19 ms·

When HDMI 2.1 isn't HDMI 2.1 – the confusing world of the standard

232 points by cbg0 5y ago

mey 5y ago
Quick summary, HDMI 2.0 no longer "exists", and HDMI 2.1 only features are now optional according to the certifying body. Manufactures are supposed to indicate which features they support. Whelp, I guess we should just stick to Display Port 1.4
- ChrisRR 5y ago
  Same is done with the bluetooth standard. Manufacturers proudly claim their product uses bluetooth 5.0, reviewers assume that means higher speed and lower power, but the products don't actually have to support all sections of the standard and can pretty much just be bluetooth classic with 5.0 slapped on the label
- jayflux 5y ago
  I see they went to the USB school of standardisation
  - Roritharr 5y ago
    I really wonder what is up with that. These standards become increasingly frustratingly complex even for people who deal with them daily.
    - colechristensen 5y ago
      It’s just design by committee and long-standing efforts for backwards compatibility. Also the people writing the standards are far too familiar with them and thus a bit lost when it comes to making practical decisions. Whenever you make changes there will be compromises and someone will have reason to be unhappy.
      - mannykannot 5y ago
        There is no way that eliminating 2.0, in such a way that everything that used to rate as 2.0 now qualifies as conforming to some subset of 2.1, can be justified as backwards compatibility.
    - jjoonathan 5y ago
      Cable manufacturers probably realized that the secret to profits lay in resisting commoditization, beat a path to the table, and made it happen.
    - zokier 5y ago
      I don't think its particularly odd that the specifications are supersets of old versions; indeed that feels pretty common in the standards world. IETF specs are maybe the odd ones out where you typically have to read like ten different RFCs to get good picture of some standard.
    - ClumsyPilot 5y ago
      Free market of fraud is going on, there should be a cost to lying
  - mavhc 5y ago
    USB 2 Full Speed = USB 1 speed This has been going on for 20 years
    - _abox 5y ago
      That was already called full speed in the USB 1 days though. There's a mode called low speed too which was 1.5mbit and meant for low throughput devices. Full speed was 12Mbit. This way stuff like keyboards and mice didn't need to include the highest end controller chips. Calling it Full Speed was more a lack of vision than a deliberate attempt to confuse consumers as this HDMI thing seems to me. So then when 2 and 3 came out they were forced to find superlatives. Hi-Speed, Superspeed. The same happened with radio. High Frequency was up to 30Mhz. Then they found they could go higher. Very High Frequency, VHF. Then came more advances and Ultra-High Frequency. Eventually they gave up after XHF and SHF and started using band letters :) Using relative terms for constantly changing technologies is just a bad idea :)
      - dylan604 5y ago
        The same people are the ones that decide to put _final/_latest/_etc in file names
    - tgsovlerkhgsel 5y ago
      The newest version of USB seems to have solved this by labeling ports with the supported speed (5, 10 or 20). Assuming they start doing the same with the cables, at least that part of the problem becomes tractable.
      - brnt 5y ago
        There is a reserved seat in hell for whoever allowed usb-c cables to have a max speed of usb2 (480Mbit/s).
        mananaysiempre 5y ago
        In theory, the choice of USB-C or -{,mini,micro}A/B is orthogonal to the choice of version: there are USB-A and even USB-B connectors (with additional pins) that support USB 3.0 as well. There is a logic to this. In practice, it is confusing.
  - errcorrectcode 5y ago
    LOL. Why don't you understand superduperspeed USB 3.4.6 2x2x6? It's so eeeasy. And, coming soon to an Amazon app near you: (Pack of 3) USB4 Thunderbolt 4, 98 ft / 30m, 240W charging, gold plated! $19.99! Free Shipping! It's not like anyone is checking products are what they say they are.
    - spicybright 5y ago
      You can't legally call your food a hamburger if it's 90% meat glue and 10% cow. Can we not do the same with cables? Whoever is responsible for setting standards must be getting some good kickbacks from all this...
      - colejohnson66 5y ago
        The people writing the standards are also the ones implementing it. That’s the kickback. That’s why every USB 3.whatever device suddenly became USB4 ones.
- IshKebab 5y ago
  Seems kind of reasonable. DisplayPort is exactly the same - just because your display supported DisplayPort 1.4 doesn't mean it is required to support 10 bit colour, VRR, etc.
  - thaumasiotes 5y ago
    I replaced my old Dell XPS laptop with a newer Dell laptop. The old one had been happily driving an external monitor through a thunderbolt cable (USB-C on the laptop side; DisplayPort on the monitor side.) The new laptop still has a thunderbolt port proudly advertised in the specs. But I'm not allowed to use it. It won't send video data out that way. And when I called tech support, the most they would offer me for my newly-purchased laptop was "Don't use the thunderbolt port. Use the HDMI port."
    - Osiris 5y ago
      I thought that the primary feature of Thunderbolt was DisplayPort signaling. What makes it Thunderbolt vs USB-C if it doesn't have DP support?
      - eropple 5y ago
        So Thunderbolt 2 included a DisplayPort signaling layer but (AFAIK; may be wrong) Thunderbolt 3 does not -- instead, the hardware must implement the USB-C Alternate Mode for DisplayPort (and in turn is also implementing the USB-C Alternate Mode for Thunderbolt; it's not USB, at least not in USB3). Thunderbolt's main draw is being "PCIe on a wire".
        thaumasiotes 5y ago
        My port is advertised as "Thunderbolt 3/USB 3.2 Gen 1 Type-C port with DisplayPort with alt mode". I guess the problem is that it can't speak whatever version the cable understands? The cable says "thunderbolt 3 compatible". And like I said, the same cable worked fine with the same monitor and a previous Dell laptop. :-/
theandrewbailey 5y ago
It sounds like the marketing people who kept renaming USB 3.x Gen FU got hired to mess up HDMI.
kup0 5y ago
Why does it feel like it is inevitable that standardization/licensing organizations in tech will always eventually turn into a user-hostile mess? USB, HDMI, what can we screw up next? Is it incompetence? Malice? I'd really like to see an in-depth investigation of this phenomenon
- ASalazarMX 5y ago
  HDMI was never a pro-user protocol, it was made to encumber a digital display signal with DRM.
  - ChuckNorris89 5y ago
    This. HDMI was cooked as a proprietary connector with DRM by the big manufacturers in the DVD/Blu-Ray, TV, home-entertainment business and the big movie studios to enforce stronger content protections to protect their IP, at wich it miserably failed, as I can still torrent every Hollywood blockbuster and every Netflix series. IIRC, every manufacturer must pay a fee to the HDMI consortium for every device with HDMI they sell. DisplayPort, by contrast is a more open standard only requiring a flat fee for the standard documentation and membership instead of a fee per unit sold IIRC.
    - babypuncher 5y ago
      DisplayPort and DVI both support HDCP. This wasn't the purpose behind HDMI, though support for it was no doubt a requirement for adoption. It was designed to be a single cable for carrying video and audio between playback devices, receivers, and displays. For this purpose, it succeeded and did a much better job at it than alternatives. HDMI still makes far more sense for use in a home theater environment than DisplayPort thanks to features like ARC.
      - MomoXenosaga 5y ago
        On that note are there TVs with displayport? I'm using my LG TV as monitor for a PC and forced to use HDMI.
astraloverflow 5y ago
As the saying goes, a camel is a horse designed by a committee
smithza 5y ago
LTT did some manual testing of HDMI cables [0] in hopes of answering the last question of this article, "how do consumers know if a cable supports v2.1 features?" Does anyone know of other tests or more comprehensive data sets? [0] https://linustechtips.com/topic/1387053-i-spent-a-thousand-dollars-on-hdmi-cables-for-science/ https://linustechtips.com/topic/1387053-i-spent-a-thousand-d...
floatingatoll 5y ago
This is a stellar example of how catering to everyone results in the destruction of a brand. “HDMI 2.1” will be with us for years and it’s essentially irrelevant now, and they aren’t willing to admit they were wrong, so their only hope is to release an HDMI 2.2 that is just “all of the optional features of HDMI 2.1 are now required”, which will cause howls of outrage and confusion. I’m guessing they are too captive to manufacturers to have the courage to do that. Oh well.
- IshKebab 5y ago
  It wasn't an oversight to make the features optional. They're deliberately optional so device manufacturers aren't forced into a ridiculous all or nothing situation.
  - nomel 5y ago
    I see it as a problem with the abstractions being on the wrong layer.
    - IshKebab 5y ago
