What's the Matter with HDMI?

M

MDS

Audioholic Spartan
Let's look at the 'worth' of HDMI from the perspective of the consumer.

Say you belive that HD-DVD/BluRay will survive and the future will also bring 1080p native content to cable/satellite and you want to purchase equipment that will allow you to view such content.

Now also sssume that you are someone who has a need for a very long cable run (even if you are perceived to be in the 'minority') and also accept the fact that maybe HDMI wasn't designed for your needs. No problem you think - just use component video which is perfectly capable of carrying 1080p60 over great distances.

But it IS a big problem:
- All of the new formats require HDCP due to the irrational copyright 'concerns' of the content producers and that requires HDMI.
- With few exceptions, dvd players that can upconvert do not do so over component.
- Most TVs that can accept 1080p will only do so over HDMI (although that is slowly changing).

So basically you are screwed because 'HDMI wasn't intended for long runs' even though everything REQUIRES its use. Add to that the fact that it isn't all that reliable even over shorter distances and I think it's pretty safe to say that Kurt is correct in his assertion that it was a standard produced for the content providers and not the consumer - regardless of whether you believe that UTP is adequate for carrying HDMI signals.

craigsj: CAT6 requires certification to 200 MHz. You say there are CAT6 cables certified to 1000MHz (1 GHz)? Are you sure you aren't confusing the bit rate with the clock rate or do such cables actually exist?
 
jonnythan

jonnythan

Audioholic Ninja
You guys made the home page of Slashdot.

I'm surprised the server is doing so well.
 
K

KurtBJC

Audioholic
and then there are twinax cables.
Which, of course, are not coax. Controlling impedance and skew in twinax is just as difficult as in other twisted-pair cables.

The complaint here seems to be that twisted pair can't offer sufficiently constant impedance. There is Cat6 cable certified to 1000MHz, so citing Cat5 being certified to only 100MHz as evidence that twisted pair isn't adequate certainly is misleading if not outright fraudulent.
Oh, for goodness' sake. I didn't indicate that CAT5, or similar constructions, can't be used at higher bitrates and frequencies. In fact, I have said nothing negative about the use of CAT5, 6, etc. in network cabling. But this is a different application, and it works really quite badly.

I'd rather not pay test probe prices for HDMI cabling.
I don't know what you're talking about. I haven't suggested that the cable best suited for carrying HDMI must be costly. In fact, in about a month we'll be introducing our Belden bonded-pair HDMI cable; it will be demonstrably technically superior to anything currently on the market; and it will be cheaper than most of what's currently available. If the HDMI standard had been done using unbalanced coaxial lines, the cable would be less expensive still, and easier to make, to boot.

By quoting heresay from unidentified sources whose expertise is unknown and unconfirmable, you have totally failed to do that.
Oh, fine. David DeSmidt and Martin VanDeBurgt, both at Belden, would be two of the engineers who've remarked on the inappropriateness of the choice of coax.

As for your claim that you have not purposefully misled, are you saying that you are unaware that there are Cat5+ certifications well in excess of 100MHz? You do realize that GigE runs on twisted pair, right? You realize that Infiniband does, that 10GigE will, etc? SATA? SAS? Future PCI Express? It is a joke to suggest that twisted pair isn't up to the task, and anyone who claims to be expert yet says so deserves to have his motives questioned.
But again, I have not said twisted-pair can't be used in high-speed data applications. I have said it was a poor choice for HDMI, and I have stated my reasons. And those issues which I have pointed to, and for which coax would provide substantial relief, that is, skew and return loss, are in fact the major problems with getting HDMI to perform consistently over distance.

As for lack of error correction, that is not a twisted pair issue, and the design choice understood that the data stream was not critical in nature.
Error correction is not a twisted pair issue. But if you're not going to provide error correction, using cable with relatively poor impedance control is a good way to produce dropouts. Knowing that the signal would be run real-time and without error correction would guide a reasonable designer to the conclusion that control over impedance and timing were too critical to rely upon twisted pairs.

I disagree with your assertion that 50 foot runs are "very common".
And you are wrong. We sell a lot of 50-foot cables. And 45, 40, 35, and 30, as well. All of these lengths are very popular and are in common use, and it is a challenge to meet spec at these lengths.

Since when did HDMI ever promise to support the needs of your church users? What were they happily using before they were stabbed in the back by the false promises of twisted pair?
We have lots of customers who have run analog component video 200 feet or more with excellent results; they haven't even needed booster amps at those distances. Analog video is phenomenally robust; and digital video, designed right, would be more, not less, so.

What is the volume of equipment today offering 16 bit 1080p?
Zero. But the standard calls for it, and it's coming. I know that there are a handful of 12-bit 1080p devices on the market now.

