Keeping the Audience Happy

By Matt Hines, Audio/Multi-Media Producer, Post Production @ iZotope Inc. The job of the re-recording mixer is an extremely challenging proposition. Working in audio post production often involves creativity and compromise in equal parts, as mixers must balance a variety of aesthetic opinions with the limitations of their delivery requirement or playback medium. That’s not an indictment of loudness standards. I believe standards encourage aesthetically good, dynamic mixes. Mixing for loudness compliance also helps mitigate inconsistent perceptual volume between program materials—a common consumer complaint. But looking at the statistics in reports by Ofcom, the FCC, and other regulatory bodies within the broadcast industry, it’s clear there are other audience complaints over which we, as the content creators, still have the ability to influence. Revealing statistics At a conference a few years ago, I heard a troubling statement by a major international broadcast television network. Of the thousands of audience complaints they received that year in their biggest market, more than 13% concerned dialogue intelligibility. This number, in my view, is alarmingly high. Here’s a closer look at the breakdown of that 13%:

• 19% of intelligibility complaints are due to accents. iZotope is certainly at the forefront of digital signal processing, but accent removal is not yet within our wheelhouse!
• 11% regard the level of audible background noise in broadcasts. This issue is easily and instantly addressable via noise reduction solutions, such as iZotope’s RX.
• Many of the remaining complaints are related to the more nebulous issue of “clarity.”

Interestingly, these figures seemed consistent across all demographics. This is despite the fact that a 20-year-old viewer listens to televisual content at an average volume of 55 dB SPL whereas an 80-year-old viewer listens at more than 80 dB SPL. Researching consumer habits The BBC has conducted extensive research into consumer listening environments for several decades. Much of that research is shared publicly for content creators and audio product manufacturers to digest. We know so much about how our audiences consume content, we consistently produce the best-sounding mixes we possibly can, and we take great pains in ensuring loudness compliance... so why is intelligibility still an issue? One reason may be that intelligibility is more compromised in stereo playback systems. These configurations allow a listener to easily perceive extreme left and extreme right audio information, yet the center is more muddy. This experience is further worsened by the typical resonances and reflections of an acoustically untreated home environment. (We’ve had 5.1, with a dedicated center channel, for years, and yet we haven’t seen widespread adoption of this format in most listeners’ homes.) To me, this suggests that a clear and continuing emphasis should be placed on dialogue intelligibility. Though the proportion of intelligibility complaints may be somewhat revelatory, the idea of mixing for it is certainly not new. Indeed, everyone I speak with in iZotope’s post community always mixes with this goal in mind following the fundamental basics of channel placement, frequency adjustment, and stem-based workflows. Proposing a solution I believe that we, as an industry, need a more standardized metric for measuring dialogue intelligibility. I believe that this metric should anticipate as best as possible the wide variety of consumer playback mediums, as well as how they affect the sound of a mix and its elements. I also believe that this metric should result from a collaborative community. This community includes manufacturers like iZotope, the audio post professionals we’re privileged to call our users, the standards and regulatory bodies, and of course the broadcast networks themselves. We should all work together to introduce this new standard, mix towards it, and QC against it. Final thoughts Indeed, there is a large volume of research that’s been conducted in related arenas (see papers by the ISO and TNO, for example). Yet the focus of the existing research typically surrounds public address systems and loudspeakers. I’m keen to see a renewed focus on dialogue intelligibility. We should apply research toward meaningful, actionable tools and standards. Then mixers can analyze and objectively report on a dialogue intelligibility metric that considers the balance of dialogue within a mix (dialogue, music, sound effects, etc.) as well as the potential playback mediums. Because in this day and age, “it sounds good enough to me” doesn’t cut it.   About Matt Hines Matthew is an audio technologist specializing in post production and audio mastering, and a keyboardist by night. Originally hailing from England, where he cut his teeth editing MIDEM award-winning classical recordings and dialogue and audio restoration for Nimbus Records, he's called USA home for the last eight years. Freelancing as a sound designer and dialogue editor for independent film soon led him to a role as Audio/Multi-media Producer and Product Manager with iZotope, working with a dedicated team to deliver key innovations such the RX Post Production Suite.