      It's kind of going in that direction with HDMI and DisplayPort over USB 3. But.. not exactly because they're not actually encoding video data into USB packets. It's too difficult to have abstractions like that when you're dealing with such insanely high data rates.
  - rocqua 5y ago
    AND manufacturers not wanting to be stuck with old branding on devices. the standard coukd have made some things optional. But they made everything optional.
  - floatingatoll 5y ago
    I think they would have been better off forcing manufacturers into that situation, and that the feature list has grown so large that it’s no longer a sensible specification for purchasing decisions, which will erode consumer trust.
PedroBatista 5y ago
I'm a "tech guy" and I'll be in the market soon for a screen. Just the thought I'll have to learn about this while HDMI disaster in order to not get burned gives me anxiety. Also, never quite liked HDMI when it came out, but from what I'm reading they really outdone themselves during these years.
- zamadatix 5y ago
  If you're getting a high performance monitor DP is overwhelmingly preferred and available anyways, e.g. this monitor can be operated at full speed via DP 1.2 (though it requires reduced timings at least you don't have to drop chroma quality). If you're getting a high performance TV then be prepared to dig into which input port set to which mode supports which actual data rate and features on which firmware because literally none of it is consistent. Also be sure to check technical reviews to see that it actually works, e.g. 4k120 HDR with freesync from my Xbox Series X didn't work with my TV for quite a while even though all of those things worked individually.
- bjoli 5y ago
  I have been avoiding HDMI like the plague due to various issues with devices and cables. The only way it has reliably worked has been where the use has been WAY under what devices and cables have been specced for. Anything remotely close to max speeds/features has been riddled with problems. DisplayPort has been less of a hassle, but probably mostly because I settled for 4k60 which HDMI 2.0 supposedly handles with ease. Devices with DP2.0 will appear any time now. On paper DP2.0 shound handle 10k60 displays without compression in 8 bit, or 8k60 (10bit) also without compression.
anonymousiam 5y ago
I remember when USB2 came out and similar mischief ensued. All the hardware manufacturers got together and pushed the standards body to re-brand USB 1.1 hardware as USB 2.0 (full-speed vs. high-speed). It allowed hardware retailers to empty their shelves, while consumers thought they were getting the latest technology. https://arstechnica.com/uncategorized/2003/10/2927-2/ https://arstechnica.com/uncategorized/2003/10/2927-2/
- sixothree 5y ago
  Same thing exists for USB3. Every time a new version is released, all cables and products suddenly support that revision. They just don't have any new features. Not to mention that I've _never_ had a cable identify what it is capable of. Thus USB is a shitshow of crapiness.
  - thescriptkiddie 5y ago
    I recently upgraded the NVME SSD in my machine. The motherboard only has a single NVME compatable M.2 port, so I bought a USB 3.1 enclosure [0] to put the old drive in while I cloned it to the new drive. The enclosure has a USB type-C connector so I also had to use a USB 3.1 A-to-C adapter [1] to connect it to my motherboard's [2] USB 3.1 type-A port. Anyway something somewhere went wrong and it took over 5 hours to copy 750 GB instead of the expected 10 minutes. Absolute shitshow. [0] https://www.newegg.com/rosewill-rhub-20001-enclosure/p/N82E16817182427?Item=N82E16817182427 https://www.newegg.com/rosewill-rhub-20001-enclosure/p/N82E1... [1] https://www.amazon.com/dp/B07L92KPBB https://www.amazon.com/dp/B07L92KPBB [2] https://www.asus.com/us/Motherboards/Z97AUSB_31/ https://www.asus.com/us/Motherboards/Z97AUSB_31/
    - zokier 5y ago
      From the review section of that enclosure: > Cons: - Included USB-C to USB-C cable is only capable of USB 2.0 speeds (40 MB/s as measured with crystaldiskmark) yeah, that would explain it.
    - anonymousiam 5y ago
jon-wood 5y ago
Has anyone ever seen a device that actually uses Ethernet over HDMI? The thought of being able to plug a single network cable into the back of your display and then anything plugged into that has a wired connection is lovely, but as far as I can tell absolutely nothing actually supports it, despite the ever growing set of internet connected devices sitting underneath people's TVs.
- tjohns 5y ago
  Ethernet Over HDMI is used by newer AV receivers to support eARC (extended audio return channel). The older ARC spec would work with any HDMI cable, but bandwidth limitations only allowed compressed 5.1 surround sound. eARC uses the higher bandwidth from Ethernet Over HDMI, allowing uncompressed 7.1 surround and Dolby Atomos streams. (If you're not familiar with ARC/eARC, this lets the TV send audio from its native inputs back to the AV receiver over an HDMI cable. Without ARC, you need to plug everything directly into the AV receiver.)
  - josteink 5y ago
    eARC is neat in theory, but my experience with it has been that’s too unreliable and unstable to actually use in practice. I even bought new cables to make sure there wouldn’t be issues, but eARC audio regularly falls out in ways other sources (including regular ARC) doesn’t. And when it fails there’s literally zero tools for diagnosing it either. Maybe around the time of eARC2 we’ll have something working as well as Bluetooth does today. (Yes, that’s me being snarky)
    - ApolIllo 5y ago
      That's unfortunate. I've been hoping to simplify my HT setup and eARC was something I wanted to target in an upgrade
    - isatty 5y ago
      I’ve had no problems with eARC when used together with my Appletv and LG C1 TV - which I’m happy about because the appletv does not offer any other wired audio ports.
MomoXenosaga 5y ago
Is it really confusing? The people who need the specs of HDMI2.1 (like gamers) will do their research.
Strom 5y ago
As with any tech, you can't trust marketing if you plan on pushing it to the limit. You need to either test it yourself, or in some rare cases there are reviewers who have already tested it. Most "reviews" for tech are extremely superficial though and certainly won't be testing HDMI inner workings. For HDMI 2.1, there are a bunch of monitors being sold under that banner that don't have the full 48 Gbps bandwidth. For example the Gigabyte M28U is limited to half of that at 24 Gbps. [1] Gigabyte certainly doesn't want you to know this. On their specificaion page they just list it as HDMI 2.1. [2] Similar nonsense was going on during the transition from HDMI 1.4 to 2.0. I really wanted HDMI 2.0 output on my laptop and held off on a purchase until that was possible. I finally bought an ASUS Zephyrus GX501 laptop. It has a Nvidia GTX 1080, which does support HDMI 2.0 output. The marketing for this laptop also seems to suggest that they're utillizing this potential, with claims like "You can also run a large ROG gaming monitor with NVIDIA G-SYNC™ via DisplayPort™ over Type-C™ USB or use HDMI 2.0 to connect 4K TVs at 60Hz." [3] The specification page mentions the HDMI 2.0 port. [4] However in reality I found out that this HDMI 2.0 port is limited to HDMI 1.4 bandwidth. It supports HDMI 2.0 features like HDR, but not the bandwidth. 4K @ 60Hz is possible only with 4:2:0 chroma subsampling and you're limited to 8 bits, so no HDR. I'm not the only one who found this out either. There are plenty of others on the ASUS forums. [5] Hard to say whether this was intentional misleading by ASUS marketing, or whether engineering messed up, or whether the feature ended up being cut due to time constraints. In any case, they still haven't adjusted the old marketing pages for this laptop that never shipped with HDMI 2.0 bandwidth. Reviewers don't tend to check things like this either. For example The Verge reviwed this laptop [6] and wrote: "Asus has a nice array of ports on the Zephyrus — four traditional USB 3.0 ports, a Thunderbolt 3 USB Type-C port, HDMI, and a headphone jack." They're just regurgitating the marketing material. There's no depth to it, the claims aren't verified. So people end up buying the product and then get confused why it isn't working. -- [1] https://www.rtings.com/monitor/discussions/q-D1CBeE2EiGMYgn/additional-review-notes-supported-resolutions https://www.rtings.com/monitor/discussions/q-D1CBeE2EiGMYgn/... [2] https://www.gigabyte.com/Monitor/M28U/sp#sp https://www.gigabyte.com/Monitor/M28U/sp#sp [3] https://rog.asus.com/laptops/rog-zephyrus/rog-zephyrus-gx501-series/ https://rog.asus.com/laptops/rog-zephyrus/rog-zephyrus-gx501... [4] https://rog.asus.com/laptops/rog-zephyrus/rog-zephyrus-gx501-series/spec https://rog.asus.com/laptops/rog-zephyrus/rog-zephyrus-gx501... [5] https://rog.asus.com/forum/showthread.php?96916-GX501-Zephyrus-unable-to-output-4k60hz-HDMI https://rog.asus.com/forum/showthread.php?96916-GX501-Zephyr... [6] https://www.theverge.com/2017/8/25/16201656/asus-rog-zephyrus-review-gaming-laptop-nvidia-max-q https://www.theverge.com/2017/8/25/16201656/asus-rog-zephyru...
alin23 5y ago
