Well, thanks for being insulting. Actually, Blue Jeans Cable is the only brand in the world to use bonded pairs for high impedance stability and low skew -- that's technology you can measure, and we see it when we look at bonded pair Cat 6 and 6a data pairs as well as in the long certification lengths for our HDMI cables. Far from being a "$3 cable," our bulk HDMI cable is (I'm reasonably sure) the costliest HDMI cable stock in the world to manufacture as well as being the only stock that is US-made. Belden can't sell it to anyone else, because nobody wants the cost bite; we take the cost bite and still turn the cable around to the consumer much cheaper than boutique brands that don't have the technology, and that just rely on Chinese manufacture, do.
Now, I don't know how well the WW people have been able to control skew and impedance stability using this four-conductor design but I can tell you that such an approach is fraught with difficulties as it introduces new considerations into the already difficult problem of dimensional control. I can't find any published specs on their cable so I don't know what they claim; and I have found that it's chancy to rely upon people's published specs in the absence of test data to confirm. I can tell you, having been involved in some of these R&D projects (why do you assume we haven't? For crying out loud, a lot of R&D both on our part and Belden's went into bonded-pair HDMI.), that the translation of a theoretical spec to a working, manufactured cable that exhibits the specs projected for it is really tough. I would not want to have to do that with a Chinese wire factory. Every aspect of production becomes critical when you get up into the multi-Gigabit range.
Now, "speed." I have no idea what you mean. You don't want a foamed dielectric, which will increase propagation speed, because at these frequencies and in a paired-cable configuration the dimensional and foaming characteristics are really, really hard to control and you tend to get return loss spikes way up in the high frequencies. If what you mean is not propagation speed but data rate, the fact is that still today there is no specification for cable carrying data rates over 10.2 Gbps (three channels at 3.4 Gbps summed). The pass criteria change from Category 1 to Category 2 HDMI cable, and if higher data rates were to be tested there would likely be a Category 3 spec. This is a complicated issue, but the gist of it is that HDMI Licensing has set the 2.0 spec to employ Category 2 cable even at the higher data rates which exceed Category 2 testing, and they've created an S-parameter model they call the "Worst Cable Emulator" against which sources and sinks will be tested. The point is, at any rate, that there simply are no specs for higher data rates other than the Category 2 spec, and a Category 2 cable should pass all signals error-free under any existing HDMI spec version, including 2.0.
The result is of course that anybody can claim that he has a "28 Gbps HDMI cable." The term has no definition; nobody that has used that sort of term (and there have been several) has ever attempted, in my experience, to define what they mean by it; all cables will pass SOMETHING at 28 Gbps, and the only question is how well. To know how well is good enough, you've got to have a spec which integrates sources, sinks and cables. What I am quite sure of is that bonded pairs perform better than conventional pairs as we get up into the crazy-high bitrates. You can see it at 250 MHz (equivalent, in HDMI terms, to 500Mbps/pair, 1.5 Gbps total) on Cat 6 cable, well short of high-def HDMI frequencies.
All of this about "jitter": Well, jitter's a problem in digital signaling, for sure. In fact, when you go sticking extra conductors into the pair, you're liable to have considerably greater difficulty with intrapair skew, which really throws a wrench into the jitter problem because the two sides of the differential signal are arriving out of time; picoseconds matter here. But what must be understood about jitter is that it is a problem which arises BEFORE the receiving circuit, and if the receiving circuit is able to fully recover the data, it does not matter at all how good or how bad the jitter was. The bits come out the same, and it is indeed "all ones and zeros." Now, if you get bit errors due to jitter, of course, that's another story--you've got data loss, and as HDMI has no way of correcting the error that can easily be visible or audible.
Now, audio is not carried separately on HDMI cables, as many people think (except ARC audio, going back from a TV to a receiver). It's embedded in the video data stream. The result is that audio depends exactly, and only, on the same data stream upon which video depends. You can have a cable deliver perfect video, and if it does, it will deliver perfect audio. If a cable delivers bad video, it will deliver bad audio. The two ride together and the principles governing fidelity for the one are the same as for the other.
There's always going to be a lot of hucksterism where cable is concerned, and the dry and dull features of a well-engineered product are always going to sound less exciting than the extravagant claims associated with boutique products. To understand what's true, you've got to have a detailed appreciation of this stuff at the technical level.
Kurt
Blue Jeans Cable
