Dr Lech's computer program is scanning through the voice data, using pattern recognition techniques to seek telltale "harmonics" unique to a depressed person's voice. It can detect harmonics related to seven different emotions, including fear, excitement or anxiety...
Professor Allen said the research was at an early stage but showed great promise. "In clinical work, the voice is often one of the first things you notice (in depressed patients)," he said.
"I can tell with my patients on the phone, sometimes after only a few words, how they are going. You can hear a flatness, a tension coming through. The voice changes very dramatically when they are depressed."
This is called listening.
I find it upsetting and a little disgusting that there is room for demand that such a voice-reading program exist. We're all creatures of conversation, even me, Little Miss Hermit Crab, have enough and sit through enough conversations in a day to understand that the dialogue is only part of the whole interaction. Facial inflection, louder body language, pauses for consideration, the words you choose, the words you don't choose, and how you say what you say make the whole. These things make a face to face conversation so much more meaningful than some quick IM of txt speak, and even a hand written letter.
We all read these things. All you need to do in order to understand them is pay attention.
No one should ever need a computer to tell them what is contained in another person's voice.