Why don't you simply write an article explaining that you don't like that HDMI has a 50 foot limit? That's what it all boils down to after all.
The limitations on effective distance runs are a mess; they were completely avoidable; and that is, indeed, a significant part of my complaint. I am joined in that complaint by many, many people who have had trouble getting HDMI to perform over distance.

You are welcome, of course, not to care that the HDMI standard is so fragile that it doesn't work reliably at distances which the designers could easily foresee users would ask it to work at. But the frustrated users do care, whether you do or not. It is not unreasonable for people to care. What I find puzzling is that it seems to make you very angry that other people do care about this issue.

Now, apart from writing articles, I am actually DOING something about it. As our Belden HDMI update article describes, we actually have, by using Belden's patented bonded-pair technology, managed to make an HDMI cable which, in our in-house use tests, has been capable of carrying 720p and 1080i video flawlessly for a distance of 150 feet on a purely passive cable. This cable was designed principally by David DeSmidt, referenced above (one of those engineers, as you'll recall, who has stated outright to me that HDMI should have been run unbalanced, on coax), and in the meantime we are resolving manufacturing issues in an effort to make the cable economical to terminate in the U.S. rather than China. Nobody here is merely griping and whining; we are solving the problem, and trying to deliver a product to the consumer at the lowest practical cost. But the problem did not need to exist in the first place.
 
gene

gene

Audioholics Master Chief
Administrator
You guys made the home page of Slashdot.

I'm surprised the server is doing so well.
Actually we have FIVE (5) dedicated servers running Audioholics now and they seem to be properly configured. Bring on Slashdot, Yahoo or any other big dogs, we can handle them :D

BTW, our servers are not connected with any HDMI cables :eek:
 
no. 5

no. 5

Audioholic Field Marshall
The complaint here seems to be that twisted pair can't offer sufficiently constant impedance.
If I may, I think that the complaint is that HDMI itself is a flawed interface, the fact that the cable design seems to be adding to the problems just fuels the desire for a better system.
 
DavidW

DavidW

Audioholics Contributing Writer
Mr. craigsj seems to be spending a lot of discourse impugning KurtBJC's integrity and dignity by harping about the basis of his engineering knowledge.

Would Mr. craigsj please tell us where he obtained his engineering degree? Would he also be interested in telling the forum members how long he has been in practice and in what specialties?

If he is in fact an engineer, that's great, but what he believes in is still just an opinion, educated or not.

If he can't, then he should not use such arguments as basis for criticism of others for the same.

Let's all play nice, can't we?
 
MACCA350

MACCA350

Audioholic Chief
Now, apart from writing articles, I am actually DOING something about it. As our Belden HDMI update article describes, we actually have, by using Belden's patented bonded-pair technology, managed to make an HDMI cable which, in our in-house use tests, has been capable of carrying 720p and 1080i video flawlessly for a distance of 150 feet on a purely passive cable. This cable was designed principally by David DeSmidt, referenced above (one of those engineers, as you'll recall, who has stated outright to me that HDMI should have been run unbalanced, on coax), and in the meantime we are resolving manufacturing issues in an effort to make the cable economical to terminate in the U.S. rather than China. Nobody here is merely griping and whining; we are solving the problem, and trying to deliver a product to the consumer at the lowest practical cost. But the problem did not need to exist in the first place.
Great to see,
I am currently using a 50 foot run of component cable(never had any issues) and when I upgrade(hopefully very soon:D ) to a 1080p projector and High-Def player I'll be needing a 50 foot HDMI cable to transfer a 1080p signal.

I'd be rather pi$$ed off if a standard HDMI cable has problems with this since the players won't output 1080p over component, and it seems that I may run into these problems.

cheers:)
 
mike c

mike c

Audioholic Warlord
haha, even though I have no projector, no HD player, no nothing ... I have already bought a 40 foot cable from BJC :)

I bought a 30 footer from them some time back for my dedicated HT#1, and it's one fat cable.
 
K

KurtBJC

Audioholic
David DeSmidt and Martin VanDeBurgt, both at Belden, would be two of the engineers who've remarked on the inappropriateness of the choice of coax.
Oops...what I meant to say, of course, was "the inappropriateness of the choice to use twisted pair rather than coax."
 
K

KurtBJC

Audioholic
Btw, the article mentions new HDMI cables coming out by BlueJeans last month or this month. Would that be the current BJC Series 2 cables mentioned on this page?

http://www.bluejeanscable.com/store/dvi/index.htm

Just wondering. Thanks!
My apologies for not picking up on this sooner. As I think I may have mentioned in another of the responses here, the Belden bonded-pair HDMI cables are expected to be in stock and available for sale somewhere around the very end of June or the first half of July (more likely the latter than the former). The current Series-2 HDMI is, like all of the other HDMI cables I've ever encountered, Chinese-made. The Belden bonded-pair cable will be called the Series-1. I am awaiting some testing reports and what-not, and hoping that people respond well to it.

