Over the past several years, SMPTE has been engaged in studies of cinema audio. For over a year, a Study Group was conducted to review B-Chain alignment in cinema. This year, a new working group was formed to further this work, and presumably form standards. Most worrisome is that many of the participants haven’t taken time to read the existing standards.
In October 2012, this publication produced a report titled
“X-Curve Is Not An EQ Curve,” highlighting a common misunderstanding of the 35-year-old cinema sound alignment method called “X-Curve.” To help the organization of video engineers along in its understanding, it was decided to offer a revised version report to SMPTE as an education piece. Ioan Allen, one of the original VPs of Dolby and the creator of X-Curve, and John Allen (not related), a cinema sound guru and creator of the HPS-4000 speaker system, were kind enough to review and contribute. After months of cajoling, the article finally appeared on the SMPTE site this month. (Oddly, published as a fancy white paper.) It was both enlightening and polarizing.
The point of the article is that the response curve known as “X-Curve” is only useful in regards to the entire measurement method prescribed in SMPTE standard ST202 B-Chain Electroacoustic Response. The mixed reactions to this news are noteworthy. Nearly all, if not all, US-based cinema sound experts applauded it. But for some people, the message of the article was truly news. It appears that “experts” outside of the country don’t understand how to apply the alignment technique correctly, which has become fuel for all manner of complaints, most of which are targeted incorrectly at the standard itself. Worse yet, the very existence of a standard has downgraded in many people’s minds the value of alternative methods to aligning sound. After 35 years, one would expect improvements. They exist, and they shouldn’t be discarded.
The worrisome part is that the studies under SMPTE are not of the quality of those by experts of the past. SMPTE has the ability to gather lots of data, but the ability of its members to crunch and evaluate remain be seen. Members like to talk about “science,” as if measurement alone can always reveal what sounds good and what doesn’t. This is the reason why so many reasonable people stay away from the field of sound: because sound quality has a strong subjective component. It’s an inherent property of sound in commercial environments.
In 1994, Claude Fortier and Pierre Cote, in a paper for the Canadian Acoustical Society titled “Digital Signal Processing Applied to the Equalization of the Loudspeaker Room Interaction”, made the noteworthy statement: “Equalization in the frequency domain effectively only equalizes the minimum phase part of the response due to the presence of all-pass phase components in the very complex room response.” That may seem like a lot to digest, but what it says, in simpler words, is that the 1/3 octave equalizers used to align cinema sound systems will never correct all of the acoustic irregularities imposed by the room. That’s why those who insist on applying extreme equalization to obtain a “perfect” response curve on their measurement gear always get it wrong. And this is why subjective values must always come into play.
The SMPTE 25CSS Calibration Ad Hoc Group is conducting a significant study of sound in cinemas, undertaken by the cinemas themselves. It’s an impressive undertaking, and there will be a tremendous amount of data collected. But there is this nagging issue of capturing how the cinema sounds, the quality of which, after all, cannot be reliably measured. Binaural head recordings of the auditoriums are being considered, a valid technique for capturing how the human ear hears sound. The goal would be to study the recording of the cinema auditorium against the measurements made of the same auditorium, hoping to find correlation. But this is where this author departs from the committee. New tricks are unlikely to reveal new magic. Science says that 1/3 octave equalization isn’t perfect, and that means that subjective values must be applied during system alignment to achieve good sound. The SMPTE committee, sadly, is spinning a lot of wheels but is unlikely to be going anywhere.
SMTPE ST202 and X-Curve may not be perfect, but theory says that no method ever will be. That doesn’t mean it’s not possible to get better and/or more consistent results using other methods, or that other methods shouldn’t be used. The enlightened approach would be to embrace many methods, so that there’s an opportunity to match skill with technique. Variety is much more fun. But the SMPTE committee is looking for the Holy Grail. Its members could easily end up in a snake pit.
