Moving to a 48G envelope is massive and the folks at HDMI have done it and kudos to them. So how do we protect from the inevitable false marketing claims that are sure to surface even before the specification is finalized?
How can anyone legitimately claim their HDMI transmission lines (cables, fiber, wireless, etc.) are ready for the HDMI Rev 2.1 48G? Well they really can’t since the actual spec has not been finalized; however, there are basic rules and fundamentals that can allow for some pretty accurate scientific predictions.
If we look back at the tests used for our old friend audio, one can claim all kinds of specifications depending on how the product was measured. Remember the days where power output was all over the map because some rated in IHF (Institute of High Fidelity) and others in RMS (Root Mean Square). Then there were those that decided to pick particular frequencies along the entire bandwidth (20Hz to 20kHz) that best represented their products.
So, for customers that did not quite understand these numbers, many were swayed into believing things that really were not true.
For instance, you can measure power amplifiers with an IHF rating at a mid-band frequency of 1kHz and come up with some incredible output numbers. Whereas amps far more expensive and higher in quality were given an RMS rating taking into consideration the entire audio spectrum and not focusing in on a sweet spot.
Then you have the products that really stretch the bandwidth specs by not mentioning the rolloff points. Again, use audio. If a manufacturer built an amp that produced 100 watts per channel out to 20kHz, how would that company determine the “standard” for this achievement? That’s pretty simple, it has always been at the -3dB rolloff.
So, if your amplifier rolled off, say at 25khz at -3dB and was still producing 100 watts per channel, can you not claim that your amp is in fact good to 25KHz?
Sure, you can! But there are still tricks like only measuring one channel at a time. Now the power supply is only powering one channel with far more current stability. You have to think smart and understand how these specs are derived.
Let’s look at HDMI Rev 2.1 bandwidth, but instead of using 20kHz, let’s use the predicted channel bandwidth of 12Gbps. Have you seen anyone describe their HDMI products with a deviation rolloff number?
NO! Why? There are limits for this in bandwidth. The figure above shows this. You can see that the HDMI response curve (Insertion Loss) for Rev 1.4 can be as much as 25dB down @ 5.1Gbps and still be within limits.
This is due to some corrective equalization inside each display (or HDMI receiver). Think of it as HDMI tone control with a response curve shaped opposite the loss you see here. Now extend this curve out to what we expect to be 12Gbps per channel and run the numbers — wow!
We know that if we get any throughput in comparison to today’s HDMI Rev 2.0 response rules we can pretty much prove that a particular transmission line can run at 12Gbps. These products will be limited by large gauge sizes and short lengths.
What we don’t know is if and what kind of corrective EQ HDMI will provide compared to what we have used in the past. Who knows, we may find there is tons of correction, and even longer lengths and smaller gauges can meet the standards.
We can produce some very close approximations by using conservative guidelines that provide enough data to predict 12Gbps per channel capability. A good example is today’s HDMI limits. We can compare the new 12Gbps data integrity to today’s 6Gbps rules and compute that particular transmission line’s performance in this worst-case testing environment. We can run these numbers well beyond 12Gbps per channel … all the way to 20Gbps. We will be running these numbers on existing Rev 2.0 cables to see how they fare as we approach this huge 48G transmission.