Our next task, and it's a tricky one: we are trying to work out how to do all of the final assembly of these cables in the United States, rather than shopping out termination to China. For the time being, until we can work out a solution to that problem, we actually have the cable shipped from the point of manufacture in Richmond, Indiana to Seattle, then we put it in a container, send it to China, have it terminated, and have it shipped back. Bringing the final assembly home is likely going to require that we have a connector custom-made for us; right now, most of the connectors are very labor-intensive to prepare and attach, and that makes sense when you consider how little the people who work in Chinese factories make. For us to have a competitive all-American solution will require a connector that's designed for ease and speed of termination, and with any luck, we'll be able to put together the right combination and bring this project completely in-house. It will never be as cheap to assemble one of these in America as it is to assemble in China; but if we can get anywhere close to the mark, we'd rather build 'em here.
 
pepar

pepar

Junior Audioholic
Our next task, and it's a tricky one: we are trying to work out how to do all of the final assembly of these cables in the United States, rather than shopping out termination to China. For the time being, until we can work out a solution to that problem, we actually have the cable shipped from the point of manufacture in Richmond, Indiana to Seattle, then we put it in a container, send it to China, have it terminated, and have it shipped back. Bringing the final assembly home is likely going to require that we have a connector custom-made for us; right now, most of the connectors are very labor-intensive to prepare and attach, and that makes sense when you consider how little the people who work in Chinese factories make. For us to have a competitive all-American solution will require a connector that's designed for ease and speed of termination, and with any luck, we'll be able to put together the right combination and bring this project completely in-house. It will never be as cheap to assemble one of these in America as it is to assemble in China; but if we can get anywhere close to the mark, we'd rather build 'em here.
Not having them terminated in China will increase the time that it will take them to knock off your idea, but probably not for long. :(
 
K

KurtBJC

Audioholic
Not having them terminated in China will increase the time that it will take them to knock off your idea, but probably not for long. :(
You know, that did occur to us. But this is a pretty hard product to duplicate, for two reasons.

First, the bonded pairs are the subject of a Belden patent. There's not much intellectual property law in China, so the Chinese, if they could duplicate it, could probably get away with marketing it over there. But if they exported it to America, it wouldn't be long before the incoming shipping containers were being seized and destroyed by customs.

Second--and this, I think, is the big, big barrier--just knowing that the trick is in sticking the pairs together isn't enough to duplicate the product. Doing something like this requires some pretty burdensome in-process controls, relating to everything from handling the wire to preparing the dielectric to the method of extrusion and jacketing. Just grabbing a run of telephone wire and a pot of glue won't get you a high-performance bonded pair. Dimensional control in a product like this is exceedingly difficult; when I visited the Belden engineering lab and factory, one of the things that really impressed me was the sophistication of the systems for controlling dimensions; they can actually electrically monitor the characteristics of the dielectric as the wire flows through the extruder (even though there's no shield on the cable yet!), and adjust any parameter on-the-fly to keep the product within a tight tolerance. The traditional way to handle things was to control dimensions the best you could, and then sweep-test the resulting product (they still sweep test--but at Belden, they pretty well know the result of the test beforehand because of the process controls); when a million feet of cable fail a sweep test, that's a pretty expensive goof.

One of these days, maybe, the Chinese will catch up. But right now, the money for them is mostly in making high-volume, less quality-critical products, and they don't have a lot of high-quality capabilities. That will change, no doubt; but by that time, their wages will have gone up and they'll have lost the competitive advantage of cheap labor.
 
W

westcott

Audioholic General
Sign me up for a 50 foot run of the new Belden HDMI cable. I don't live in a church but I still need a long run.

Kurt, if I did not know any better, I would swear craigsj is "mas" in disguise!:eek:

I guess there a lot of people out there that think they can run an unbalanced cable with shielding that is not properly connected at both ends.
 
K

KurtBJC

Audioholic
Sign me up for a 50 foot run of the new Belden HDMI cable.
It's looking like we'll have these in stock around the first week of July; I'll try to remember to drop you a note when that happens. If I forget, though: just stop by our site, and we'll have them available to order as soon as they show up.

As for all of the criticism...you know, I have found something odd in the reaction to HDMI. I write articles on all sorts of audio and video cable issues, and while people quarrel with me on one point or another, for the most part the debate, when there is one, is in a friendly spirit. But something about HDMI gets people really emotional, and particularly hostile.

I've been trying to guess at just why that is. I think there are two elements, perhaps. First, I have encountered people who seem to consider themselves early adopters; they have an emotional stake in HDMI because they've read a lot of material about it from the promoters of the HDMI interface, and they've bought in to that to some degree. When someone says something bad about it, they get defensive (especially if they've spent $150 for a cable at the store!).

