|
 |
Do I Need a New HDMI Cable?
Probably Not...Here's Why
Since we're a cable vendor, people tend to assume that we will always be in favor of buying a new HDMI cable. And, heck, we never mind if you do. We won't turn down your money. But we also like to tell the truth, and the truth is that most of the time, when people ask us whether they need a new HDMI cable, the answer is "no." The point of this piece is to help illuminate the issue and let you know why that is, as well as helping you figure out if your situation is one of the less common ones where the answer is indeed "yes."
It's "pressure sales" season, as we write this, and we know what happens. Time and again, people get sold new HDMI cables they don't need, because they are told they'll need it for new features and whatnot. Let's have a look.
Bad Reasons to Buy an HDMI Cable:
Bad Reason No. 1: new HDMI Specification Version
Probably the number-one cause of concerns about needing a new HDMI cable is the version-number thing. The HDMI specification has had a variety of versions, from 1.0 through 2.1, and at each stage, we get concerned phone calls: "My cable is a 1.3 cable. Will it handle 1.4?"
Here's the thing: up through version 2.1, all conventional passive HDMI cables have had the same inner structure. Details of construction may differ from one supplier to another, but there's only ever been one update to the inner cable architecture, and that was 1.4's addition of the optional "Ethernet over HDMI" feature. That particular feature, as it happens, is very rarely seen on actual source and receiving devices, making the question whether a cable is "with-Ethernet" or without something of a moot point for the vast, vast majority of consumers.
In other words, there is nothing physically special about a cable built after the arrival of 1.3 as opposed to 1.2, or any other spec-version change. The layout's the same; the pins connect in the same way. 1.3 introduced two performance tiers for testing, but many cables which existed before the two-tier spec function at the higher tier without modification. 2.1 introduced a new third tier, but many cables built before 2.1 are handling new higher-speed signals just fine. How can you tell if yours isn't up to snuff? We'll get to that, in a bit.
But before we leave this question of version-number compatibility, one further fact will underscore the point. HDMI Licensing, which licenses and polices use of the HDMI specification, prohibits HDMI Adopters from designating a cable by spec version. Why? Because it's deceptive. It doesn't mean what people think it means, for various reasons (including the fact that ALL cables certified under any version of the spec remain compliant, albeit with conditions, under ALL spec versions). When you see someone offering an "HDMI 1.4 cable" or an "HDMI 2.0 cable," that's a violation of HDMI's naming standards, a real red flag; and that alone is a good reason to hang on to your money.
Bad Reason No. 2: Features, Protocols, and Resolutions
All right -- so "spec version" is not itself a reason to need a new HDMI cable. But a visit to any of a number of online vendors will show that HDMI cables are described as supporting different lists of features. One of them may list new Dolby surround standards which others do not. Another will say it handles 4K video. Another says it handles deep color, another says it'll handle different colorspaces supported by later spec versions. Surely feature support in an HDMI cable is important?
Well, no. In fact, HDMI cables do not vary in their ability to support any of these sorts of protocols and features. You don't need to fill the cable with "deep color" fluid, or wrap it in special Dolby foil, to get it to support deep color or Dolby surround protocols. All of the features and protocols available in HDMI depend upon the source and destination devices, but the cable's role in conveying them is only to connect the same wires to the same connector pins they'd be connected to anyway. When a vendor tells you that a particular cable supports, say, some Dolby standard, it's rather like selling you cholesterol-free peanut butter. The claim is deceptive even though, strictly speaking, it's true; it's just that there's no such thing as peanut butter that DOES contain cholesterol, so it's not a reason for picking one peanut butter over another.
The one partial exception to the above is in the realm of resolution and color depth. The more pixels there are on a screen, the more screens being sent per second, and/or the more data that must be sent for each pixel, the more bandwidth the cable needs (audio bandwidth is so much smaller than video that it is a non-factor here). There now are three regular testing tiers for HDMI cable and one optional tier. The lower tier, "Standard" or "Category 1," certifies performance up to the bitrate required to convey conventional HD (720p or 1080i) video at standard color depth. Anything beyond that calls, at least in theory, for a "High Speed" or "Category 2" cable, which is tested to the 10.2 Gbps bandwidth called for by HDMI 1.4 and which, through a bit of specification prestidigitation, is supposed also to be good for the 18 Gbps bandwidth called for by HDMI 2.0. Then there are "Premium HDMI Cable" certifications, where the testing is actually conducted out to 18 Gbps, to make sure the bandwidth really is there as it's supposed to be. Lastly, for data rates exceeding 18 Gbps and all the way up to 48 Gbps, there's the Ultra High Speed certification.
