Recent research has raised concerns about legal procedures for admitting 'enhanced' versions of indistinct covert recordings used as evidence in Australian criminal trials. This paper seeks to deepen these concerns via an experimental study using two enhancements that were admitted in a trial on the grounds that the judge, listening personally, considered they assisted him to hear words from the prosecution transcript more clearly than he could in the original. Results demonstrate that, as in previous studies, the apparent clarity of the 'enhanced' audio depends on listeners having been primed by a transcript. However, this study goes further, by demonstrating that the two enhancements are not neutral, incremental improvements on the original, but incorporate unnoticed artefacts that cause them to be perceived differently from the original, and from each other. These effects were not acknowledged by the judge, who endorsed the prosecution's assurance that the enhancements had simply increased the clarity of the audio. Of course, the aim here is not to single out one particular judge, but to demonstrate that current procedures for admitting transcripts and enhancements are insufficient to protect judges, juries and defendants from potentially misleading interpretation of indistinct covert recordings.