Hacker News Reader

Comment by matwood | Hacker News Reader

I think the point of the "either works or it doesn't" is a comparison versus analog. With an analog signal you could get slightly better results by getting better cables (or worse with worse cables). If you have a poor digital cable it flat out doesn't work. If the cable you're using works, buying a more expensive cable isn't going to make it work any better. Either the 1s and 0s made it from the source to the destination or they didn't.

No one is claiming that there isn't engineering that needs to happen to make a digital signal reliably travel from the source to the receiver.

View on HN

That's still not correct.

For HDMI in particular, there's no error correction at all on the video data (though there is on the audio and control data).

So if one bit gets flipped in the video data as it goes through the cable, you're going to see that wrong bit on screen. You'd have to have an insanely high bit error rate for your eyes to notice it, but it's there, and it gradually gets worse as the error rate increases.

What could flip a bit in a cable? Any number of things. EM interference from outside a badly shielded cable. EM crosstalk between conductors inside a badly designed cable. Signal reflections at badly designed connectors. Basically anything that can happen to any other cable, whether the signal that goes over it is considered to be analog or digital.

I assume the reason HDMI applies EC to the audio and control data is that errors in those channels are more noticeable to the user, whereas errors at reasonable rates in the video will mostly sneak by.

The HDMI spec at http://www.hdmi.org/download/HDMI_Spec_1.3_GM1.pdf is a great source of information about this.

I've had this happen. You see what are called "sparkles" when you get a lot of failures (and it truly looks like sparkling glitter on your TV). It turned out to be the TV's connector more than the cable itself. The same $4 cable worked fine on a replacement set.

Indeed. However it uses TMDS [1] which is a "blind mans" error reduction technique.

1. http://en.wikipedia.org/wiki/Transition_Minimized_Differenti...

Totally true. HDMI takes many steps to try to reduce errors to a reasonable rate. In practice, this rate is usually low enough to ignore, even with cheap cables.

I was just making the point that there is indeed a progressive degradation in image quality with increasing error rate, even in HDMI.

IMHO, seeing sparkles on the screen means 'does not work.' Hearing audio errors means does not work. When pushing a digital signal around if you have significant errors you'll know quickly at the receivers end if the cable is bad. If you hook up your cheap cable and don't notice any of these errors, getting a better cable isn't going to make the picture sharper or sound clearer like in the days of pushing analog signals around.

Clearly you've never had an Ethernet Cable that got caught in a door. Or was terminated poorly. Or doesn't have the insulator sheath under the crimp. Or run parallel to a set of Florescent light ballasts. All of these events can reduce performance on Ethernet without causing it to fail outright.

There is a reason why every single TIA/EIA 568 cable plant is carefully tested for things like Near End Cross Talk (NEXT), Latency, Skew, etc..

There is also a reason why there is an entire _discipline_ dedicated to ensuring that the "Digital" signal not only has a channel, but a meticulously crafted one. (See: http://www.linktionary.com/t/tia_cabling.html for details)

To reiterates the OP - there is no "Digital Signal" - signals are analog in nature, and they need an analog channel to transmit them. That analog channel needs to be engineered carefully to get the lowest EBR (Error Bit Rate) possible.

What's interesting is the Failure case of the Digital CODEC on the other end - some may reduce transmit rate in the presence of an increased EBR resulting from dropped packets. Or some may seize up and cease to function. In the degraded case, having a pristine analog channel from the transmitter to the receiver will certainly result in a better digital signal.

There is also the issue of future proofing - We're installing an entire plant of Category 6a cable right now in one of our data centers for a gigabit environment - Category 6a is like the "Monster" version of Category 5e/6 - It seems like a douchey way to do a network install, except when you consider the following - (1) this cable plant will be in use for 10-15 years, (2) We're testing a bunch of Nexus 5010s that will likely make good use of that 6a for 10 Gigabit. So, in addition to passing _today's_ signal flawlessly, it will also pass tomorrow's. This might be an important factor if your HDMI cable is being run through hard to access cabinets/walls.

Now, with all that said, it is still probably the case that the $150 Monster Cables (under normal circumstances) don't have a lower EBR than a $10 cable, but, in the face of RF interference, It may actually be a better performer. I don't know, but I wouldn't discard the possibility without a bit of research and education.

Anything running over Ethernet has packets with checksums. Yes, noise will cause packets to drop, and that results in lower overall throughput, while connectivity still remains possible.

The signal running over HDMI is not packet-based. You don't have any mechanism like with TCP where packets can get dropped and resent later. So if interference causes bits to flip on an HDMI connection, you will immediately see the interference on your TV, and it won't be pretty; it would probably look something like an HDTV antenna signal that loses reception--missing or discolored chunks of the screen or sparkly pixels. The effect is immediate and disruptive.

The end result is that Monster Cable might do a better job of keeping your picture intact during a freak electrical storm, but there is no practical difference otherwise. If you are able to see a picture with HDMI, and you don't expect any wildly varying levels of RF interference in your environment, your cable is as good as anything else.

Again, nobody is disagreeing. This isn't a defense of Monster, it's a typical It's More Complicated Than That jibe. Don't buy Monster unless that flowchart checks out, but also realize that it's not impossible to see signal degradation on those cheaper digital cables.

