Senior Member Research Staff Dolby Laboratories, Inc San Francisco, California, United States
It is generally known that increasing the relative dialog level of content (such as via Dialog Enhancement (DE)) can increase intelligibility and reduce listening effort. But what exactly is the expected benefit for a real-world audience which could have tens or hundreds of millions of people, some of whom have hearing impairments? We introduce a model that predicts which parts of a large audience benefit by how much from changing the speech-to-background mix ratio of a piece of content. The model is intended to allow decision makers to predict what impact changes in audio production guidelines or DE technologies will have on their audience.