This paper reports on an empirical study which evaluated SoundHelper, a multimedia application to demonstrate how to pronounce target speech sounds. Two prototypes, involving either video or animation, were developed and evaluated with 20 Speech and Language Therapists. Participants responded positively to both, with the video being preferred because of the perceived extra information provided. The potential for the use on portable devices, since internet access is limited in hospitals, is explored in the light of opinions of Augmented and Alternative Communication (AAC) device users in the UK nd Europe who have expressed a strong desire for more use of internet services.