When the new MacBook Pro came out this year, everyone was puzzled as to why the newly included HDMI port was only 2.0. Well it turns out they lied. It’s 2.1 after all! \s Jokes aside, it’s actually only 2.0 because internally they transmit a DisplayPort video signal and convert it to HDMI using an MCDP2900 chip[0], which is the same chip usually seen inside USB-C hubs. So the new MacBook basically got rid of the HDMI dongle by integrating it inside the laptop. This also breaks DDC/CI on that port and now I get a ton of support emails for Lunar (https://lunar.fyi https://lunar.fyi) because people can’t control their monitor brightness/volume and think that the app is broken. [0] https://www.kinet-ic.com/mcdp2900/ https://www.kinet-ic.com/mcdp2900/
- eatYourFood 5y ago
  How can I control screen brightness from my m1 max? Willing to pay for a solution.
  - alin23 5y ago
    Lunar can do that: https://lunar.fyi https://lunar.fyi It’s free for manual brightness adjustments. Just make sure to use one of the Thunderbolt ports of the MacBook. And if it still doesn’t work, check if there’s any monitor setting that could block DDC by going through this FAQ: https://lunar.fyi/faq#brightness-not-changing https://lunar.fyi/faq#brightness-not-changing
    - MrSourz 5y ago
      Neat. I've been quite fond of Monitor Control: https://github.com/MonitorControl/MonitorControl https://github.com/MonitorControl/MonitorControl
      - ApolIllo 5y ago
        How does Lunar and Monitor Control get around the Tbolt -> HDMI dongle baked into the logic board problem?
wyager 5y ago
2021 HDMI is a disaster. I'm using a Sony flagship TV, a PC, and a popular Sony 7.1 receiver. I had to update my graphics card to get 4k120 444 10bit and eARC. Only eARC is totally broken - audio often doesn't work at all without restarting PC/TV/receiver a few times. And then once it "works" it will randomly cut out for 5-10 seconds at a time. HDR on windows is also totally broken. It was a nightmare to get something that correctly rendered full 10bit HDR video (I ended up having to use MPC-HC with madvr and a ton of tweaking). You also have to turn off windows HDR support and use D3D exclusive mode. After updating my TV to get DRR, the audio for this setup stopped working. Linux also has zero HDR support. Didn't have luck getting 5.1 or 7.1 working either. MacOS can at least handle HDR on Apple displays - not sure if it works on normal displays. Haven't tried surround sound.
- plus 5y ago
  Were you able to get 4k120Hz 444 working on Linux? What GPU do you have? I can only do 4k60 444 or 4k120 420 on my LG C1 connected to my Radeon RX 6900xt.
  - wyager 5y ago
    I don't remember, but I had to sidegrade from a 2070Ti to a 3060Ti to get HDMI2.1 to get 4k120 444.
- MayeulC 5y ago
  Not to be offensive, but -- first world problems: where did you find a new graphics carts, for starters? Now, a bit more on a serious tone: this is all bleeding edge. And combining multiple recent development together is a recipe for corner cases and untested combinations. That said, did you try Variable Refresh Rate with that? Bur reduction technologies (backlight strobing) are also interesting, but thankfully they require little software interaction (for now).
- zamadatix 5y ago
  > HDR on windows is also totally broken. It was a nightmare to get something that correctly rendered full 10bit HDR video (I ended up having to use MPC-HC with madvr and a ton of tweaking). You also have to turn off windows HDR support and use D3D exclusive mode. After updating my TV to get DRR, the audio for this setup stopped working. Odd, I've got 3 different HDR monitors across 2 Windows computers (one Nvidia, one AMD) and also 2 different HDR TVs occasionally connected to each and I've never had to turn HDR off or do tweaking in madvr to get it to look right. Having HDR disabled in Windows globally makes me wonder if the adjustments you're talking about are referring to tonemapping HDR content to an exclusive SDR output. One thing I will say is TVs default to some god awful HDR settings. Well I guess you could say that about picture settings on TVs in general but it's even worse for HDR. It took me a solid 40 minutes to figure out how to get the picture settings for a Samsung Q900R to display HDR properly instead of "worlds darkest picture mode" (turns out the same values take different actual effect depending on the type of source you identify it as and "PC" is not what you want even though you specify HDR picture mode elsewhere and it detects an HDR signal...). Also for SDR content you'll need to adjust the SDR bridghtness slider depending on the actual brightness of your display.
theshrike79 5y ago
Linus Tech Tips used a $expensive dedicated device to test a ton of HDMI cables. Most of them were shit: https://youtu.be/XFbJD6RE4EY https://youtu.be/XFbJD6RE4EY And what was most interesting is that price and quality didn't always correlate at all.
errcorrectcode 5y ago
The fundamental problem is a lack of supply chain integrity. Customers can buy a million cables or laptop batteries directly from (country that shall not be named), but they have no idea if they're getting fakes or not. The fix isn't "authorized" suppliers only, but requiring a reputable someone in the supply chain to maintain evidence of continually testing products advertising trademarked standards for compliance. If it's too much work, then boohoo, sad day for them, they don't get to be traded or sold in country X. In all honesty, flooding a market with cheap, substandard products claiming standards they don't comply with is dumping. https://en.wikipedia.org/wiki/Dumping_(pricing_policy) https://en.wikipedia.org/wiki/Dumping_(pricing_policy)
- jml7c5 5y ago
  >In all honesty, flooding a market with cheap, substandard products claiming standards they don't comply with is dumping. I think you may have misread the article. The problem isn't that manufacturers are lying about HDMI 2.1 compliance. They are complying with the HDMI 2.1 requirements; the standards body chose a weird set of requirements, that's all.
b3lvedere 5y ago
When i had a very old Samsung tv, my Nvidia Geforce videocard produced a nice image to the tv and Dolby AC3 sound to my even older surround set via a nice hdmi to optical out converter in between. Now i have a not-so-old Philips tv and suddenly i can't get dolby ac3 sound anymore. Why? Because the GeForce communicates with the tv and the tv responds it only has stereo. The surround set has no hdmi input or output so it cannot communicate with GeForce. I have tried everything from hacking drivers to changing EDID with some weird devices. Nothing works. Stereo only. Very frustrating. I was recommended to replace my surround set or my tv. Both pretty expensive solutions for some weird hdmi communication bug/standard. So i bought a $20 usb sound device to get Dolby AC3 sound to my suround set. All because i replaced my old tv which couldn't communicate with the GeForce about its speaker setup.
nvarsj 5y ago
This article doesn't mention the most infuriating aspect of HDMI - it's not an open standard! It's a closed, proprietary standard that requires licensing fees and prevents any open source drivers from existing. This is why the Linux AMD open source drivers don't support HDMI 2.1 - so you can't display 120hz@4K on Linux with AMD.
- alvarlagerlof 5y ago
  Can displayport do it?
  - the_pwner224 5y ago
    Of course. Using 4k / 144 / 10bit right now on a RX 5700xt.
  - nvarsj 5y ago
    You can’t do 4:4:4 chroma with displayport 1.4. Not enough bandwidth. As a result, fonts look pretty bad. Next gen displayport should be able to do it but no cards support it yet afaik.
pseudosavant 5y ago
This HDMI bandwidth calculator helped me understand HDMI 1.4/2.0/2.1 far better than anything else. It is worth noting that many resolution/depth/rate configurations easily fit within TDMS limitations, especially if DSC is enabled (which it should be). FRL isn't required, unless you want to be certain you can support all HDMI 2.1 situations (mostly over 8K or 240Hz+). https://www.murideo.com/cody.html https://www.murideo.com/cody.html
- tverbeure 5y ago
  If you want to compare video bandwidth requirements not only for HDMI but also for DisplayPort and some other video transports, you can also use my video timings calculator: https://tomverbeure.github.io/video_timings_calculator https://tomverbeure.github.io/video_timings_calculator.
  - fomine3 5y ago
    Looks useful, thank you. Could you add DisplayPort with DSC?
    - tverbeure 5y ago
      The DSC compression ratio is selected by the source as a function of the available and desired bandwidth. One thing I could add is the theoretical maximum bandwidth if DSC is dialed up to the maximum compression setting of just 8 bits per pixel.
tehabe 5y ago
Maybe an analogue VGA cable connection wasn't the worst idea
Night_Thastus 5y ago
Wait, according to that chart...does that mean that HDMI 2.0 cannot support Dolby Vision, as it is a form of Dynamic HDR? I am very confused, as several devices I own are only supporting 2.0, but claim to support Dolby Vision.
shmerl 5y ago
I'm still waiting for DisplayPort to replace HDMI, but some profit a lot on the patents for it, that's why you don't see DP in TVs and such.
gumby 5y ago
Shouldn’t consumer AV be adopting USB C by now?
- shmerl 5y ago
