We’ve all had a good seven years to figure out why our interconnected devices refused to work properly with the HDMI 2.1 specification. The HDMI Forum announced at CES today that it’s time to start considering new headaches. HDMI 2.2 will require new cables for full compatibility, but it has the same physical connectors. Tiny QR codes are suggested to help with that, however.
The new specification is named HDMI 2.2, but compatible cables will carry an “Ultra96” marker to indicate that they can carry 96GBps, double the 48 of HDMI 2.1b. The Forum anticipates this will result in higher resolutions and refresh rates and a “next-gen HDMI Fixed Rate Link.” The Forum cited “AR/VR/MR, spatial reality, and light field displays” as benefiting from increased bandwidth, along with medical imaging and machine vision.
↫ Kevin Purdey at Ars Technica
I’m sure this will not pose any problems whatsoever, and that no shady no-name manufacturers will abuse this situation at all. DisplayPort is the better standard and connector anyway.
No, I will not be taking questions.
HDMI, as long as it continues to stay a closed expensive licensable spec, needs to die.
DP doesn’t have this problem.
Out of sheer curiosity, what was the HDMI Forum supposed to do? You cannot have higher data rates without a more expensive cable, either we’re talking about a cable that has more conductors in it (the USB 3.0 way) or a cable that has higher-quality conductors in it (the HDMI way, since the HDMI connector is pretty packed with pins already). The best the HDMi Forum could do was to trademark a name such as “Ultra96” and enforce it as best they can. Which is what they did.
DisplayPort is irrelevant for TVs because it has no ARC/eARC, which is the only way to connect your device to a Dolby Atmos receiver (unless your device has a second dedicated HDMI port just for audio, which only a handful of Blu-ray players have).
For the nitpickers: Yes ARC can work with lossless Dolby Atmos if your source has the ability to transcode it to lossy Dolby Digital Atmos (aka E-AC3+JOC), you don’t necessarily need eARC.
A problem to solve instead of investing time and resources in a closed format.
What is the “problem to solve”?
– If the “problem to solve” is the lack of ARC/eARC support in DisplayPort, then the VESA dullards would have added it years ago (back in 2009 when HDMI got ARC), they just don’t care about such “consumer concerns”. This makes HDMI a must-have for setups with a Dolby Atmos receivers.
– If the “problem to solve” is ARC/eARC itself, it’s unfortunately a requirement due to HDCP. You see, the HDCP people charge a royalty per port, and ARC/eARC allows source devices to have only one HDMI output (that outputs both video and Atmos audio) and then let the TV split the Atmos audio from the video and pass the Atmos audio to the receiver via ARC/eARC. Atmos will never happen over SPIDF or even USB-C because those don’t do HDCP, and Hollywood requires that Atmos only travel over an HDCP link. Even GPUs don’t have a second dedicated HDMI port for audio, precisely because they want to save on royalties, only a handful of expensive Blu-ray players have it.
Really, I don’t see how the problem can be solved, regardless of the definition of the problem.
HDCP is horse shit, that makes everything expensive an incompatible – without ANY benefit, to anyone, including the content owners. Let’s not pretend that has any merit or value.
I don’t think kurkosdr was defending HDCP, just pointing out that we all have to deal with it. I’m sure they would agree that it has no merit or value. It certainly doesn’t for us mere consumers!
I agree with Thom that DisplayPort is overall the better standard, and if not for media companies’ desire for control over anything media related, it would be the default connector on consumer devices just as it is on PCs and other IT equipment.
“Out of sheer curiosity, what was the HDMI Forum supposed to do? You cannot have higher data rates without a more expensive cable, either we’re talking about a cable that has more conductors in it (the USB 3.0 way) or a cable that has higher-quality conductors in it (the HDMI way, since the HDMI connector is pretty packed with pins already). ”
No, that’s not true obviously. The other option, would be to develop a better encoding for the data. Think about how upgrading from a 14.4d to a 28.8k modem surprisingly didn’t require you to rewire your phone cord. Of course, if it were easy they would have done it. Changing the cable format is a worse choice for most of the participants. So I assume that they didn’t have any easy protocol improvement they could throw at it. Or the tin foil hats are right and the accessory minded participants won the argument. But I don’t believe in tin foil hats.
All good practical encodings for short-to-medium cables have already been invented. Display port and USB-C have the same problem that HDMI has (you need a better cable for the higher data rates).
