Why 6 MHz Channels Take up 6 MHz: Part 1
by Leslie Ellis // January 10 2005
One of the most familiar of cable technology terms is the well-worn “6 MHz.” It is, to the cable engineer, what the inch is to the carpenter: An enormously familiar unit of measure.
It’s hard to go more than two sentences deep into a technology discussion, in fact, without bumping into 6 MHz. Usually, the context is “how many.” How many 6 MHz channels are left in the analog tier? How many 6 MHz channels are on the digital tier? Which services use how many 6 MHz channels?
As a scribe for this industry (with a thing for verbatim notes), I thought it might be amusing to count the number of times “6 MHz” crept into my laptop in 2004. The search took 40 minutes; it found 1,589 instances.
As cable lingo goes, 6 MHz is the good old wagon: Steady. Reliable. Fundamental.
But why are 6 MHz channels 6 MHz, and do they continue to matter, in an all-digital world?
Going Back to 1941
The answer to the first part of the question goes back 64 years, to 1941. That’s when the National Television Systems Committee (NTSC) put out the Big Plan for television signals. (It modified the big plan a dozen years later, to graft in color — but the widths were set, so to speak, in ’41, the pioneers say.)
As required bandwidth goes, the NTSC’s work to define how broadcast, color television would work translated into a need for 4.2 MHz.
Huh? How’d we get to 6 MHz, then? In cable systems, the extra 1.8 MHz goes toward modulation — the process of imprinting a video channel onto the carrier that conveys it to subscribing TVs.
And six decades later, here we still are, with 6 MHz.
Meanwhile, the “all digital” momentum builds, with digital simulcast as Chapter One. Already, technologists are beginning to wonder: Is Old Reliable 6 MHz becoming a vestige?
The thinking: If all channels are available digitally, and if the carrying capacity for digital is measured in Megabits per second (Mbps), not Megahertz (MHz), is it still relevant to think in those old-fashioned, analog, 6 MHz chunks?
The Bandwidth Shelf
Cable’s bandwidth, as this column has described previously, is like a bookshelf. For the sake of illustration, let’s say the bookshelf is 750 inches long, partitioned into 6-inch slots.
The first analog channel begins at 54 inches, and ends at 60 inches. (“Left” of it is the space allocated for upstream stuff, plus some padding.) The next one occupies the slot from 60 to 66 inches. And so on, until the last slot, which stops at the 550-inch mark.
That’s the analog shelf space. It holds around 75 channels.
The digital section of the shelf is also cubby-holed into 6-inch slots. They begin at the 550 mark, and extend to 750. The digital shelf holds around 33 slots — except that each 6-inch slot on the digital shelf can carry enough bits for around 10-12 TV channels.
It is a vestigial, if odd, fact of life that the digital shelf doesn’t really need to be segmented into 6 MHz slices. Digital transit is more interested in Megabits-per-second than it is in the rudiments of a 64-year-old standard, which reflects different constraints.
Translating MHz to Mbps
When digital video arrived, more than a decade ago, a key component of it was digital modulation — the process by which digital information is imprinted onto a carrier for transit to homes.
Cable uses a digital modulation method known as quadrature amplitude modulation, or QAM. (Some say it as a word that rhymes with “mom;” others as a word that rhymes with “slam.”)
QAM is essentially plumbing — hard-core, brainiac plumbing, but still plumbing. It’s the digital equivalent of FM (frequency modulation) or AM (amplitude modulation).
QAM never really developed quite the sex appeal of, say, compression techniques, which continue to generate technical headlines to this day.
Yet any cable service that is rooted in digital — video, data, voice — moves through QAM modulators to homes. These days, most cable providers run, or plan to run, a version known as 256-QAM.
The details can cause immediate and binding glaze-over (trust me). The thing to remember is this: 256 QAM, as a conveyor belt, moves 38 Mbps worth of data, per 6 MHz slot.
One digital video program, using the most common of compression methods — MPEG-2 — uses 3.75 Mbps. Multiply by that by 10, and you’re pretty close to 38 Mbps. That’s why about 10 digital programs (more with available tweaks) fit into a 6 MHz slice.
A digital channel earmarked for broadband data also moves at 38 Mbps per 6 MHz channel, sliced according to downstream speed ceilings. Ditto for 6 MHz channels that carry voice-over-IP calls.
Huh, you say. Would it make sense to eliminate the 6 MHz chunks, or make bigger groupings of chunks? If it’s just bits, why worry with 6 MHz slices! Just treat cable bandwidth as the big fat pipe that it is.
That’s the thinking behind DOCSIS 3.0, also known as “wideband DOCSIS” and “channel bonding.”
Here to Stay
Like the inch, the 6 MHz marker will likely never go completely away — too many things rely on it as such, and behave accordingly.
So, like a good old wagon, 6 MHz will keep showing up at work. It’ll just do so differently. More on that next time.
This column originally appeared in the Broadband Week section of Multichannel News.