A Few Technical Errors Related To "Digital Performance Testing"
Dear Gary:
I’ve always been a longtime loyal reader of your magazine. The content is informative, technical, and quite often cutting-edge.
However, my colleges and I at AudioQuest feel that the article titled “Digital Performance Testing” by Jeffrey Boccaccio in your December 2007 issue contains more than a few technical errors that need to be addressed.
Although there is some useful information in the article, Mr. Boccaccio portrays his company as “a few brave men” who are the sole entity with a clear understanding of HDMI technology. Consider these quotes from the article: “The days of selling snake oil and pretty packaging were over. It was time to play with the big boys!” and “The specs are laid out, anyone can read them. There is no call for this or for the countless other ways firms push the limit and spoil a fantastic system… What a joke!” Perhaps it was unintentional, but Mr. Boccaccio appears to be saying that the aftermarket sellers of high-performance have no in-house expertise with the technical qualifications and knowledge to properly determine (through testing) the quality of the HDMI cables that they make and sell. This is certainly not true of AudioQuest, and while I can’t speak for all of our competitors, I believe it is also not true for a number of them.
I feel duty bound to point out that Mr. Boccaccio’s article contains more than a few technical errors. Some of these errors must be considered serious since they pertain to the very specifications that the DPL program is testing.
On page 82, Mr. Boccaccio explains how the DPL rating is given: “The ratings are qualified by each of the following categories: • Eye Pattern • Inter Pair Skew • DDC Line Performance • 5 Volt Lines Capability • Hot Plug Detect Return Dynamics.” Yet Mr. Boccaccio is wrong about two of the five specifications DPL is testing!
Let’s take a look at these examples one by one: On page 78, Mr. Boccaccio says: “Hot Plug Detect… The voltage that returns to the source originates from the original 5 volts sent down from the source. This return voltage is limited to no less than 3.3 volts.” This is clearly incorrect. Table 4-26 on page 52 of the HDMI 1.3a spec says: “Required Detection Levels for Hot Plug Detect Signal…Item Value…High voltage level (Source) Minimum 2.0 Volts, Maximum 5.3 Volts.” This discrepancy would directly affect the accuracy of the DPL test.
On page 78 of the article, Mr. Boccaccio says this about the +5 V line in the HDMI interface: “The tolerance of the voltage is ±10 percent (0.5 volts). It must remain stable from 4.5 to 5.5 volt. We explained that without this 5 volts (±10 percent) the unit may have problems initializing the system.” Again, this is incorrect. The Table 4-22 from the HDMI 1.3a spec states “+5V Power Pin Voltage. Item Min Max. TP1 voltage 4.8 Volts 5.3 Volts.” Again, being one of the five specifications that DPL measures, this discrepancy would directly affect the accuracy of the DPL test.
Mr. Boccaccio’s article contains other errors that may not directly affect the DPL test results, but they do raise some doubt about InVision’s true suitability to act as a kind of governing body for the testing of HDMI technology.
On page 77 of the magazine, Figure 2, the power line is shown as a two-way path between the source and sink devices. This is incorrect, the +5 V power is clearly a one-way (source to sink) feature. More errors are contained in the section following, relating to power. Mr. Boccaccio states “Fuel: The HDMI interface requires energy to operate. It is an active-system interface unlike the analog systems from the past. It uses a DC voltage of 5 volts, which is mandatory for the interface to operate and to stay operating.” This is not true. The +5 V power in the HDMI interface is only designed to power the DDC communication portion of the circuit in the sink device, but only when the sink device is turned off. When the sink is turned on, it should not draw much power from the source anymore, as is clearly stated on page 50 of the HDMI 1.3a spec; “A Sink shall not draw more than 50 mA of current from the +5V Power pin. When the Sink is powered on, it can draw no more than 10mA of current from the +5 V Power signal.”
In addition, there are indications that Mr. Boccaccio may have a less-than-complete understanding of how the HPD (hot plug detection) works. On page 78, Mr. Boccaccio states “Hot Plug Detect: This wire is used to inform the source that the sink received a request for operation and is ready for further discussions.” HPD is a sink-initiated action; when the cable is plugged into the sink, the sink HPD pin inside the cable carries the voltage to tell the source to start “handshaking.”
On page 77 of the article, Mr. Boccaccio states: “For HDTV, there can be about two million pixels with 256 levels of brightness per pixel.” The minimum bit depth of an HDTV format is 8-bit per color or 24-bit total for RGB colors. Therefore, HDTV has 16.7 million possible colors, or 16.7 million brightness levels (each color has a unique brightness level). This is not esoteric information...it’s common knowledge.
Continuing on page 77 of the article Mr. Boccaccio says: “1080p would be over a trillion instructions per second. The only transport system that could successfully move this level of data was digital.” Again, this is not accurate. Any video signal format could be transmitted in either digital or analog form. As a matter of fact, the 1080p signal is generated from most cameras in the analog format before being converted to digital. The drive voltage of each pixel in most TVs is analog form. The reason that there is no 1080p capability on component video is not the limitation of the analog signal; rather it’s a copyright protection imposed by Hollywood.
These are a few of the examples of inaccurate information contained in Mr. Boccaccio’s article. When Mr. Boccaccio approached AudioQuest at the CEDIA show last year, I gave him the following reasons why AudioQuest had decided not to join the program:
1) Most of the leading companies making HDMI cables already have all of the necessary test equipment (eye pattern tester, Bit-Error Rate “BER” tester, etc.) and the expertise in-house to perform the same tests DPL would do. Why would we pay a fee to have DPL do the test?
2) The cable product alone cannot define one of the major HDMI specifications, the DDC load character. The entire system—source, cable, sink, etc. must be considered. The tests that DPL proposes cannot guarantee the cable would work even with very high scores.