Now, if you know you're running a "Category 1" cable but you have a "Category 2" need for bandwidth, do you need to worry, and do you need a new cable? Maybe, and we'll get to that in a moment. But quite possibly not. Before we get to that, there's another snake-oily problem to address.
Bad Reason No. 3: Stated Bandwidth
We've been seeing a lot of cables that advertise various levels of stated "bandwidth." 15 Gbps, 20 Gbps, 24 Gbps, 33.6 Gbps, 36 Gbps, and so on, and so on. Surely THAT's got to be for real, hasn't it? We were, after all, just speaking about how bandwidth can be important.
Well, bandwidth IS important. No question about that. Shoving a few billion ones and zeros through a cable every second without error isn't as easy as some people think it is. But these stated bandwidth numbers don't mean anything -- nothing, zilch, nada. This one really seems to shock people -- after all, the number's on the package. Doesn't that mean the cable's been tested and has the bandwidth stated? Isn't that an electrically objective measurement?
To understand, let's back up a bit. What is "bandwidth"? In a cable, it's the ability to faithfully convey the signal over the frequency range where the spectrum of the signal falls. We don't run SDI video over rusty coathangers, because the high-frequency loss properties of rusty coathangers aren't very good (though we just might get away with running, say, S/PDIF audio, with much narrower bandwidth, down that rusty coathanger if we don't need it to go very far). Digital signals like HDMI video have interesting spectral properties -- they're very broad, so one can't just look at a single frequency, as we might when testing a radio antenna. But how do we MEASURE this "bandwidth," and in what units?
If I said that a cable had resistance of one ohm at DC, we'd all know more or less what that means. And we could get an ohmmeter and measure it. If we wanted to be precise, we could get a fancy $25,000 ohmmeter that would check it down to the thousandth of an ohm accurately, and see just how close we came to the target. And resistance being a basic property of wire, we'd find that it's not that hard to achieve broad agreement among electrical engineers as to whether our wire came within an ohm, plus or minus some tolerance. A good ohmmeter, properly calibrated, would satisfy everyone.
But now that we've got the electrical engineers in the room, let's ask them a different question. Instead of a wire of which we want to know resistance, we present a data pair, like one of the pairs in an HDMI cable. And we ask these electrical engineers: how much bandwidth does this data pair have? No matter how much test gear we have in the room, this question isn't answerable as stated. The engineers will ask us, in response: by what criteria are you judging bandwidth? Bandwidth, you see, is not a simple property like resistance. There is no single definition of it. Instead, there are specifications for conveying data in accordance with various protocols, and these put the burden of conveying a signal, under the particular terms of the specification, upon sending circuits, cable, and receiving circuits. One spec writer might write a spec that goes easy on the cable, and puts more burden on the circuitry; another might write one that places heavy demands on the cable, and less burden on the circuitry. But without a specification to judge what is deemed sufficient to pass some bandwidth, the word "bandwidth" has no common or pre-existing meaning.
Now, the HDMI specifications do indeed give testing criteria of this type. But these various non-conforming bandwidths offered up by cable vendors do not exist in the specs; there are, even with 2.1, no criteria that would assist a person in saying whether a cable had 24 Gbps bandwidth, for example. There are only a few test speeds. Category 1 testing is at about 2.25 Gbps; Category 2 testing runs at 4.95 Gbps and 10.2 Gbps, with an equalization formula (the very kind of thing one simply can't derive from first principles, but must have the spec, to know) applied to the 10.2 Gbps testing. "Premium" certification testing is done at those rates and at 18 Gbps, again with specified equalization, and Category 3 testing runs signals at 48 Gbps and takes other factors into account besides the basic eye-pattern testing. A claim that a cable meets HDMI standards at one of these bitrates, accompanied by a relevant certification from the Authorized Testing Center showing the appropriate testing category, is genuine (and, incidentally: the version number shown on that certificate is meaningless; a Category 2 certification under 1.3, for example, is valid for all other versions of the spec). Other claims of bandwidth are meaningless.
