A DEFINITIVE GUIDE TO HDMI® 2.1A
- Published , by Tom Devine
As a manufacturer of video distribution products sold internationally, at AVPro Edge we are hyper-confident our affiliations with HDMI, HDCP, HDBaseT and the Imaging Science Foundation enable our engineers to design best-in-class products containing HDMI inputs, outputs and without-a-doubt-puts. Yet at times in our Sioux Falls headquarters, walking past the tech assistance offices that monitor our phone support lines, and pausing to hear integrators calling from project sites with a myriad of aural puzzles, I become re-amazed (sure, feel free to use that one...). Has our industry become so devoid of training it is taking for granted all products with HDMI ins / outs / ups and downs now simply work, regardless of their born-on date? In one sense that is true, as HDMI was designed for licensed products to produce an image when connected together.
As a major update, HDMI 2.1, was announced on January 4th, 2017, and while for some that may have passed unheralded, the industry is now on the cusp for 8K resolution and other significant enhancements to become mainstream.
Inattention to HDMI designations with legacy devices, particularly cabling, produces unanticipated compatibility issues for consumers and integrators alike, as next generation products attempt to mesh within existing systems. And adding to that, HDMI LA (HDMI Licensing Administrator, Inc.) announced prior to the opening of CES 2022 a minor revision, but a change in designation nonetheless, to HDMI 2.1a.
AVPro Edge sponsors AVPro Academy regional trainings and presentations at CEDIA Tech Summits, affording the integration community opportunities for keeping pace with new technology convergence. HDMI 2.1a crosses the HDMI 2.0 series Rubicon, ushering forward a profound collection of advancements to enhance the viewing experience; it is still not, however, sans pitfalls.
Below I detail these features and the effect they impart on the future of video distribution and playback.
HDMI 2.0/a/b
The previous major leap forward with HDMI was the step up from the 1.3/1.4 versions maximum data rate of 8.16 Gbit/s and maximum transmission bit rate of 10.2 Gbit/s to version 2.0 with a data rate of 14.4 Gbit/s and a maximum transmission bit rate of 18.0 Gbit/s, suitably positioning HDMI 2.0 into the UHD era. While teasing BT. 2020 color space, no consumer products emerged for that advantage. The audio sample frequency doubled to 1536kHz, and the format made provisions for 32 audio channels.
HDMI 2.0 retained TMDS encoding (Transition-Minimized Differential Signaling) for video signal transmission as in previous versions, relying on powerful clock recovery in the receiver for high skew tolerance to accommodate longer cable lengths.
Version 2.0a, released in April of 2015, added HDR10 support, while version 2.0b in December 2016 added capabilities for hybrid log-gamma (HLG). Worthy of mentioning is the establishment of the HDMI Forum, by HDMI founding members, in October of 2011. From October 25, 2011, forward, additions and changes to the HDMI specification are the responsibility of the HDMI Forum, with more on that to follow.
HDMI 2.1/HDMI 2.1a
By the numbers, data capacity for HDMI2.1 catapults to a maximum data rate of 42.6Gbit/s and maximum transmission bit rate of 48.0 Gbit/s, and how this is accomplished within the constraint of using the same 19 pin form factor is remarkable.
As with previous version updates initiating substantial changes, implementation often trails press day fanfare by years. HDMI 2.1’s entry delay was hampered by chip availability prior, and unrelated to, the pandemic.
A quantitatively significant leap in bandwidth is made possible in HDMI 2.1, in part attributable by shifting from three data channels of 6 Gbps (the 18Gbps total for HDMI 2.0) into three data channels each doubling the signal rate to 12 Gbps. How the data is structured also changed, using a packet-based model embedding the TMDS clock signal into the data 3 lane, and converting the TMDS clocking channel (which in previous HDMI versions never transmitted audio, video, or data), into a fourth data channel of 12Gbps. Encoding efficiency gained in the packet-based format allotted more bandwidth to data transfer compared to TMDS, enabling the four channels an aggregate maximum data rate of 42.6 Gbps, nearly three times that of HDMI 2.0 with a maximum transmission bit rate of 48.0Gbps.