The other thing I've run into is people who have a computer network-specific technical background. These people sometimes don't have a good general grounding in electronics, and rarely know anything about transmission line theory, but they know that twisted-pair cable works well in the applications with which they work, and they feel that an argument against the HDMI cable design is an argument against data cabling generally, which of course it is not, or that the argument is negated by the success of twisted-pair cable in various data applications. What these people fail to appreciate, I think, is that twisted-pair cabling, despite its limitations, works in the various high-speed data applications to which it is put BECAUSE engineers like the people responsible for the heavy-lifting work on our HDMI cable have spent a lot of time and sweat working out the problems. If the engineers need the cable to have different twist rates on different pairs to minimize crosstalk, for example, they will work out how much skew that's going to cause, and they design the protocol around the fact that there will be a certain amount of unavoidable skew. If they need the cable to have zero skew, they account for the problems with crosstalk that that'll cause. And they set specs for cable length, for signal strength and quality, and for quality of data recovery, and, importantly, for error correction when something goes wrong, that all work together to produce a reliable interface. This all makes it easy to build a network, or what-have-you, around these technologies without having to wonder whether it'll work; and that ease of use translates into the comfortable illusion that the whole system was trouble-free to begin with. So, if I suggest that in a system where none of those things have been worked out satisfactorily, cheap Chinese twisted-pair cable is an inappropriate choice, I'm an idiot, a liar, a scoundrel, or whatever the epithet-of-the-day is.

I really have never minded people disagreeing with me about things. But it is disconcerting, sometimes, to have the kind of temper-flaring displays that result when the fragility of HDMI is the subject of discussion.

Ach!

Kurt
 
jonnythan

jonnythan

Audioholic Ninja
"So, if I suggest that in a system where none of those things have been worked out satisfactorily, cheap Chinese twisted-pair cable is an inappropriate choice, I'm an idiot, a liar, a scoundrel, or whatever the epithet-of-the-day is."

Maybe you're a liar et al. because the cheap Chinese twisted-pair cable works perfectly well for those people ;)
 
K

KurtBJC

Audioholic
Maybe you're a liar et al. because the cheap Chinese twisted-pair cable works perfectly well for those people ;)
You know, that's a factor, too. People who find it working well wonder what all the fuss is about, and as often as not, they're running modest lengths of cable. The funny thing is, we actually take pains to tell people the truth about this: if the cheap cable seems to work well in your application (no "sparkle" dropouts, no loss of sync, etc) then it's working perfectly; and no improvement in cable quality, at least with your current setup, will make any difference to your picture. If we actually WERE going to be dishonest about this stuff, and make up some line about how better cable is going to improve something-or-other, we might sell more cable. But in the long run, we figure that we sell more cable because of the credibility we earn from telling people the truth than we could sell by taking that sort of approach.
 
F

FantasyGhost

Enthusiast
Reading the article i just have to ask. Ok here goes;

There's this company that provides internet and telephone over twised pair cable. Only a few years ago they went into the buisness to provide their clients with digital television over that same cable. Now they're trying to get it over vdsl 2. They're trying to provide the whole country with optical cable. But still the cables going into the homes is twisted pair. Now my question is;

Will they ever get to HD TV?

Because when i'm reading this article i'm not convinced they ever will. I mean they already have problems to get even digital television working in whole cities. And the bitrate they're sending from some of the channels, you can see it's so pixelated. Chanching channels is so slow etc...
 
M

MDS

Audioholic Spartan
There's this company that provides internet and telephone over twised pair cable. Only a few years ago they went into the buisness to provide their clients with digital television over that same cable. Now they're trying to get it over vdsl 2. They're trying to provide the whole country with optical cable. But still the cables going into the homes is twisted pair. Now my question is;

Will they ever get to HD TV?
That's a bit different than what is being discussed here. There is no problem with UTP for voice/data and it is widely used. As speed increases, the requirements go up and that is why we now have CAT6 specs and even CAT7 is being discussed (although it probably won't see the light of day for some time as it isn't really necessary just yet).

So there is no issue with getting HD video and audio to the home. The issue is once the data arrives at your home there are now multiple devices connected together using HDMI and there is now a whole lot more data to shuffle around - the bits that are the actual audio/video signal, all of the control signals, HDCP copy protection protocol with exchanging of encryption keys to verify 'authorized' devices, etc.
 
F

FantasyGhost

Enthusiast
Thanks for answering.

The thing is they ain't using a CAT5/internet cable, they're using a 4 wired cable (blue, white, purple, marine blue) in most homes.
 

Latest posts

newsletter

  • RBHsound.com
  • BlueJeansCable.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
Top