INDIAN WELLS, CALIF.: One complication of controlling TV loudness is rendering stereo audio from surround sound. Funneling 5.1 channels into two can result in louder audio if not done properly, said Bruce Jacobs, chief technologists of Twin Cities Public TV. Jacobs was at the HPA Tech Retreat addressing regulations associated with the Commercial Advertising Mitigation Act, which go into effect next December. CALM Act rules direct TV stations and cable systems to deliver audio with consistent loudness.
The big networks started implementing loudness controls and facilities for 5.1 surround sound long ago, but there’s a different dynamic at the station level. Jacobs said that more than two-thirds of PBS member stations don’t have the infrastructure to handle 5.1 surround sound. Those stations have to downmix, and they may end up with noncompliant material as a result.
“A station can be deemed noncompliant when broadcasting ‘compliant’ content,” he said.
Surround sound delivers audio in 5.1 channels—left and front, center, left and right surround; and low-frequency effects—the point one. In downmixing the left and right front are preserved, Jacobs said, “but what do you want to do with the rest?”
The center channel typically is added equally to the stereo left and right channels, usually with a 3 dB drop in level to compensate for the average 3 dB boost experienced in a room when feeding the same signal to two speakers.
The rear right and left channels must also be considered. Feeding them into the front channels can increase the volume of sound effects, possibly making it hard to hear the dialog. So the rear channels are usually dropped 3 dB as well, but in this case to reduce the level for the listener, not to match the center compensation.
Consequently, a show mixed with a main singer or speaker in just the center channel will produce the same listener loudness in both surround and stereo. However, if the show is mixed with the singer equally in all three front channels, the downmix loudness will be 2.8 dB louder than the surround mix.
One option to avoid this problem is to customize the downmix, Jacobs said, but that raises new issues. Metadata values in an encoder can be set to subtract -12 dB, but that doesn’t do much for TV networks and stations without a metadata channel. And producers don’t want to bother with metadata because all of the distributors would have to deal with that metadata.
“A lot of equipment that could carry the metadata channel would have to be configured,” Jacobs said.
Instead, Jacobs suggests that the industry define a standardized downmix for systems without metadata--using -3 dB for center and rear channels--and requiring loudness measurement of the downmix. This, he said, would encourage producers to mix using techniques that insure consistent loudness for both the surround mix, and the consumer down mix.
~ Deborah D. McAdams, with thanks to Bruce Jacobs.
See Bruce's presentation at:http://data.memberclicks.com/site/hopa/2012_TR_Pres_BJacobs_Calm_Down.ppt
