How the law treats ‘enhancing’ of poor quality audio

By Chris Potter (Flickr: 3D Judges Gavel) [CC BY 2.0], via Wikimedia Commons

Like most things in the law, the handling of ‘enhanced’ audio is determined by a precedent – a ruling made by a judge in relation to one particular case, which is then applied in many other cases that seem superficially similar. The problem with the legal practice of following precedents is that if one judge is wrong, many others will be too. Unfortunately, that is what happened in relation to ‘enhancing’.

In the 2002 judgment commonly cited as a precedent in regard to ‘enhancing’, the judge opined that

the techniques used were merely ‘the aural equivalent of the use of a magnifying glass to enhance an individual’s capacity to perceive the relevant record’.

This is a very serious misunderstanding

To see why, consider what is involved in ‘enhancing’ a poor quality photograph or messy handwriting. Just making the photograph bigger is rarely useful. This is because magnification cannot add information that was never captured in the first place. And indeed forensic recordings are often found to be just as unintelligible after ‘enhancing’ as before.

This judgment has had several very unfortunate effects for the use of ‘enhanced’ audio.

First, the level of scrutiny given to ‘enhanced’ recordings is extremely low. On the misunderstanding that an ‘enhanced’ version is much like the original only clearer, it is just waived through with no questions, certainly no cross examination of the audio engineer. Evaluating whether the enhancing was effective or not is yet another task left for the long-suffering jury, who now have not one but two versions of the unintelligible audio to deal with.

Second, while ‘enhancing’ frequently makes no improvement, that doesn’t mean it makes no difference. The problem is the difference can create a misleading representation of the original conversation, with potential for significant injustice. We have an example shortly.

Third, the ‘enhanced’ version is provided to the jury, who listen to it with the detective’s transcript in front of them. The fact they believe themselves to be hearing an ‘enhanced’ version can confidently be expected to increase their acceptance of the (often inaccurate) transcript.

A particularly alarming observation

I’ve spoken to the judge who made the 2002 ruling and he had no recollection of stating these words, and no knowledge of their subsequent effect.

Of course this is no slur on the judge himself. In fact he is a particularly ethical and thoughtful judge. But surely it does give cause for concern about the process that allows judges to make rulings on language and speech with no consultation of experts in linguistic or phonetic science.

Unaccountable ‘enhancing’: some real examples

The following quote from a recent murder trial shows how low the bar is for having ‘enhanced’ audio accepted as evidence.

  • CROWN PROSECUTOR: {…}  Is this one of the discs that was enhanced?
  • DETECTIVE: Yes, it was.
  • CP: By an out‑company; is that right?
  • D:  Yes.
  • CP:  That is called {NAME OF MUSIC COMPANY}?
  • D: Yes.
  • HIS HONOUR: What does that involve, can I ask?
  • DETECTIVE:  My understanding is we gave them a copy of the master CD and they placed it in the computer programme and they put a number of filters through the audio recording in an attempt to enhance the sound.
  • {END OF DISCUSSION – EVERYONE WAS HAPPY WITH THIS EXPLANATION AND THE ‘ENHANCED’ VERSION WAS ADMITTED AS EVIDENCE}

Another instructive example

One recent rape case featured a recording in which a male voice could be faintly and indistinctly heard among the loud cries of the victim.

The recording was sent to be ‘enhanced’, in full expectation this could make it possible to hear what the male was saying, and who he was. Rather than explaining that this expectation was unrealistic and misguided, the audio engineer took on the job.

His ‘enhancing’ consisted in boosting the loudness of the male voice. This of course made no improvement whatever to its clarity, but did have the (unacknowledged) effect of giving the illusion the male was far closer to the victim than he appeared to be in the original – surely a matter of some significance in a rape case.

It might be useful to point out that in this case the audio was part of the defence case. It is not only the police that get these things very wrong.