A Mask That Can Give You Superhuman Abilities 68
An anonymous reader writes "The students at Royal College of Art in London have developed masks that can increase your sight and hearing senses. They allow you to choose one conversation or one visual among a cluster of sounds and visuals, then hear or see the one which you want to. There are two masks developed by them: Eidos Vision and Eidos Audio. Eidos Audio allows a wearer to hear a specific conversation in a crowd and could be developed as a hearing aid and help ADHD sufferers. Eidos Vision improves vision allowing wearer to see 'time trails' similar to a timelapse photography."
So glasses+hearing aides (Score:4, Insightful)
We already have glasses and hearing aides that don't make you look like a complete dork.
Re:I'll stick with... (Score:1, Insightful)
You mean your gonna take a web server offline for five minutes?
Re:So glasses+hearing aides (Score:4, Insightful)
We already have glasses and hearing aides that don't make you look like a complete dork.
If you watch the video in the article you'll see that these devices are much more sophisticated than normal glasses and hearing aids.
The sight mask appears to allow you to see the trail of motion (allowing a football manager to see where his players have moved from) and the hearing mask allow you to focus on a specific sound (allow a child with ADHD to focus on useful sounds). Although these designs do look odd they are probably not the final designs and aren't supposed to worn all the time.
Re:So glasses+hearing aides (Score:4, Insightful)
It also can't work well due to the ongoing gain control problems with speech amplification. Digitally managed gain control has been an amazing cluster !@#$ and a complete waste of processing power. You can't twitch the gain up and down to maximize distant speech without smearing "plosive" sounds, and you can't do it well with the limited digital sampling rates available in most low power A-D conversion systems.
Of course, there is a solution to this. Simply amplify the analog signal and clip it when it's too loud. Keep the digital systems out of it, their 1-bit minimum signals throw out low level signals that matter for low frequency sounds. This approach was documented by Robert Licklider in the early 1960's. But it's not amenable to digital processing because A-D conversion throws out all the low level signals and the zero-crossing analog data that contains the actual information. And it's got prior art, over 50 years old, so there are no patents available. So instead we get these sucky, smeared A-D based overprocessed gain systems that have already thrown out most of the important information.