  They should (USB-C with DisplayPort protocol). But HDMI group makes a lot of money on patents, that's why DP adoption is so slow in that segment. https://en.wikipedia.org/wiki/HDMI#History https://en.wikipedia.org/wiki/HDMI#History It's simply anti-progress stance at this point.
01100011 5y ago
HDMI has always been a shitshow. I spent years of my life playing whack-a-mole for a major set top box manufacturer because of incomplete or braindead implementations. CEC added a whole 'nother layer of shit to go wrong. There is a reason CE mfgs go to "plug fests"(https://www.cta.tech/Events/CTA-861-PlugFest https://www.cta.tech/Events/CTA-861-PlugFest) instead of being able to trust their implementation will work with other devices as long as they follow the rules.
- TheJoeMan 5y ago
  Whoa I had never heard of a plug fest before, the concept seems pretty interesting. Silly of me to assume the mega corps just buy their competitor’s products internally.
  - myself248 5y ago
    Buying each other's products only gives you what's actually made it to market. Plug-fests are full of unreleased stuff. (Buying each other's stuff, to tear it down for competitive analysis, is exceedingly common. But you don't do that at a plugfest.) Trivia: Plug-fests used to be called Bake-offs, but Pillsbury sued to stop it: https://www.cs.columbia.edu/sip/sipit/pillsbury.html https://www.cs.columbia.edu/sip/sipit/pillsbury.html Trivia 2: Bluetooth calls theirs an "unplugfest". https://www.bluetooth.com/bluetooth-resources/unplugfest-event-guide/ https://www.bluetooth.com/bluetooth-resources/unplugfest-eve...
- tacoman 5y ago
  I'm still living this dream today. It's amazing how terrible brand spankin' new "hospitality" (hotel) TVs are when it comes to poor implementations. It is still whack-a-mole and often has to be compensated for on the set-top side because the TV manufacturers are even worse at firmware.
Nevermark 5y ago
My design would remove all ambiguity, by having different rules for describing port vs. device capabilities. 1. HDMI ports should be described as "HDMI X.XX Compatible". This indicates that the ports themselves (and the device that contains them) will work when connected to any other HDMI X.XX device, and that the description is of a port. This is the low bar on 2.1 devices. Their ports will sensibly negotiate optional features with other HDMI 2.1 devices. 2. HDMI devices should be described as "Full HDMI X.XX" or "Limited HDMI X.XX up to [limitations]" to distinguish devices that support all features of X.XX that apply to the device, or have applicable limitations. (Audio devices would not be considered "Limited" by non-applicable features such as image resolution.) The ""Limited HDMI X.XX up to ..." disclosure would need to be prominently displayed in any description. One place where all limitations can be found. Limitation phrases (like "up to 4k resolution") would have standard wording supplied by the HDMI standards body. Additional references to HDMI versions, such as product listing titles, can be shortened to "Limited HDMI X.XX" without any limitations listed. So all limitations listed, or none, to avoid any confusion. Done! Next problem, Internet? ...
- andrecarini 5y ago
  My design would be HDMI 3.Y where Y is a variable-length Base32 encoding of the value of a bitmask representing the presence of the underlying features of an HDMI port/cable/device.
  - rowanG077 5y ago
    This is terrible. You expect an average consumer to start converting base32 to binary to check what features their cable support? I mean I'm all for educating the public but this is just completely unreasonable
    - hayanno 5y ago
      I'm pretty sure it's a joke, maybe your comment is also one but it's harder to tell.
    - andrecarini 5y ago
porphyra 5y ago
I bought an HDMI 2.1 8K display, only to find that the Nvidia RTX 3090 wouldn't run it at 8k 60 Hz on Linux, only 30 Hz. On Windows, however, 8k 60 Hz works perfectly. It was a bummer.
dwaite 5y ago
As someone who works on a lot of standards, I can say there are two common misconceptions 1. Typically, standards and certification are entirely different. 2. Standards may not follow semantic versioning - e.g. a major version may not indicate a loss of either backward or forward compatibility. The protocols defined in USB 1 are still allowed in USB4, as are the cables and connectors. HDMI is the same way. Saying something is USB4 or HDMI 2.2 compliant is not a stronger statement than saying they are USB 1.0 or HDMI 1.0 compliant. Likewise, statements like "USB 3.2 Gen 2x2" are garbage. The correct terminology is "SuperSpeed USB 20Gbps". Why do so many products use the incorrect, more confusing terminology while omitting the official marketing name? Often because they are not certified products.
- Raqbit 5y ago
  Was "USB 3.2 Gen 2x2" not the name of SuperSpeed 20Gbps before USB 4 got released?
  - zokier 5y ago
    No, "SuperSpeed 20 Gbps" has been the official nomenclature since day 1. From USB 3.2 language use guidelines • USB 3.2 Gen 1 o Product capability: product signals at 5Gbps o Marketing name: SuperSpeed USB • USB 3.2 Gen 2 o Product capability: product signals at 10Gbps o Marketing name: SuperSpeed USB 10Gbps • USB 3.2 Gen 2x2 o Product capability: product signals at 20Gbps o Marketing name: SuperSpeed USB 20Gbps [...] To avoid consumer confusion, USB-IF’s recommended nomenclature for consumers is “SuperSpeed USB” for 5Gbps products, “SuperSpeed USB 10Gbps” for 10Gbps products and “SuperSpeed USB 20Gbps” for 20Gbps products. https://www.usb.org/sites/default/files/usb_3_2_language_product_and_packaging_guidelines_final.pdf https://www.usb.org/sites/default/files/usb_3_2_language_pro... The SuperSpeed name was originally introduced with USB 3.0
ksec 5y ago
The first thing, this isn't the first time Xiaomi has done something like this. Earlier this year they launched a WiFi 6E router, every single press ran their PR without fact checking and looking through the data. The router doesn't even support 6Ghz along with a few other optional features. After hours of searching, reading through spec apparently no one on the internet gives a damn, I wrote to Wi-Fi Alliance for clarification, and Xiaomi finally retracted their PR and relabel their router as WiFI 6 only. The second baffling thing is how HN's react to USB-C and HDMI so differently. With USB-C / USB 4, more than half of HN expect the consumers to know which cables and port offer which features. It is the consumers's fault for not choosing the right cable or knowing what their port does. I dont see how that is different to HDMI here. All it takes is one bad actor in the market. And it will be a race to the bottom. That is partly why Apple brought back MagSafe. You dont tell the user to go and find a USB-C 120W ( or 240W ) capable cable that may or may not be compliant to do quick charging. You tell them to use MagSafe cable.
- smcl 5y ago
  Weird, I haven’t seen HNers blaming the user for USB confusion. In fact I am sure I’ve seen articles here getting shared and upvoted which discuss the USB confusion and express sympathy for users while trying to disambiguate the situation
  - ksec 5y ago
    Yes you get these article submission, but you still see many of HN comments defending USB-C as a non-issue or an user issue. It is the user's fault for buying the wrong cable. Just get the Thunderbolt Cable. For most part of 2014 to 2019. USB-C was hailed as the silver bullet. The increasing backlash against USB-C was a very very recent thing.
    - ximeng 5y ago
      https://news.ycombinator.com/item?id=29445844 https://news.ycombinator.com/item?id=29445844 this was the last discussion I saw on HN about this, very much not hiding the complexity or laying blame.
taurath 5y ago
Topic aside - why can’t cable companies write the spec of the cable on the cable? Like it’s all the same connector, I can’t tell the difference between 2.0 and 2.1. seems like it should be standard
- atoav 5y ago
  It is HF. A propper tester for that stuff costs at least 10k and this would test one cable at a time. Throwing cables on the market and hope they work is much easier. Getting cheap HDMI cables is a gamble for that reason and I wish it would be different.
taubek 5y ago
This reminds me so much to the USB standard story. I don't know why is it so hard to have some kind of normal naming convention?
dataengineer56 5y ago
I blame Sony. They pushed expensive HDMI 2.1-compatible TVs and receivers alongside the PS5, strongly implying that without a 2.1-compatible living room then you might as well be playing your PS5 through a black-and-white TV. Of course the PS5 can't really exploit any of the features of HDMI 2.1, and then it turned out that most/all 2.1-compatible receivers and TVs have glaring errors and incompatibilities that render them effectively useless.
- Mindwipe 5y ago
  LG pushed HDMI 2.1 on their TVs a full year before Sony did.
clarge1120 5y ago
HDMI is contributing to consumer confusion at least, and fraud at most. In order to get the benefits of this protocol, you need the source, the monitor, and the cables to meet the spec. That’s a lot of opportunities for confusion, waste, and loss of consumer confidence. The protocol specs may be great, but the confusion will eventually lead consumers to think that HDMI 2.1 is just industry jargon, like “hi-def”.