Do the vendors know these numbers are meaningless? In most cases, yes, they do know, and that means that the use of these bogus claims is dishonest. In some cases they may not know -- they may simply be repeating what a Chinese factory rep has told them. That may do something for their honesty, but not for their credibility. Again: hang on to your money.
When You Do NOT Need a New HDMI Cable, Summarized:
Before we finally get to the issue of when you do need a new HDMI cable, let's review. These are the three important points above:
(1) You do NOT need a new HDMI cable simply because your gear supports a new version of the HDMI spec. If you see something advertised as an "HDMI 1.4 cable," "HDMI 2.0 cable," or "HDMI 2.1 cable," this is dishonesty, pure and simple.
(2) You do NOT need a new HDMI cable because your gear supports some new feature or protocol, with the (very unlikely) exception of Ethernet over HDMI. If you see a cable advertised as supporting some new standard or other for digital audio or some such thing, that claim isn't "false," but it's misleading and meaningless -- your existing HDMI cable supports that, too.
(3) You do NOT need a new HDMI cable because the new cable makes some claim about stated bandwidth. The only need (and maybe not a need, as we'll discuss) is for a "high-speed" Category 2 cable if your signal will exceed Category 1 bandwidth limits, and you can identify that cable very easily: does it say "High Speed HDMI Cable" on the box? If it does, you're good. Other claims of bandwidth are meaningless. (There will also be a new tier of cables, for HDMI 2.1 Category 3, but we don't yet know how those will be labeled or when the first will hit the market).
When You Do, Actually, Need a New HDMI Cable:
All of that said, is it possible that on some occasion you need a new HDMI cable? Yes, it certainly is. But, fortunately, it's very easy to check.
When you've made some change to your configuration and are now putting more bandwidth demand on your cable, the best thing to do first is to just plug it in, using your existing cable, and see what happens. If you plug it in, set your gear to the highest supported resolution/framerate/color depth combination, and find that your picture looks good, you're done. It really is that simple.
HDMI cables do fail sometimes when asked to handle faster signals than previously, and so it's possible that you will do this, and find problems. But the handy thing to know here is that HDMI cables generally don't fail subtly. You won't plug it in and find that the contrast in the dark areas has become muddy, or that the red hues just aren't as lively as before. You won't even see other artifacts like "macroblocking," because HDMI doesn't use compression. If something's the matter, you're going to see plain signs of it, and if you do, you very likely do need another HDMI cable. These signs of HDMI cable failure, in order from least-severe to most-severe, are:
-- (1) "Sparkles," which are individual pixel dropouts, where the pixel data didn't get through. These usually will be white but can appear in other colors as well -- they're a bit like what we used to call, in the old analog-TV days, "snow."
-- (2) Line drop-outs. This is the same as "sparkles," but after the beginning of the dropout a significant number of consecutive pixels fail, leaving a horizontal streak.
-- (3) Intermittent, flashing picture -- this tells you that you're losing so much data that the receiving circuit is having trouble even syncing up with the signal.
-- (4) No picture at all.
In such cases, it doesn't mean the cable is "broken" -- turn the resolution back down, and everything will be fine -- only that the high-frequency losses in the cable are too high to let you get a good picture. You might suppose that these things will happen whenever you try using a "Standard" HDMI cable in a "High Speed" HDMI cable setting. That seems like a reasonable assumption, but in practice, it's frequently not the case. There's often enough headroom in sending and receiving circuits that the cable can be somewhat below the specified standard. We have often seen "Category 1" cables handle "Category 2" signals over considerable lengths -- sometimes as much as triple the Category 1 certified length for that cable -- without problems. However, as the bitrates go up, this headroom factor does seem to have greatly diminished. We are finding that the practical limits for 18 Gbps HDMI 2.0 signals and still-higher-rate HDMI 2.1 signals are normally quite close to the certified limits.
It is, in fact, pretty likely that you may indeed need a new cable if/when you start using equipment that makes full use of HDMI 2.1's extended bandwidth. But with the information above, you can be the judge of that yourself; don't let anybody push you into buying a cable you don't really need. As it is so often in life, you're best to err on the side of hanging on to your money.