Where Are We? This is Where!

Last month, I reviewed the Modulation Sciences SpiderVision Surround Sound signal level meter. As the lead-in to that review, I noted Lord Kelvin’s dictum that to measure is to know.
Author:
Publish date:

Last month, I reviewed the Modulation Sciences SpiderVision Surround Sound signal level meter. As the lead-in to that review, I noted Lord Kelvin’s dictum that to measure is to know.

Well, I spent a couple of days this past month using the SpiderVision meter to measure in considerable detail the various channels emerging from my set-top box. The results are quite interesting and, as Kelvin noted, I now know a good bit more than I did.

At the same time, Randy Hoffner, another columnist for our magnificent magazine, wrote to suggest that I was being a little unfair in my March column in suggesting that HD might not be viable yet. In a subsequent exchange of letters, we sort of identified the problem as existing in the difference between what content producers generate and what end users actually get to perceive in their living rooms.

So, it is timely and relevant to share with you something about the reality of the couch potato experience, as measured by your somewhat trepid reporter.

I am a subscriber to DirecTV, a satellite service. I have an expanded service with lots o’channels. These are advertised and described as having “crisp digital picture and sound.” For the purpose of this study, I auditioned 184 of these channels, making notes about the measured nature of the audio signals of all of them.

I did not audition the XM music bundle, and I skipped some of the instructional channels supplied by DirecTV. I did include various shopping channels, the NASA channel, the RFD channel and the C-Spans, as well as the major networks, news, sports and movie channels.

Note also that I reside in a somewhat narcissistic and well-to-do cultural ghetto (Boston metro). I’m assuming that the technical quality of our offerings in such a media hot bed is approximately as good as it gets.

Before we actually start picking into the details, note that this is not intended as a rant about DirecTV. It is not possible for me to determine what, if any, of the problems I found are the fault of DirecTV, and I make no case that any of them are. I have quibbles about DirecTV’s service, but it is a considerable improvement over its cable predecessor, Charter Communications.

Let’s consider what I found: There were 184 channels that I observed and documented. Of those, only nine (5 percent) were HD channels. One-hundred-seventy-one (93 percent) had analog audio that triggered the Dolby Pro Logic upmix on my TV, while 18 (10 percent) had Dolby Digital audio, (some channels did both).

Eighty-two channels (45 percent) had mono signals, 59 (32 percent) were in stereo and 43 (23 percent) had a significant enough surround component to be regarded as surround sound. Keep in mind that this distribution is program-dependent and will vary over time. Also, what I observed does not necessarily reflect the intent of the source producer—only what was coming out of the set-top box while I was observing it.

I typically auditioned each channel for several minutes and tried to get at least some program content (as opposed to commercials). Some stereo signals appeared to be simulated, and on numerous channels, there was a significant amount of HF phase shift between L and R on several mono signals, resulting in a fake, fairly spitty-sounding narrow stereo image.

So, the basics are this--approximately half the TV audio being broadcast is mono, slightly more than a quarter is stereo and slightly less than a quarter is surround sound. Only 5 percent of the content is received as HD, although perhaps 10 percent may be generating HD.

The good news is that 138 of the channels (75 percent) appeared to have no problems with their audio signals.

The bad news is that only 138 (75 percent) had no problems, which means, of course, that 46 channels (25 percent) did have observable errors and/or problems. Uh-oh!

Note also that I did not measure level variance nor did I measure dialnorm error. I suspect that had I done so and specified that anything greater than 3 dB variance was an error, the number of problem-free channels would have been much smaller.

Many of the problems had to do with imbalances between left and right channels, often as much as 6 dB. One channel had significant and pervasive clipping, while another had pronounced overshoots, but without the clipping artifacts.

INTERPRETING THE DATA

Much of the stereo was quite narrow, and some of the surround consisted of mono voice and surround audience noises. One old mono movie suffered from embarrassing synthesized surround. At least three other channels were transmitting legacy mono movies in mono (which is fine with me). A few channels (nine, or 5 percent) had sound that inspired me to write “good” or “really good.”

Finally, I made an arbitrary and somewhat subjective distinction between “major” and “minor” errors. The major errors were, in my opinion, embarrassingly audible and representative of very poor engineering craft. Seventeen channels (9 percent) had such errors. Minor errors are probably audible, but more representative of sloppiness than incompetence. Twenty-nine channels (16 percent) exhibited such errors.

This was an informal, nonscientific study. We have only one sample point, for instance. It is reasonable to assume that what I found will not reproduce very precisely at other sites, with other service providers, etc.

On the other hand, this particular couch potato has some professional expertise and, further, sought professional installation. Also, the service provider under test broadcasts in parallel to (I am guessing conservatively) more than 1 million subscribers, so a certain amount of consistency across set-top boxes can be expected.

I suggest then that the errors in my findings will probably not be much greater than plus or minus 10 percent vis-a-vis the average condition for all viewers. That in turn suggests that the number of channels with errors will constitute between 22 and 28 percent of the total. The amount of mono being broadcast ranges more widely (it being program-dependent, remember), probably between 35 and 55 percent.

Mono is still with us, in spades. This is true even though we broadcast using stereo or multichannel. Given that voice tracks are predominant and mono by nature, that isn’t surprising. I was, however, a little surprised by the number of mono music beds under mono voiceovers. And, sadly, at this point in time, HD and surround sound (synthesized and real) remain very much in the minority.

This reality is a little sobering. We really truly aren’t there yet.

And thanks for listening.