It's just a lot less likely.

A $10 cable should fail before a $250 cable, however the frequency of failures for an infrequently moved or manipulated cable should still give it a longevity of a few years.

A $10 cable wont fail 25 times within the lifespan of the technology it's attached to. HDMI won't exist before you get 25 failures, unless you enjoy hammering nails into you cables or some other bizarre behaviour.

I think most people will gladly pay more for a better cable, but they'd like to pay something comensurate to the engineering effort, and production and material costs. The markup is insane (nearly $200 apparently).

One has to wonder if Monster and their retail partners would make just as much money selling them at $50, due to higher volume, with the added bonus of people not totally misleading everyone with the value proposition.

I spent about $5 more on my HDMI cables than I would have to get ones with a heavy-duty rubberised outside instead of the cheap and tacky bendy-plastic that breaks easily from over-manipulation.

You are right, people will pay more for better quality but misleading people into believing something is better quality by outright lies actually violates most consumer protection laws. I've seen no $250 monster cables at best buy here in ontario, the most expensive I saw was a $60 10ft cable by monster, however the consumer protection act is very strong and very strict here. Best Buy has screwed me over before and the mention of the CPA to a manager gets the problem resolved quickly. It appears they know they're in the wrong in many cases, but people let them get away with it - this is usually on their extended warranties from my experience.

You're confusing what I was saying. In an HDMI situation if you use a cheap cable and the picture and sound are fine, buying a more expensive cable isn't going to change anything. With HDMI in particular, if the cable is failing you'll know right away with either a poor picture or sound. It won't have the gradations that were possible when pushing an analog signal around - i.e. the picture looks could, but it could be sharper with a better cable.

I'm not sure why you brought ethernet into the mix, but generally you're talking about much longer runs than HDMI and it's often not as easy to see when your are losing packets without specifically checking if you are dropping packets (IMHO poor network performance IS failing outright, but not as easily seen as a bunch of blocks on the TV). Even then, the cable you use is either delivering all the packets or it isn't (up to an acceptable level of errors/loss for the length you are running). Delivering a stronger 1/0 isn't going to change anything as long as it's strong enough for the receiver to differentiate. That's where it either works or it doesn't comes from.

But did you use audiophile, linear crystal aligned cables? And did you install them the correct way round? https://usa.denon.com/ProductDetails/3429.asp

Presumably on ethernet you have to decide if you want the Tx or Rx signal to have the extra digital nuance?

"Or run parallel to a set of Florescent light ballasts."

Really? Crap. I've just run the cable down to a back room (for an x360) - and it goes straight alongside a couple of flouro tubes.

I think your understanding of cables is off.

All signals are analog. Every single one. (Actually, I take that back. With the exception of quantum computers, to date, all signals are analog).

A "digital" cable doesn't just work or not work. It is possible to receive spurious bit errors for many reasons. For example, you turn up the volume to your awesome surround sound speakers, which causes more current to flow through your speaker wires that happen to be running parallel to your horrible HDMI cable. The increased current causes interference in your low-power HDMI cable, causing bit errors in the transmission.

>> "It is possible to receive spurious bit errors for many reasons."

Surely that's why there's error correction layered on top?

Instead of arguing about what might happen, why not find an HDMI device that'll show you how many unrecoverable bit errors it has detected in the stream... Can't be that hard to do.

If my hard disk can send data over a tiny crappy unshielded cheap USB cable without a single bit error, I'd be willing to bet money that an HDMI cable isn't that much different.

HDMI doesn't do error correction for video.

ugh, I hated signals theory at uni... but let me dust off that knowledge...

> All signals are analog.

I see your reasoning in saying this ("on the wire" etc.); your incorrect though. Not all signals are considered analog (which is important to the point being made)

Analog signals are generally susceptible to noise because they are usually non-recoverable; error checking is hard because it is a non-discrete continuous signal.

>It is possible to receive spurious bit errors for many reasons.

Yes, which is why error checking exists in all these systems. In the context of the wire alone, yes, it is not a case of "it just works". But in the case of the communications subsystem (i.e. the transmitter, medium, receiver) it really does either work or not work :)

Ultimately thought the phrase "it's digital; either it works or it doesnt" is making no comment about the actual signal. It's really saying: an expensive cable is not, due to the digital sub-system, likely to make any difference to error rates (or more succinctly: the performance of the error correction is better than the noise affect of the transport medium)

Isn't HDMI supposed to do error correction on both sides, though? I could be totally making that up, I thought I remember hearing that a while back.

There is this weird thing called error detection/correction.

which HDMI doesn't do.

Hm, no idea, but I find that very hard to believe. I suspect digital transmission of information is almost impossible without error correction. IANAEE, though (I am not an electronics engineer).

HDMI does have error correction. Every 8 bits of video data are sent as 10 bits on the wire.

Note, it's traditional on HN to add the phrase [EDIT] when making substantive modifications to comments, particularly if someone has already responded to your original comment.

OT: The site should really help to enforce such niceties. When you edit you should get a "reason for editing" box.

Reply on news.ycombinator.com