3) DPL only tests a set quantity of samples for each model, and it can only sample a very small quantity of cables throughout the year. At AudioQuest, we do 100 percent BER tests with every long cable we sell.
AudioQuest has no problem at all with the idea of a testing body to verify that HDMI standards are met. We welcome this, since it would benefit all. But we believe it is important that this testing body, in addition to being neutral, be 100 percent technically competent.
Xiaozheng Lu, Senior Vice President, Product Development, AudioQuest
Editor-In-Chief and Publisher Gary Reber Comments:
Xiaozheng, thank you for your probing letter pertaining to DPL testing. I asked Jeffrey Boccaccio if he would like to respond to our letter. Below is his response.
Jeffrey Boccaccio, President, In-Visions Technology International Comments: I would like to thank Xiaozheng Lu of AudioQuest for his response to my article in Issue 126, December 2007 of Widescreen Review on HDMI and the need for an industry-wide performance testing and certification program. I especially appreciate his comment that “AudioQuest has no problem at all with the idea of a testing body to verify that HDMI standards are met. We welcome this since it would benefit all.” On this point, we all agree.
Mr. Lu’s letter also clearly demonstrates that, although the DPL program is only just now launching, it is already having a substantial impact on the industry, and that also is good news for us all.
Before I respond to some of the specific points from Mr. Lu’s letter, it should be stated up front that the article provided to Widescreen Review was lengthy, detailed, complicated, and—by necessity—produced quickly due to short production schedules. As a result of this process, we simply did not have the amount of time that we would normally have liked for a full proofing process. Thanks to an unlimited time for review and Mr. Lu’s sharp eyes, clearly some inadvertent misstatements made it through to publication.
That being said, I would like to assure Mr. Lu and all Widescreen Review readers that DPL testing engineers are all quite well versed in the HDMI 1.3 specification, and all testing will be conducted in strict accordance with this standard. However, the real power of the DPL program is not that we know the HDMI 1.3 spec better than AudioQuest does, but that we have developed a testing regime that is universally applied to all submitted products and brands. In other words, we offer a testing and performance ranking service that provides a basis of comparison among all brands—to the benefit of the consumer.
We at In-Visions are taking significant “heat” from some cable companies that—like a shady used-car salesman—ask consumers to simply “trust me” with sometimes outrageous performance claims coming not from their laboratories, but from their marketing departments. The DPL testing program, however, will eliminate FUD—Fear, Uncertainty, and Doubt—from the consumer’s buying process. Through an independent, fair, and accurate testing program applied to all brands, the consumer will be presented with a performance ranking that will help ensure that they buy the right product, from the right brand, at the right level of performance, for the right price.
In-Visions Technology has invested literally hundreds of thousands of dollars in the latest state-of-the-art Tektronix test equipment and is well prepared to put submitted cable products through rigorous and fair testing. In fact, we are now in the process of conducting testing on products submitted by several participating manufacturers.
I am pleased to hear that AudioQuest claims to possess an equivalent grade of testing equipment and thoroughly tests (100 percent testing, according to Lu) their products. However, we have good reason to believe that not all companies are similarly equipped or follow precisely the same testing procedures. With this being the case, consumers are left to their own recourse to try and compare and contrast competing products, sometimes receiving unsatisfactory results. It is this unhappy reality that In-Visions Technology seeks to address with the DPL performance testing and certification program.
In closing, I would like to address Mr. Lu’s three summation points:
1. Although Mr. Lu states that he cannot speak for other companies, in his first point Mr. Lu states that “Most of the leading
companies making HDMI cables already have all of the necessary test equipment…and the expertise…Why would we pay a fee to have DPL do the test?”
There is no way Mr. Lu could know what type of testing equipment and regimes all of the other brands have, and in fact, we have good reason to believe that not ALL companies are so equipped.
Nevertheless, a company like AudioQuest would pay a fee to have its products tested and certified if it wanted to give its customers the benefit of being able to objectively compare AudioQuest products with other brands on the market. If you do not desire to so assist your customers, you need not participate. No law mandates DPL certification. Companies freely choose DPL’s independent, objective performance testing and ranking for many reasons, including: demonstrating the superiority of their products relative to others on the market; helping their customers clearly identify the quality steps within their line; and, finally, adhering to a universal testing standard to aid their customers in comparative shopping.
2. Mr. Lu states, “The entire system; source, cable, sink, etc. must be considered.”
Not a bad point. In fact, DPL is planning on expanding testing to other elements in the system. However, until that time, it still makes sense to offer a performance-ranking system that will aid consumers in comparing existing cable brands and products today.
3. Mr. Lu states that “DPL only tests a set quantity of samples for each model, and it can only sample a very small quantity of cables throughout the year. At AudioQuest, we do 100 percent BER test with every long cable we sell.”
DPL will test multiple samples of each manufacturer’s products. In fact, DPL will not only test multiple samples provided by the company, but our testing regime includes multiple product samples purchased at retailers around the country via a “secret shopper,” to make sure unscrupulous manufacturers cannot get a favorable ranking by supplying “hand-picked and tweaked” samples. This will allow us to ensure that the products actually being delivered by the manufacturer are meeting promised performance standards.
I want to thank Widescreen Review for the opportunity to comment on Mr. Lu’s letter. It is interesting to note that DPL performance testing is a service that will primarily aid consumers in their quest to purchase products that perform as promised. And yet, DPL even helps companies like AudioQuest to objectively demonstrate the superior performance of their products by displaying a DPL performance-ranking seal placing them above other companies who either don’t support DPL testing or whose products rank lower than AudioQuest’s.
You can E-mail Widescreen Review @ editorgary@widescreenreview.com