232 points by cbg0 5y ago

mey 5y ago

Quick summary, HDMI 2.0 no longer "exists", and HDMI 2.1 only features are now optional according to the certifying body. Manufactures are supposed to indicate which features they support. Whelp, I guess we should just stick to Display Port 1.4

ChrisRR 5y ago
Same is done with the bluetooth standard. Manufacturers proudly claim their product uses bluetooth 5.0, reviewers assume that means higher speed and lower power, but the products don't actually have to support all sections of the standard and can pretty much just be bluetooth classic with 5.0 slapped on the label
jayflux 5y ago
I see they went to the USB school of standardisation
- Roritharr 5y ago
  I really wonder what is up with that. These standards become increasingly frustratingly complex even for people who deal with them daily.
  - colechristensen 5y ago
    It’s just design by committee and long-standing efforts for backwards compatibility. Also the people writing the standards are far too familiar with them and thus a bit lost when it comes to making practical decisions. Whenever you make changes there will be compromises and someone will have reason to be unhappy.
    - mannykannot 5y ago
      There is no way that eliminating 2.0, in such a way that everything that used to rate as 2.0 now qualifies as conforming to some subset of 2.1, can be justified as backwards compatibility.
  - jjoonathan 5y ago
    Cable manufacturers probably realized that the secret to profits lay in resisting commoditization, beat a path to the table, and made it happen.
  - zokier 5y ago
    I don't think its particularly odd that the specifications are supersets of old versions; indeed that feels pretty common in the standards world. IETF specs are maybe the odd ones out where you typically have to read like ten different RFCs to get good picture of some standard.
  - ClumsyPilot 5y ago
    Free market of fraud is going on, there should be a cost to lying
- mavhc 5y ago
  USB 2 Full Speed = USB 1 speed This has been going on for 20 years
  - _abox 5y ago
    That was already called full speed in the USB 1 days though. There's a mode called low speed too which was 1.5mbit and meant for low throughput devices. Full speed was 12Mbit. This way stuff like keyboards and mice didn't need to include the highest end controller chips. Calling it Full Speed was more a lack of vision than a deliberate attempt to confuse consumers as this HDMI thing seems to me. So then when 2 and 3 came out they were forced to find superlatives. Hi-Speed, Superspeed. The same happened with radio. High Frequency was up to 30Mhz. Then they found they could go higher. Very High Frequency, VHF. Then came more advances and Ultra-High Frequency. Eventually they gave up after XHF and SHF and started using band letters :) Using relative terms for constantly changing technologies is just a bad idea :)
    - dylan604 5y ago
      The same people are the ones that decide to put _final/_latest/_etc in file names
  - tgsovlerkhgsel 5y ago
    The newest version of USB seems to have solved this by labeling ports with the supported speed (5, 10 or 20). Assuming they start doing the same with the cables, at least that part of the problem becomes tractable.
    - brnt 5y ago
      There is a reserved seat in hell for whoever allowed usb-c cables to have a max speed of usb2 (480Mbit/s).
      - mananaysiempre 5y ago
        In theory, the choice of USB-C or -{,mini,micro}A/B is orthogonal to the choice of version: there are USB-A and even USB-B connectors (with additional pins) that support USB 3.0 as well. There is a logic to this. In practice, it is confusing.
- errcorrectcode 5y ago
  LOL. Why don't you understand superduperspeed USB 3.4.6 2x2x6? It's so eeeasy. And, coming soon to an Amazon app near you: (Pack of 3) USB4 Thunderbolt 4, 98 ft / 30m, 240W charging, gold plated! $19.99! Free Shipping! It's not like anyone is checking products are what they say they are.
  - spicybright 5y ago
    You can't legally call your food a hamburger if it's 90% meat glue and 10% cow. Can we not do the same with cables? Whoever is responsible for setting standards must be getting some good kickbacks from all this...
    - colejohnson66 5y ago
      The people writing the standards are also the ones implementing it. That’s the kickback. That’s why every USB 3.whatever device suddenly became USB4 ones.
IshKebab 5y ago
Seems kind of reasonable. DisplayPort is exactly the same - just because your display supported DisplayPort 1.4 doesn't mean it is required to support 10 bit colour, VRR, etc.
- thaumasiotes 5y ago
  I replaced my old Dell XPS laptop with a newer Dell laptop. The old one had been happily driving an external monitor through a thunderbolt cable (USB-C on the laptop side; DisplayPort on the monitor side.) The new laptop still has a thunderbolt port proudly advertised in the specs. But I'm not allowed to use it. It won't send video data out that way. And when I called tech support, the most they would offer me for my newly-purchased laptop was "Don't use the thunderbolt port. Use the HDMI port."
  - Osiris 5y ago
    I thought that the primary feature of Thunderbolt was DisplayPort signaling. What makes it Thunderbolt vs USB-C if it doesn't have DP support?
    - eropple 5y ago
      So Thunderbolt 2 included a DisplayPort signaling layer but (AFAIK; may be wrong) Thunderbolt 3 does not -- instead, the hardware must implement the USB-C Alternate Mode for DisplayPort (and in turn is also implementing the USB-C Alternate Mode for Thunderbolt; it's not USB, at least not in USB3). Thunderbolt's main draw is being "PCIe on a wire".
      - thaumasiotes 5y ago
        My port is advertised as "Thunderbolt 3/USB 3.2 Gen 1 Type-C port with DisplayPort with alt mode". I guess the problem is that it can't speak whatever version the cable understands? The cable says "thunderbolt 3 compatible". And like I said, the same cable worked fine with the same monitor and a previous Dell laptop. :-/

theandrewbailey 5y ago

It sounds like the marketing people who kept renaming USB 3.x Gen FU got hired to mess up HDMI.

kup0 5y ago

Why does it feel like it is inevitable that standardization/licensing organizations in tech will always eventually turn into a user-hostile mess? USB, HDMI, what can we screw up next? Is it incompetence? Malice? I'd really like to see an in-depth investigation of this phenomenon

ASalazarMX 5y ago
HDMI was never a pro-user protocol, it was made to encumber a digital display signal with DRM.
- ChuckNorris89 5y ago
  This. HDMI was cooked as a proprietary connector with DRM by the big manufacturers in the DVD/Blu-Ray, TV, home-entertainment business and the big movie studios to enforce stronger content protections to protect their IP, at wich it miserably failed, as I can still torrent every Hollywood blockbuster and every Netflix series. IIRC, every manufacturer must pay a fee to the HDMI consortium for every device with HDMI they sell. DisplayPort, by contrast is a more open standard only requiring a flat fee for the standard documentation and membership instead of a fee per unit sold IIRC.
  - babypuncher 5y ago
    DisplayPort and DVI both support HDCP. This wasn't the purpose behind HDMI, though support for it was no doubt a requirement for adoption. It was designed to be a single cable for carrying video and audio between playback devices, receivers, and displays. For this purpose, it succeeded and did a much better job at it than alternatives. HDMI still makes far more sense for use in a home theater environment than DisplayPort thanks to features like ARC.
    - MomoXenosaga 5y ago
      On that note are there TVs with displayport? I'm using my LG TV as monitor for a PC and forced to use HDMI.

astraloverflow 5y ago

As the saying goes, a camel is a horse designed by a committee

smithza 5y ago

LTT did some manual testing of HDMI cables [0] in hopes of answering the last question of this article, "how do consumers know if a cable supports v2.1 features?" Does anyone know of other tests or more comprehensive data sets? [0] https://linustechtips.com/topic/1387053-i-spent-a-thousand-dollars-on-hdmi-cables-for-science/ https://linustechtips.com/topic/1387053-i-spent-a-thousand-d...

floatingatoll 5y ago

This is a stellar example of how catering to everyone results in the destruction of a brand. “HDMI 2.1” will be with us for years and it’s essentially irrelevant now, and they aren’t willing to admit they were wrong, so their only hope is to release an HDMI 2.2 that is just “all of the optional features of HDMI 2.1 are now required”, which will cause howls of outrage and confusion. I’m guessing they are too captive to manufacturers to have the courage to do that. Oh well.

IshKebab 5y ago
It wasn't an oversight to make the features optional. They're deliberately optional so device manufacturers aren't forced into a ridiculous all or nothing situation.
- nomel 5y ago
  I see it as a problem with the abstractions being on the wrong layer.
  - IshKebab 5y ago
    It's kind of going in that direction with HDMI and DisplayPort over USB 3. But.. not exactly because they're not actually encoding video data into USB packets. It's too difficult to have abstractions like that when you're dealing with such insanely high data rates.
- rocqua 5y ago
  AND manufacturers not wanting to be stuck with old branding on devices. the standard coukd have made some things optional. But they made everything optional.
- floatingatoll 5y ago
  I think they would have been better off forcing manufacturers into that situation, and that the feature list has grown so large that it’s no longer a sensible specification for purchasing decisions, which will erode consumer trust.