The difference is that HDMI has trademarked certifications for cables that can do the higher data rate while DisplayPort and USB-C are a wild west.
Bill Shooter of Bul,
The range of a cable is extremely dependent on signal frequency. You can go higher if the bandwidth isn’t already close to being maxed out, but most modern data cables are already being pushed near the max frequencies they’re engineered for. Look at the chart for “Values of primary parameters for telephone cable”…
https://en.wikipedia.org/wiki/Telegrapher's_equations
Notice that at low frequencies, doubling has very little impact on resistance. This is why modem speeds could be increased without upgrading wires. But as frequencies keep going up the resistance gets exponentially more pronounced. If the run is short enough, then there isn’t much signal degradation and it can still overpower the noise. But as the cable gets longer the signal drops and the noise increases. This is where higher quality cables are needed.
You can go beyond the cable’s specs, but you’ll start to get errors. For example some of us have tried 10gbe on cat 5e cables. It actually works ok for short runs in environments with low electrical noise. But it doesn’t take much to start getting data errors. Even these errors might be tolerable depending on what higher level layers do. Computer networks are quite tolerant of this and if your not explicitly testing for lost packets you may not even notice. As far as I know HDMI does not do retransmission, errors will turn into audio/video artifacts. If you find someone who’s tested this stuff specifically, it could be an interesting read.
Personally I haven’t tested HDMI cables, but I have tested very long USB cables – around 30m, both passive and amplified. In my testing I found that active cables solve the signal quality problems, but it appears that some hardware/drivers have critical timing requirements. If you try to go too far (around 10m with two repeaters IIRC) data packets can start to drop. For example hard drives would work absolutely fine the full length, but my logitech web cams wouldn’t work reliably. As far as I can tell this was not electrical signal quality but the hardware/driver’s inability to deal with latency.
kurkosdr,
This makes it even more confusing for USB-C / Thunderbolt connector users. Even though my monitor has support, and my computer, too, they would not negotiate properly.
Add in docking stations to the mix, and it gets even worse. Which ports work? Which don’t? Do we need adapters? And then they are sometimes hidden. The type-C cable can carry HDMI, but most cables are internally type-C -> DP and then DP-> HDMI which “erases” capabilities.
Wish they could come up with a better “Lowest common denominator” between these three standards.
HDMI is more than 20 years old, in most regions there is nothing they can enforce patent related and the connector is unchanged, so I suppose a new standard is the way to make sure the fees keep rolling in. Even if the new claims are as bogus as an ISP that can suddenly pump 2GB down the same fibres that could only delivery 200MB a few years ago!
I suspect it’s just another case of “Our gold cables are better, if you listen carefully you can hear it!”
Up until fairly recently (2017), new HDMI versions were very welcome, as they introduced things like eARC and the ability to do 4K HDR at up to 144Hz framerate. This is especially true for TVs, since TV manufacturers don’t want to waste an entire port for a DisplayPort connector that can’t do Atmos passthrough, so HDMI is your only choice when it comes to TVs (so, progress on the HDMi front is very welcome). This is objectively measurable performance btw, not fake claims like audiofool cables.
So, HDMI is set until at least until 2037 as far as patents are concerned.
Now whether HDMI 2.2 gets adopted, that’s dependent on whether 8K or 4K HDR stereoscopic gaming find a market.
kurkosdr,
One advantage of higher generation cabling is being able to “over provision” to avoid signal issues.
For connecting a 4K monitor at 60Hz, HDMI 2.0 might be sufficient. However I would step up to 8K capable cables that are overkill for the purpose, but will give me significant headroom for the signal. And in the future if I ever upgrade, then the cable will already be there.
(With I could just use type-C, though, we are so close to having a single standard cable for everything).
It’s also worth mentioning that even if the HDMI patents have expired for early versions of HDMI, HDCP still remains royalty-encumbered because of the DMCA (in countries where the DMCA or similar laws apply), so you can only implement cleartext HDMI without paying royalties (either in a source or a receiver). Also, you can’t use the term “HDMI” since it’s trademarked, so you have to either leave the port unlabelled or label it something generic like “Digital AV out”.
Cleartext HDMI is practical on a source device if your device isn’t meant to officially implement any DRM, that’s how some FPGA projects have HDMI output: they only output cleartext HDMI (and they also leave the port unlabelled or have it generically-labelled). Cleartext HDMI on receiver devices (for example TVs and monitors) isn’t practical obv.