Hacker News Reader

I am talking about fixing everything from "my colors look wrong" to "I have no picture!" with this recommendation.

"I have no picture" will definitely be fixed with a better cable.

"My colors look wrong" will not, and cannot, be-- for the reasons discussed in the article.

Well, it could be fixed due to the placebo effect.

'fixed' :)

my colors look wrong == sparkles

Yep, seen the "sparkles" issue as well a handful of times and was corrected with a new cable.

> "My colors look wrong" will not, and cannot, be--

Believe me, I would agree with you had I not been troubleshooting the issue for the last 4 years, but it did fix it.

It is entirely possible the guy ended up changing more than just the cable and didn't mention it.

This experience reminds me a lot of the Google Ops Team talk at Google I/O this year where they gave stats on things that caused outages and kept punctuating the point that things you think are impossible, will happen at-scale.

"HDMI is a digital signal, it either works or doesn't, degraded analog signal artifacts are impossible!"

Apparently not =/

Never underestimate the power of the placebo effect.

I'd also like to say that digital signals should work or not work, however, when someones map doesn’t match the actual terrain, it’s usually not the terrain that’s wrong.

Could it be related to how the cable affects the transmitting or receiving circuitry? (Electrical impedance mismatch or crosstalk comes to mind.)

And this "it's digital" argument is bogus. The individual bit is either transmittes correctly or not, but the whole stream can have arbirary many bit errors which the receiver can only compensate up to a certain rate. There is usually good error correction on the audio streams (both listeners and speakers don't react well to sudden loud clicks), but errors in the video stream usually lead to artefacts that may be visible for are few fractions of a second or so (until the next full frame in the streams).

Whoever makes this "digital streams cannot degrade" argument has obviously never tried to watch DVB-T behind a hill.

DVB != HDMI.

I believe DVB-T uses MPEG-2 or some variation. While HDMI video doesn't use a motion vector based encoding. So when you have errors with MPEG-2, it could be on B or P frame which results in the delta from the previous I frame being drawn incorrectly (or not at all). With HDMI, the error won't show up in a delta frame (because there aren't any), rather just white pixels.

Do you understand digital vs. analog? If yes, then in fact you do agree with the message of the article: a non-defective cable transmits a digital signal just as correctly as any other non-defective cable. Your experience just indicates that slightly more expensive cables are less likely to be defective.

I wish HDMI equipment would display a counter of digital errors. For all non-defective cables, this counter should be 0 all the time.

I wish HDMI equipment would display a counter of digital errors. For all non-defective cables, this counter should be 0 all the time.

Does the HDMI signal provide a checksum of any kind? I was under the impression that it's just a raw TMDS bitstream of color values with no redundancy.

I read the specs myself and found out that:

  * control packets (non-video, non-audio) are protected using TMDS Error Reduction Coding (TERC4)
  * audio packets seem error-protected at a higher level (by Dolby or DTS?)
  * video packets are *not* error-protected

Wow. I am shocked. How can anyone design a high-speed video digital transport layer these days without error correction? Source: http://arstechnica.com/civis/viewtopic.php?p=6271817#p627181...

Good call on the display of errors...why is that not available in today's TVs?

Or is it? Maybe in some special diagnostic mode?

Here's what the article is saying - HDMI is like the internet, or digital TV; not like radio.

If your internet connection is working, HN looks the same with a good connection or a bad connection. If your internet connection is down, you get nothing. 0.01% of the time, you might get a garbled page from a bad connection, but it's not a common problem.

But people think that the cables are like radio antennas, and a better cable = better quality picture.

Digital stuff is actually far more annoying to fix. If your TV is black (not fuzzy, like it would be with a bad analog connection), you don't really know if it's the cable, the port, the video card, or something else.

> you might get a garbled page from a bad connection

But if you do, it isn't because the internet is bad and needs to be replaced (the cable in the analogy), it's because either the server or the client interpreted the signal incorrectly, and tried to display the page as correctly as possible, but failed.

Yes. There are poor quality cables in the world, and using them can cause problems like intermittent signal drop or failure to detect the display. And this is no more or less true than it is with other cable types.

The real problem though is that in my experience cable build quality simply doesn't correlate with price. All the cables are built at the same handful of east asian factories, and subject to the same variance. If some component supplier gets swapped (more brittle casings for the wire, say, or a different plastic formula in the injection molds) then you end up with a cable that will fail unexpectedly or intermittently.

It's worth noting that you can't really spend $18 on a single HDMI cable at monoprice unless you are getting a super long one. No one should ever spend more than $7-10 on an HDMI cable.

For fun and profit: What I do when I order from monoprice is buy five or so at a time, which equals the price of a B&M cable, and then I sell my extras to friends for a buck or two over the monoprice cost, and then my cable is effectively free.

It seems to me that this is what the article is saying, actually. Really cheap cables may very well be faulty, but $10 cables are just as good as a $90 set.