PedroBatista 5y ago

I'm a "tech guy" and I'll be in the market soon for a screen. Just the thought I'll have to learn about this while HDMI disaster in order to not get burned gives me anxiety. Also, never quite liked HDMI when it came out, but from what I'm reading they really outdone themselves during these years.

zamadatix 5y ago
If you're getting a high performance monitor DP is overwhelmingly preferred and available anyways, e.g. this monitor can be operated at full speed via DP 1.2 (though it requires reduced timings at least you don't have to drop chroma quality). If you're getting a high performance TV then be prepared to dig into which input port set to which mode supports which actual data rate and features on which firmware because literally none of it is consistent. Also be sure to check technical reviews to see that it actually works, e.g. 4k120 HDR with freesync from my Xbox Series X didn't work with my TV for quite a while even though all of those things worked individually.
bjoli 5y ago
I have been avoiding HDMI like the plague due to various issues with devices and cables. The only way it has reliably worked has been where the use has been WAY under what devices and cables have been specced for. Anything remotely close to max speeds/features has been riddled with problems. DisplayPort has been less of a hassle, but probably mostly because I settled for 4k60 which HDMI 2.0 supposedly handles with ease. Devices with DP2.0 will appear any time now. On paper DP2.0 shound handle 10k60 displays without compression in 8 bit, or 8k60 (10bit) also without compression.

anonymousiam 5y ago

I remember when USB2 came out and similar mischief ensued. All the hardware manufacturers got together and pushed the standards body to re-brand USB 1.1 hardware as USB 2.0 (full-speed vs. high-speed). It allowed hardware retailers to empty their shelves, while consumers thought they were getting the latest technology. https://arstechnica.com/uncategorized/2003/10/2927-2/ https://arstechnica.com/uncategorized/2003/10/2927-2/

sixothree 5y ago
Same thing exists for USB3. Every time a new version is released, all cables and products suddenly support that revision. They just don't have any new features. Not to mention that I've _never_ had a cable identify what it is capable of. Thus USB is a shitshow of crapiness.
- thescriptkiddie 5y ago
  I recently upgraded the NVME SSD in my machine. The motherboard only has a single NVME compatable M.2 port, so I bought a USB 3.1 enclosure [0] to put the old drive in while I cloned it to the new drive. The enclosure has a USB type-C connector so I also had to use a USB 3.1 A-to-C adapter [1] to connect it to my motherboard's [2] USB 3.1 type-A port. Anyway something somewhere went wrong and it took over 5 hours to copy 750 GB instead of the expected 10 minutes. Absolute shitshow. [0] https://www.newegg.com/rosewill-rhub-20001-enclosure/p/N82E16817182427?Item=N82E16817182427 https://www.newegg.com/rosewill-rhub-20001-enclosure/p/N82E1... [1] https://www.amazon.com/dp/B07L92KPBB https://www.amazon.com/dp/B07L92KPBB [2] https://www.asus.com/us/Motherboards/Z97AUSB_31/ https://www.asus.com/us/Motherboards/Z97AUSB_31/
  - zokier 5y ago
    From the review section of that enclosure: > Cons: - Included USB-C to USB-C cable is only capable of USB 2.0 speeds (40 MB/s as measured with crystaldiskmark) yeah, that would explain it.
  - anonymousiam 5y ago

jon-wood 5y ago

Has anyone ever seen a device that actually uses Ethernet over HDMI? The thought of being able to plug a single network cable into the back of your display and then anything plugged into that has a wired connection is lovely, but as far as I can tell absolutely nothing actually supports it, despite the ever growing set of internet connected devices sitting underneath people's TVs.

tjohns 5y ago
Ethernet Over HDMI is used by newer AV receivers to support eARC (extended audio return channel). The older ARC spec would work with any HDMI cable, but bandwidth limitations only allowed compressed 5.1 surround sound. eARC uses the higher bandwidth from Ethernet Over HDMI, allowing uncompressed 7.1 surround and Dolby Atomos streams. (If you're not familiar with ARC/eARC, this lets the TV send audio from its native inputs back to the AV receiver over an HDMI cable. Without ARC, you need to plug everything directly into the AV receiver.)
- josteink 5y ago
  eARC is neat in theory, but my experience with it has been that’s too unreliable and unstable to actually use in practice. I even bought new cables to make sure there wouldn’t be issues, but eARC audio regularly falls out in ways other sources (including regular ARC) doesn’t. And when it fails there’s literally zero tools for diagnosing it either. Maybe around the time of eARC2 we’ll have something working as well as Bluetooth does today. (Yes, that’s me being snarky)
  - ApolIllo 5y ago
    That's unfortunate. I've been hoping to simplify my HT setup and eARC was something I wanted to target in an upgrade
  - isatty 5y ago
    I’ve had no problems with eARC when used together with my Appletv and LG C1 TV - which I’m happy about because the appletv does not offer any other wired audio ports.

MomoXenosaga 5y ago

Is it really confusing? The people who need the specs of HDMI2.1 (like gamers) will do their research.

Strom 5y ago

As with any tech, you can't trust marketing if you plan on pushing it to the limit. You need to either test it yourself, or in some rare cases there are reviewers who have already tested it. Most "reviews" for tech are extremely superficial though and certainly won't be testing HDMI inner workings. For HDMI 2.1, there are a bunch of monitors being sold under that banner that don't have the full 48 Gbps bandwidth. For example the Gigabyte M28U is limited to half of that at 24 Gbps. [1] Gigabyte certainly doesn't want you to know this. On their specificaion page they just list it as HDMI 2.1. [2] Similar nonsense was going on during the transition from HDMI 1.4 to 2.0. I really wanted HDMI 2.0 output on my laptop and held off on a purchase until that was possible. I finally bought an ASUS Zephyrus GX501 laptop. It has a Nvidia GTX 1080, which does support HDMI 2.0 output. The marketing for this laptop also seems to suggest that they're utillizing this potential, with claims like "You can also run a large ROG gaming monitor with NVIDIA G-SYNC™ via DisplayPort™ over Type-C™ USB or use HDMI 2.0 to connect 4K TVs at 60Hz." [3] The specification page mentions the HDMI 2.0 port. [4] However in reality I found out that this HDMI 2.0 port is limited to HDMI 1.4 bandwidth. It supports HDMI 2.0 features like HDR, but not the bandwidth. 4K @ 60Hz is possible only with 4:2:0 chroma subsampling and you're limited to 8 bits, so no HDR. I'm not the only one who found this out either. There are plenty of others on the ASUS forums. [5] Hard to say whether this was intentional misleading by ASUS marketing, or whether engineering messed up, or whether the feature ended up being cut due to time constraints. In any case, they still haven't adjusted the old marketing pages for this laptop that never shipped with HDMI 2.0 bandwidth. Reviewers don't tend to check things like this either. For example The Verge reviwed this laptop [6] and wrote: "Asus has a nice array of ports on the Zephyrus — four traditional USB 3.0 ports, a Thunderbolt 3 USB Type-C port, HDMI, and a headphone jack." They're just regurgitating the marketing material. There's no depth to it, the claims aren't verified. So people end up buying the product and then get confused why it isn't working. -- [1] https://www.rtings.com/monitor/discussions/q-D1CBeE2EiGMYgn/additional-review-notes-supported-resolutions https://www.rtings.com/monitor/discussions/q-D1CBeE2EiGMYgn/... [2] https://www.gigabyte.com/Monitor/M28U/sp#sp https://www.gigabyte.com/Monitor/M28U/sp#sp [3] https://rog.asus.com/laptops/rog-zephyrus/rog-zephyrus-gx501-series/ https://rog.asus.com/laptops/rog-zephyrus/rog-zephyrus-gx501... [4] https://rog.asus.com/laptops/rog-zephyrus/rog-zephyrus-gx501-series/spec https://rog.asus.com/laptops/rog-zephyrus/rog-zephyrus-gx501... [5] https://rog.asus.com/forum/showthread.php?96916-GX501-Zephyrus-unable-to-output-4k60hz-HDMI https://rog.asus.com/forum/showthread.php?96916-GX501-Zephyr... [6] https://www.theverge.com/2017/8/25/16201656/asus-rog-zephyrus-review-gaming-laptop-nvidia-max-q https://www.theverge.com/2017/8/25/16201656/asus-rog-zephyru...

