Two months ago we ranted about bad audio, goaded on by reader Bro Duke, who wants the FCC to DO SOMETHING about it! Well, we got a fair amount of mail, with some interesting horror stories.
Gary Davis, who has written for TV Technology, wrote to point out that in the televised concert on Sept. 21 – dedicated to the victims of Sept. 11 – the stereo was reversed on feeds in New York and Los Angeles, although he noted that in London it was correct.
Reader George Bell speculated that some of the rebroadcast stuff I described sounded like undecoded Dolby C. He also described a technique used by some engineers to keep levels down at transmission by recording a 17 Hz tone up around 0 VU (your basic slate tone), mixed with the program. Of course, any heavy limiter (how do you spell Optimod?) will turn this program down to whatever level the threshold is set – audio police or vandals?
Bro Duke also wrote in to relate the sad tale of a badly overmodulated commercial production produced while monitoring on a cheap 12-inch woofer housed only in the cardboard box it was shipped in.
At the same time, I did an informal survey of the audio for TV practices I encounter on my television (which has a "digital" cable feed with about 150 channels). And these days I'm watching/listening on a REALLY GOOD TV with excellent broadband audio configured in stereo, equivalent to a really decent home audio system.
By and large, I'm happy to report, the TV audio wasn't too bad. Nonetheless, there are bad spots and some of them seem endemic. One channel has had a loud (-15 dB?!) 60 Hz hum for about eight months.
The local access channels all have miserable audio, with massive distortion and varying amounts of hum and buzz. (It is bad enough that with 10 minutes' practice you can easily identify any given channel by its particular brand of lousy sound quality.)
More embarrassing to us pros, one major channel that broadcasts NFL football has a serious problem with stereo compatibility, summing an array of crowd signals with varying delay to mono. This, of course, yields massive and blatant comb filtering that simply destroys any entertainment quality the broadcast might have had. ("Martha! What's that danged howling noise? It's doing it again!")
But these are, by and large, isolated instances. Mostly, TV audio (and I am referring mainly to the voice-plus-bed quality of live news, talk and entertainment programming) is clean, intelligible and has decent frequency response.
However, a particular "style" of audio has emerged for voice production on TV. It has evolved as a function of the desire to have multiple isolated sources (I'm talking about voices here) emitting from the same acoustic space. Lavalier mics, because of their inconspicuous nature, have become the mics of choice for this task. Naturally, their placement is less than ideal, as is their directionality.
This leads to, er, enhancement (high-frequency EQ boost, mainly), which in turn leads to other local noises and HF leakage from other sound sources in the room. Some gating is done, and the general effort seems to be to reduce interference artifacts as much as possible.
This is all swell, in theory. In reality, voices are bone dry, very bright and edgy (what the audiophile guys call "spitty" and "sssibilant"), and often a vaguely sour-sounding low-level comb filtering in the "sssibilance" due to crosstalk with other sources that have their own open lavaliers trying to do the same thing.
This is quite a peculiar production style thing that now is how "TV sounds." Keep in mind, this is not how reality sounds. There is no longer any room tone on TV, no stereo ambience surrounding the voices, no sense of place, depth or context. Talking heads are flat, in front, in your face and "spitty."
Is this good or bad? I don't know. It doesn't seem very pleasant. What I do know is that, when you listen carefully and objectively to it, it does sound a little weird and definitely unnatural.
Is there anything we can do to make things better? The answer is, of course, yes.
The obvious stuff that would solve the problems would be more meticulous attention to the implications and limitations of stereophony, better documentation and more attention paid to said documentation, less audio vandalism on the one hand and more restraint in the use of final-stage signal processing chains on the other and – most important – better monitors and monitoring environments.
In addition, restraint, care and thoughtfulness in terms of the following topics would help.
Maybe a lavalier isn't always the best way to go. It may be possible to use a much better microphone, placed better for human voice. There are some awfully good microphones out there, with some very interesting capabilities. Lavaliers may be quick and convenient – however, there may be more to life than just quick and convenient.
Stuffing in a 6-10 dB peak at 4.5 kHz in order to emphasize the consonants for intelligibility – in a room with multiple sound sources – is not necessarily a good idea. Once again, strategies for better placement and better microphones may reduce the need for "induced edge."
Companders are both seductive and deadly, particularly when we layer them (and we do, come on, admit it!). Less can be more – so lighten up on the ratio, the threshold and the release time, and don't put a heavy limiter with a low threshold in line with a really stiff noise gate at a high threshold. Instead, get better levels at the mic, working in a quieter studio.
Stereo and Surround
These can be your friends, if you can get a handle on how they work. The big secret weapon here is ambience, the short-term room tone and early reflections that aren't perceived directly. Unfortunately, right now "ambience" is OUT on TV, except for the occasional crowd roar and car drive-by at televised stock car racing events, and other similarly ambience-laden sporting activities.
The real secret to audio quality, of course, lies in critical listening. Critical listening is the intense and careful analysis of how recorded sound works, what sounds good, what sounds bad and why, what can be improved, and what can't. TV sound faces more than its fair share of problems in this regard. The visual requirements of the medium, combined with intense time and financial pressures, all mitigate against the meticulous effort needed to manage really good audio.
For the time being, TV audio is acceptable, generally, which is to say that you can understand the words, distortion is modest, and the noise floors and tone quality aren't too bad. Bro Duke's examples reveal the bottom feeders, maybe even the bottom quartile, of the system, I suspect. Mostly, production is better than those wretched cases (and, yes, I stayed up to watch Leno, Letterman and Maher to satisfy myself in this regard).
But acceptable isn't the same as good, and good isn't, to revisit Bob Dixon of NBC for a moment, good enough! And yes, we CAN hear the difference.
Thanks for listening.
Dave Moulton is hard at work on some wild new music! You can complain to him about this or anything else at www.moultonlabs.com.
© 2002 by David Moulton