We've run multiple audio comparison tests using the PESQ (Perceptual Evaluation of Speech Quality) scale, which is an objective and recognised industry standard measure for audio quality.
Our tests show that we consistently improve results by 0.5 points across a challenging data set of noises including construction, café/office babble, babies crying, keyboard typing, wind/water, and more. This means we bring typically unintelligible noise or noise requiring a lot of effort to understand to levels which cause far less brain drain to understand — above 2.0 on the PESQ scale.
In addition, we have a very low CPU usage (average of 11.5%) and reduced latency (< 8 milliseconds). This varies based on the computer you're using, but any recent computer will see IRIS Clarity work seamlessly without disrupting your workflow. We've also tested our technology in the most extreme environments including Formula One.