alin23 5y ago

When the new MacBook Pro came out this year, everyone was puzzled as to why the newly included HDMI port was only 2.0. Well it turns out they lied. It’s 2.1 after all! \s Jokes aside, it’s actually only 2.0 because internally they transmit a DisplayPort video signal and convert it to HDMI using an MCDP2900 chip[0], which is the same chip usually seen inside USB-C hubs. So the new MacBook basically got rid of the HDMI dongle by integrating it inside the laptop. This also breaks DDC/CI on that port and now I get a ton of support emails for Lunar (https://lunar.fyi https://lunar.fyi) because people can’t control their monitor brightness/volume and think that the app is broken. [0] https://www.kinet-ic.com/mcdp2900/ https://www.kinet-ic.com/mcdp2900/

eatYourFood 5y ago
How can I control screen brightness from my m1 max? Willing to pay for a solution.
- alin23 5y ago
  Lunar can do that: https://lunar.fyi https://lunar.fyi It’s free for manual brightness adjustments. Just make sure to use one of the Thunderbolt ports of the MacBook. And if it still doesn’t work, check if there’s any monitor setting that could block DDC by going through this FAQ: https://lunar.fyi/faq#brightness-not-changing https://lunar.fyi/faq#brightness-not-changing
  - MrSourz 5y ago
    Neat. I've been quite fond of Monitor Control: https://github.com/MonitorControl/MonitorControl https://github.com/MonitorControl/MonitorControl
    - ApolIllo 5y ago
      How does Lunar and Monitor Control get around the Tbolt -> HDMI dongle baked into the logic board problem?

wyager 5y ago

2021 HDMI is a disaster. I'm using a Sony flagship TV, a PC, and a popular Sony 7.1 receiver. I had to update my graphics card to get 4k120 444 10bit and eARC. Only eARC is totally broken - audio often doesn't work at all without restarting PC/TV/receiver a few times. And then once it "works" it will randomly cut out for 5-10 seconds at a time. HDR on windows is also totally broken. It was a nightmare to get something that correctly rendered full 10bit HDR video (I ended up having to use MPC-HC with madvr and a ton of tweaking). You also have to turn off windows HDR support and use D3D exclusive mode. After updating my TV to get DRR, the audio for this setup stopped working. Linux also has zero HDR support. Didn't have luck getting 5.1 or 7.1 working either. MacOS can at least handle HDR on Apple displays - not sure if it works on normal displays. Haven't tried surround sound.

plus 5y ago
Were you able to get 4k120Hz 444 working on Linux? What GPU do you have? I can only do 4k60 444 or 4k120 420 on my LG C1 connected to my Radeon RX 6900xt.
- wyager 5y ago
  I don't remember, but I had to sidegrade from a 2070Ti to a 3060Ti to get HDMI2.1 to get 4k120 444.
MayeulC 5y ago
Not to be offensive, but -- first world problems: where did you find a new graphics carts, for starters? Now, a bit more on a serious tone: this is all bleeding edge. And combining multiple recent development together is a recipe for corner cases and untested combinations. That said, did you try Variable Refresh Rate with that? Bur reduction technologies (backlight strobing) are also interesting, but thankfully they require little software interaction (for now).
zamadatix 5y ago
> HDR on windows is also totally broken. It was a nightmare to get something that correctly rendered full 10bit HDR video (I ended up having to use MPC-HC with madvr and a ton of tweaking). You also have to turn off windows HDR support and use D3D exclusive mode. After updating my TV to get DRR, the audio for this setup stopped working. Odd, I've got 3 different HDR monitors across 2 Windows computers (one Nvidia, one AMD) and also 2 different HDR TVs occasionally connected to each and I've never had to turn HDR off or do tweaking in madvr to get it to look right. Having HDR disabled in Windows globally makes me wonder if the adjustments you're talking about are referring to tonemapping HDR content to an exclusive SDR output. One thing I will say is TVs default to some god awful HDR settings. Well I guess you could say that about picture settings on TVs in general but it's even worse for HDR. It took me a solid 40 minutes to figure out how to get the picture settings for a Samsung Q900R to display HDR properly instead of "worlds darkest picture mode" (turns out the same values take different actual effect depending on the type of source you identify it as and "PC" is not what you want even though you specify HDR picture mode elsewhere and it detects an HDR signal...). Also for SDR content you'll need to adjust the SDR bridghtness slider depending on the actual brightness of your display.

theshrike79 5y ago

Linus Tech Tips used a $expensive dedicated device to test a ton of HDMI cables. Most of them were shit: https://youtu.be/XFbJD6RE4EY https://youtu.be/XFbJD6RE4EY And what was most interesting is that price and quality didn't always correlate at all.

errcorrectcode 5y ago

The fundamental problem is a lack of supply chain integrity. Customers can buy a million cables or laptop batteries directly from (country that shall not be named), but they have no idea if they're getting fakes or not. The fix isn't "authorized" suppliers only, but requiring a reputable someone in the supply chain to maintain evidence of continually testing products advertising trademarked standards for compliance. If it's too much work, then boohoo, sad day for them, they don't get to be traded or sold in country X. In all honesty, flooding a market with cheap, substandard products claiming standards they don't comply with is dumping. https://en.wikipedia.org/wiki/Dumping_(pricing_policy) https://en.wikipedia.org/wiki/Dumping_(pricing_policy)

jml7c5 5y ago
>In all honesty, flooding a market with cheap, substandard products claiming standards they don't comply with is dumping. I think you may have misread the article. The problem isn't that manufacturers are lying about HDMI 2.1 compliance. They are complying with the HDMI 2.1 requirements; the standards body chose a weird set of requirements, that's all.

b3lvedere 5y ago

When i had a very old Samsung tv, my Nvidia Geforce videocard produced a nice image to the tv and Dolby AC3 sound to my even older surround set via a nice hdmi to optical out converter in between. Now i have a not-so-old Philips tv and suddenly i can't get dolby ac3 sound anymore. Why? Because the GeForce communicates with the tv and the tv responds it only has stereo. The surround set has no hdmi input or output so it cannot communicate with GeForce. I have tried everything from hacking drivers to changing EDID with some weird devices. Nothing works. Stereo only. Very frustrating. I was recommended to replace my surround set or my tv. Both pretty expensive solutions for some weird hdmi communication bug/standard. So i bought a $20 usb sound device to get Dolby AC3 sound to my suround set. All because i replaced my old tv which couldn't communicate with the GeForce about its speaker setup.

nvarsj 5y ago

This article doesn't mention the most infuriating aspect of HDMI - it's not an open standard! It's a closed, proprietary standard that requires licensing fees and prevents any open source drivers from existing. This is why the Linux AMD open source drivers don't support HDMI 2.1 - so you can't display 120hz@4K on Linux with AMD.

alvarlagerlof 5y ago
Can displayport do it?
- the_pwner224 5y ago
  Of course. Using 4k / 144 / 10bit right now on a RX 5700xt.
- nvarsj 5y ago
  You can’t do 4:4:4 chroma with displayport 1.4. Not enough bandwidth. As a result, fonts look pretty bad. Next gen displayport should be able to do it but no cards support it yet afaik.

pseudosavant 5y ago

This HDMI bandwidth calculator helped me understand HDMI 1.4/2.0/2.1 far better than anything else. It is worth noting that many resolution/depth/rate configurations easily fit within TDMS limitations, especially if DSC is enabled (which it should be). FRL isn't required, unless you want to be certain you can support all HDMI 2.1 situations (mostly over 8K or 240Hz+). https://www.murideo.com/cody.html https://www.murideo.com/cody.html

tverbeure 5y ago
If you want to compare video bandwidth requirements not only for HDMI but also for DisplayPort and some other video transports, you can also use my video timings calculator: https://tomverbeure.github.io/video_timings_calculator https://tomverbeure.github.io/video_timings_calculator.
- fomine3 5y ago
  Looks useful, thank you. Could you add DisplayPort with DSC?
  - tverbeure 5y ago
    The DSC compression ratio is selected by the source as a function of the available and desired bandwidth. One thing I could add is the theoretical maximum bandwidth if DSC is dialed up to the maximum compression setting of just 8 bits per pixel.

tehabe 5y ago

Maybe an analogue VGA cable connection wasn't the worst idea

Night_Thastus 5y ago

Wait, according to that chart...does that mean that HDMI 2.0 cannot support Dolby Vision, as it is a form of Dynamic HDR? I am very confused, as several devices I own are only supporting 2.0, but claim to support Dolby Vision.

shmerl 5y ago

I'm still waiting for DisplayPort to replace HDMI, but some profit a lot on the patents for it, that's why you don't see DP in TVs and such.

gumby 5y ago

Shouldn’t consumer AV be adopting USB C by now?

shmerl 5y ago
They should (USB-C with DisplayPort protocol). But HDMI group makes a lot of money on patents, that's why DP adoption is so slow in that segment. https://en.wikipedia.org/wiki/HDMI#History https://en.wikipedia.org/wiki/HDMI#History It's simply anti-progress stance at this point.

