HDMI Part 8 - Cables for 1.3

[Rodolfo]

Now this is a sensitive area, isn't it?

The purpose of this article is NOT to justify or reject the concept of spending a dollar more for "claimed" cable quality while other people rather want to save that dollar any time the term cable is mentioned.

People take all kinds of corners on this matter, and many confrontations still happen with or without blind tests, with or without factual data.

[url=http://www.hdtvmagazine.com/articles/2006/08/hdmi_part_8_-_c.php]Read the Full Article[/url]

[mstockfisch]

Last year, PC WORLD did some objective cable testing with a little help from Pixelworks, Quantum Data, Tektronix, the Imaging Science Foundation, Silicon Image, Rohde & Schwarz, Pioneer Electronics, Epson, Denon, and Audioholics.

The results can be found at:

http://www.pcworld.com/article/id,12177 ... ticle.html

[Rodolfo]

mstockfisch,

Thanks for the link. There are dozens of reports/reviews/comparisons about cables that contradict each other with apparent facts and blind tests, most with an especific agenda.

However, the PC World article did not mention the HDMI issues I discussed on my article, such as different lenghts of D-D+ within the pairs, thickness findings on HDMI, 1080p capabilities, category levels, etc. Probably because HDMI Licensing did not detect those last year and Simplay Labs did not even exist.

The first comment on the article is already stating that there are differences in cables that could affect the sound or image:

"While conventional wisdom says a wire is a wire, a poorly constructed cable can seriously degrade video quality. Used with a digital connection, a bad cable can mangle pixel data, resulting in a speckled image; with an analog link, it can distort the signal, resulting in faded colors or blurred details."

Their final comment was also interesting:

"..you are unlikely to detect any difference in picture quality between a cable with a moderate price and a luxury brand"

However, what is their definition of a moderate price cable?

The one that comes in the box of the equipment and most people use?, could not be, is free, for a reason; the one would buy at Radio Shack or Home Depot? the mid-point line of famous cable manufacturers?

That goes along with my comment of selecting the cable based on the right return for the investment, of the cable cost and relative to the overall investment of the HD system that could potentially be affected is the chosen point of return is low, a common tendency among consumers satisfying their red sale tag saving sport.

Regardless of the dozens of reports, it comes back to square one, the chore is back to the consumer to determine what is the point of a moderate price cable and what level of quality would that moderate cable mean for his/her overall HD system, without bringing an oscilloscope to the video store. One method many non-hi-end owners apply is to choose mid-line point on any manufacturer, or the point where the next step carries a prohibited jump in price.

Best Regards,

Rodolfo La Maestra

[mstockfisch]

Rodolfo,

I work for Quantum Data and design HDMI test equipment for a living. I personally supplied the test equipment used for the "Digital Perfection" HDMI testing.

Our equipment was used to test five cables at 1080p 60Hz (i.e. 148.5MHz). The equipment was setup to output random 24-bit pixel values at the transmitter and to compare received values with those transmitted - bit-for-bit at the receiver. All errors were counted - including temporal ones that you would never see with your eyeball. The test was conducted over the full voltage swing range of the HDMI standard (0.15 to 1.56 volts). All of the cables passed this noise test (over a period of several minutes) without a single error - so long as the voltage swing stayed above 0.32 volts. As it is very unlikely that any real product would output a signal below 0.32 volts at these lengths, all the cables essentially preformed flawlessly - regardless of price.

The "Cable Confidential (chart)" shows that the five cables tested ranged in length from 4 to 4.57 meters and in price from $31 to $300.

FYI, we ship Molex 88768-9810 2-meter HDMI-to-HDMI cables with our test instruments and nobody has ever complained. I think these cables cost about $20 (quantity one).

I have personally tested very long (50-meters)/expensive/high-quality cables through a pre-emphasized/equalized repeater and seen zero errors at 1080p 60Hz. So I can say for sure that perfection is possible for a price - using expensive cable and quality electronics on both ends.

To summarize, I think it would be safe to use just about any name brand off-the-shelf cable at 4-meters - so long one has properly designed source and sink products at either end. At longer lengths (e.g. 10-meter or greater) there may be a reason to spend more money. Here, I'd buy pre-emphasized/equalized electronics and cables with Simplay certification.

Mark

[Rodolfo]

Mark,

Sorry for the delay on maintaining this exchange but I have been quite busy lately.

Thank you for the insight on the testing results you performed for your company. It is always good to share actual experiences from professionals dedicated to the subject because the public gets the full benefit with actual and factual information, so people can invest the money properly, whether they need is short run for 1080i, or a long run for 1080p.

I discussed your comments with Silicon Image and their feedback mostly coincides, as follows:

Response from Joe Lee, Director of Marketing at Silicon Image.


I would agree that for most common HD content (720p/1080i video resolution), the vast majority of cables 4m or less should function and yield a perfect image, especially when using well designed source & sink devices.

We find that although the cable does contribute to the robustness of the HDMI digital signal, the quality of the electronics in the sink device (specifically, the HDMI receiver chip in the TV) has a much greater influence on robustness.

Higher quality HDMI receiver chips on the market incorporate an equalizer which can tolerate and successfully recover a signal that has been heavily attenuated.

For 1080p content, which is twice the bandwidth of 720p/1080i, the signal integrity is significantly more sensitive and cable lengths over 4-5m will begin to show differentiation in performance.
----------------------



Best Regards,

Rodolfo La Maestra