Throughout its relatively young life, digital cinema audio has been anything but elegant. Too many formats, no shortage of uncoordinated innovation, and no agreement on how to simplify distributions. Confusion reigns, starting with DCI’s omission to include SMPTE DCP audio in its specification until August 2013 (by means of an errata), long after products were tested for compliance and installed. Now that we’re on the path to create yet another SMPTE audio standard, it might be useful to think about what we can actually accomplish in the field, rather than what we can accomplish on paper.
In a recent seminar held in Las Vegas during CinemaCon, the time required to implement another audio standard was pointed to as a key issue towards solving the problem of multiple audio formats. If only it were so easy as that. What wasn’t said at CinemaCon is that the industry has yet to be 100% compliant with the original SMPTE 429-2 Annex A audio specification first published in 2009. One only has to inquire as to why Hollywood studios, creators of the DCI specification, continue to distribute their movies in a non-DCI-compliant format using Interop DCP and not SMPTE DCP to grasp the gap between paper documents and reality.
More importantly, even if another audio standard is created today, there remains the fact that equipment in the field will not be compliant without further investment. Someone has to pay, whether it is manufacturers, exhibitors, or studios. If NATO, for example, were really focused on the issue, it would be questioning who will pay, not what will be standardized.
Using economics as our guide, then perhaps we should look at the cinema audio problem differently. To have an immediate and positive impact on distribution, and no impact on current investment, the new format should be backwards compatible as a single distribution. It must be backwards compatible while meeting the promise of immersive sound.
The three formats now in discussion can be evaluated in this light. Dolby Atmos utilizes the 16-channel audio capability of both Interop and SMPTE audio. In fact, Dolby Atmos utilizes the Interop audio format, and not SMPTE audio, as Dolby’s legacy servers do not fully support SMPTE DCP. But the 16-channel audio carried by Atmos contains a full 5.1 or 7.1 mix. However, the 5.1 or 7.1 track file is a backup, not a mix intended for backwards compatibility with non-Atmos systems. So while close, Atmos is not backwards compatible.
Next is DTS MDA. As with Atmos, MDA also utilizes the 16-channel audio capability of the DCP. But as with Atmos, the 16-channel audio carried by MDA does not contain a backwards-compatible 5.1 or 7.1 mix. It, too, renders object-based sounds, carried in a separate and synchronized MDA file, on top of its channel-based bed of audio. As with Atmos, the channel-based mix carried in the MDA distribution does not contain all of the sound elements of the original mix, and therefore cannot be played on a 5.1 or 7.1 system. For this reason, MDA does not pass the backwards-compatibility test.
That leaves Auro 3D. The companies behind Auro 3D, Barco and Auro Technologies, have long bragged that it is fully backwards compatible with 5.1 audio. Auro 3D is capable of being backwards compatible with 7.1 audio, too, but this capability has not been well promoted in cinema. Backwards compatibility is assured by design. Each 5.1 channel carries additional information in the lower four bits of audio data, which, in audio terms, are below the noise floor and not reproduced by the digital-to-analog converters that drive amplifiers. All of the immersive sound elements mixed in the studio are contained in the audible 5.1 channels. The inaudible portion of the distribution is used to generate the full 11.1 mix in the cinema auditorium. It does this by subtracting sound (not adding sound) from the original 5.1 sound track and redistributing the subtracted audio to the additional speakers. The subtractive idea is clever, ensuring that the 5.1 channel-based mix carries all of the sound elements of the immersive sound mix, while enabling the re-generation of the original immersive mix in the auditorium. Auro 3D alone passes the backwards-compatibility test.
In spite of the advantage of backwards-compatibility, Auro 3D has not achieved the degree of acceptance that one might expect. There most apparent reason is that Auro 3D is a proprietary format. At the time of this writing, it appears to be the most secretive format as well. Unlike Dolby and MDA, Auro has not published the contents of its bit stream. Dolby published the contents of the Atmos bit stream in two SMPTE Registered Disclosure Documents, now in ballot in SMPTE. (Such ballots are strictly based on the technical completeness of the submission, and not on the subject of the submission.) MDA submitted details regarding contents of its bit stream in SMPTE, in consideration of a standard. Neither disclosure grants users the right to develop their own interoperable system, as both technologies contain intellectual property of their owners. Neither disclosure constitutes a standard. Auro, however, has published nothing in SMPTE that describes how its format is constructed.
Even though distributors may hate to hear this, it is unlikely that a single audio distribution will ever prevail to serve all auditoriums. The momentum to create another audio standard is set. But rather than pretend that we’re on a path to solve the industry’s immediate problems by introducing something new and different, it might be smart to evaluate the options that the industry now has, and plot a course that provides both short-term and long-term benefits. If Auro 3D were to play its cards right, it has an excellent opportunity to play a role in such a solution.