01100011 5y ago

HDMI has always been a shitshow. I spent years of my life playing whack-a-mole for a major set top box manufacturer because of incomplete or braindead implementations. CEC added a whole 'nother layer of shit to go wrong. There is a reason CE mfgs go to "plug fests"(https://www.cta.tech/Events/CTA-861-PlugFest https://www.cta.tech/Events/CTA-861-PlugFest) instead of being able to trust their implementation will work with other devices as long as they follow the rules.

TheJoeMan 5y ago
Whoa I had never heard of a plug fest before, the concept seems pretty interesting. Silly of me to assume the mega corps just buy their competitor’s products internally.
- myself248 5y ago
  Buying each other's products only gives you what's actually made it to market. Plug-fests are full of unreleased stuff. (Buying each other's stuff, to tear it down for competitive analysis, is exceedingly common. But you don't do that at a plugfest.) Trivia: Plug-fests used to be called Bake-offs, but Pillsbury sued to stop it: https://www.cs.columbia.edu/sip/sipit/pillsbury.html https://www.cs.columbia.edu/sip/sipit/pillsbury.html Trivia 2: Bluetooth calls theirs an "unplugfest". https://www.bluetooth.com/bluetooth-resources/unplugfest-event-guide/ https://www.bluetooth.com/bluetooth-resources/unplugfest-eve...
tacoman 5y ago
I'm still living this dream today. It's amazing how terrible brand spankin' new "hospitality" (hotel) TVs are when it comes to poor implementations. It is still whack-a-mole and often has to be compensated for on the set-top side because the TV manufacturers are even worse at firmware.

Nevermark 5y ago

My design would remove all ambiguity, by having different rules for describing port vs. device capabilities. 1. HDMI ports should be described as "HDMI X.XX Compatible". This indicates that the ports themselves (and the device that contains them) will work when connected to any other HDMI X.XX device, and that the description is of a port. This is the low bar on 2.1 devices. Their ports will sensibly negotiate optional features with other HDMI 2.1 devices. 2. HDMI devices should be described as "Full HDMI X.XX" or "Limited HDMI X.XX up to [limitations]" to distinguish devices that support all features of X.XX that apply to the device, or have applicable limitations. (Audio devices would not be considered "Limited" by non-applicable features such as image resolution.) The ""Limited HDMI X.XX up to ..." disclosure would need to be prominently displayed in any description. One place where all limitations can be found. Limitation phrases (like "up to 4k resolution") would have standard wording supplied by the HDMI standards body. Additional references to HDMI versions, such as product listing titles, can be shortened to "Limited HDMI X.XX" without any limitations listed. So all limitations listed, or none, to avoid any confusion. Done! Next problem, Internet? ...

andrecarini 5y ago
My design would be HDMI 3.Y where Y is a variable-length Base32 encoding of the value of a bitmask representing the presence of the underlying features of an HDMI port/cable/device.
- rowanG077 5y ago
  This is terrible. You expect an average consumer to start converting base32 to binary to check what features their cable support? I mean I'm all for educating the public but this is just completely unreasonable
  - hayanno 5y ago
    I'm pretty sure it's a joke, maybe your comment is also one but it's harder to tell.
  - andrecarini 5y ago

porphyra 5y ago

I bought an HDMI 2.1 8K display, only to find that the Nvidia RTX 3090 wouldn't run it at 8k 60 Hz on Linux, only 30 Hz. On Windows, however, 8k 60 Hz works perfectly. It was a bummer.

dwaite 5y ago

As someone who works on a lot of standards, I can say there are two common misconceptions 1. Typically, standards and certification are entirely different. 2. Standards may not follow semantic versioning - e.g. a major version may not indicate a loss of either backward or forward compatibility. The protocols defined in USB 1 are still allowed in USB4, as are the cables and connectors. HDMI is the same way. Saying something is USB4 or HDMI 2.2 compliant is not a stronger statement than saying they are USB 1.0 or HDMI 1.0 compliant. Likewise, statements like "USB 3.2 Gen 2x2" are garbage. The correct terminology is "SuperSpeed USB 20Gbps". Why do so many products use the incorrect, more confusing terminology while omitting the official marketing name? Often because they are not certified products.

Raqbit 5y ago
Was "USB 3.2 Gen 2x2" not the name of SuperSpeed 20Gbps before USB 4 got released?
- zokier 5y ago
  No, "SuperSpeed 20 Gbps" has been the official nomenclature since day 1. From USB 3.2 language use guidelines • USB 3.2 Gen 1 o Product capability: product signals at 5Gbps o Marketing name: SuperSpeed USB • USB 3.2 Gen 2 o Product capability: product signals at 10Gbps o Marketing name: SuperSpeed USB 10Gbps • USB 3.2 Gen 2x2 o Product capability: product signals at 20Gbps o Marketing name: SuperSpeed USB 20Gbps [...] To avoid consumer confusion, USB-IF’s recommended nomenclature for consumers is “SuperSpeed USB” for 5Gbps products, “SuperSpeed USB 10Gbps” for 10Gbps products and “SuperSpeed USB 20Gbps” for 20Gbps products. https://www.usb.org/sites/default/files/usb_3_2_language_product_and_packaging_guidelines_final.pdf https://www.usb.org/sites/default/files/usb_3_2_language_pro... The SuperSpeed name was originally introduced with USB 3.0

ksec 5y ago

The first thing, this isn't the first time Xiaomi has done something like this. Earlier this year they launched a WiFi 6E router, every single press ran their PR without fact checking and looking through the data. The router doesn't even support 6Ghz along with a few other optional features. After hours of searching, reading through spec apparently no one on the internet gives a damn, I wrote to Wi-Fi Alliance for clarification, and Xiaomi finally retracted their PR and relabel their router as WiFI 6 only. The second baffling thing is how HN's react to USB-C and HDMI so differently. With USB-C / USB 4, more than half of HN expect the consumers to know which cables and port offer which features. It is the consumers's fault for not choosing the right cable or knowing what their port does. I dont see how that is different to HDMI here. All it takes is one bad actor in the market. And it will be a race to the bottom. That is partly why Apple brought back MagSafe. You dont tell the user to go and find a USB-C 120W ( or 240W ) capable cable that may or may not be compliant to do quick charging. You tell them to use MagSafe cable.

smcl 5y ago
Weird, I haven’t seen HNers blaming the user for USB confusion. In fact I am sure I’ve seen articles here getting shared and upvoted which discuss the USB confusion and express sympathy for users while trying to disambiguate the situation
- ksec 5y ago
  Yes you get these article submission, but you still see many of HN comments defending USB-C as a non-issue or an user issue. It is the user's fault for buying the wrong cable. Just get the Thunderbolt Cable. For most part of 2014 to 2019. USB-C was hailed as the silver bullet. The increasing backlash against USB-C was a very very recent thing.
  - ximeng 5y ago
    https://news.ycombinator.com/item?id=29445844 https://news.ycombinator.com/item?id=29445844 this was the last discussion I saw on HN about this, very much not hiding the complexity or laying blame.

taurath 5y ago

Topic aside - why can’t cable companies write the spec of the cable on the cable? Like it’s all the same connector, I can’t tell the difference between 2.0 and 2.1. seems like it should be standard

atoav 5y ago
It is HF. A propper tester for that stuff costs at least 10k and this would test one cable at a time. Throwing cables on the market and hope they work is much easier. Getting cheap HDMI cables is a gamble for that reason and I wish it would be different.

taubek 5y ago

This reminds me so much to the USB standard story. I don't know why is it so hard to have some kind of normal naming convention?

dataengineer56 5y ago

I blame Sony. They pushed expensive HDMI 2.1-compatible TVs and receivers alongside the PS5, strongly implying that without a 2.1-compatible living room then you might as well be playing your PS5 through a black-and-white TV. Of course the PS5 can't really exploit any of the features of HDMI 2.1, and then it turned out that most/all 2.1-compatible receivers and TVs have glaring errors and incompatibilities that render them effectively useless.

Mindwipe 5y ago
LG pushed HDMI 2.1 on their TVs a full year before Sony did.

clarge1120 5y ago

HDMI is contributing to consumer confusion at least, and fraud at most. In order to get the benefits of this protocol, you need the source, the monitor, and the cables to meet the spec. That’s a lot of opportunities for confusion, waste, and loss of consumer confidence. The protocol specs may be great, but the confusion will eventually lead consumers to think that HDMI 2.1 is just industry jargon, like “hi-def”.