Fixed Rate Link
This rearrangement to the physical layer architecture, increasing HDMI data transfer, is referred to as Fixed Rate Link. Ranked from FRL1 through FRL6, it is broken out to define the number of “lanes” (using paired wires contained in the HDMI cable) and the lane rate bandwidth of each, expressed in Gbps. Combinations range from a minimum of three channels of 3Gbps with FRL1, to a maximum of four channels of 12Gbps for FRL6.
Embedding the TMDS clock signal within FRL signal packets enables the bandwidth density necessary for HDMI 2.1 to host features like Dynamic HDR, Variable Refresh Rate, Link Training, and more. Despite the transmission mode of AC-coupled FRL replacing DC-coupled TMDS, HDMI 2.1 is not exclusively free of using TMDS. FRL maintains backward compatibility were data rates to fall under FRL minimum support range. TMDS fallback encompasses FRL1 and FRL2, although a particular device may not support all possible signals in those rates.
AVPro Edge introduced the world's first 8K matrix switcher, the AC-MX-42X, which operates within Fixed Rate Link 5 parameters.
Link Training
Link Training, enmeshed in FRL for HDMI, is a protocol for switching between TMDS and FRL and is intended to create reliable and stabilized communication between a Source and a Sink. Six Link Training states exist in HDMI ultimately governing the quality level of the image to be displayed. Link Training plays no direct part in image fidelity parameters rather, it parses EDID information from the sink on behalf of the source, verifying FRL support and confirming the maximum compatible rate of data that can be exchanged (and other display criteria such as timing, color depth, etc.). Link Training can assess HDCP status but that is optional. The source initiates Link Training, with the Sink negotiating for a specific FRL rate. If the Sink requests a new Link Rate, LTS: 4 (Link Training State) can be used to change FRL rates. Upon agreement, FRL transmission commences with LTS: P (Passed). If Link Training fails (LTS: L) TMDS is initiated, with an image of some construct designed to appear.
It is at this juncture where integrators should pay heed and consider test equipment capable of verification for system signals.
As an example, say you have an enthusiast gaming client, and your design comprises of a leading game console, next Gen 8K display, switchgear for multiple locations in the dwelling, and Ultra High Speed cabling. At 4k/100/120fps, 12-bit 4:4:4 RGB, it’s a toe-tip over the 48Gbps threshold by a scant .11Gbps. Link training suppresses the signal to a still fantastic 40.1 Gbps. But with HDMI 2.1 and HDMI 2.1a, additional features can further reduce bandwidth, with the “Info” button on the display remote control alerting your client to less than state-of-the-art performance. While testing prior to deployment has always been a best practices strategy and manufacturers such as AVPro Edge sister company, Murideo, engineer and build affordable, field portable test instrumentation, design verification before large-scale, duplicate purchases could prove to be a hidden miracle.
And as I mentioned at the start regarding tech support…test gear can mitigate company downtime, isolate issues to their origin and even prevent them from arising.
High Frame Rate
The quantum uptick in bandwidth supported by HDMI 2.1 gives manufacturers freedom to develop new devices outputting or receiving resolutions as high as 10420 x 4320…10K. While unlikely the market will see ultra-high resolutions eclipsing 8K any time soon, 100/120fps is supported with HDMI 2.1, and game consoles available now are capable of outputting 4K/120fps.
As an entertainment medium, HFR movies have been met with a fair amount of skepticism. The series of Peter Jackson’s The Hobbit movies were critically received but largely commercial disappointments. Ang Lee’s two 120fps movies, Billy Lynn’s Long Halftime Walk and Gemini Man exhibited a hyperrealism which most people found disconcerting. The lack of content from Hollywood for this domain suggests that it isn’t commercially viable, though James Cameron is said to be shooting the Avitar sequels in HFR so perhaps as the consumer side of things is catching up with theatrical technology, we may see more in the future. High Frame Rate should not be confused with High Refresh Rates.
Frames per Second vs. Refresh Rate
There is an element of confusion circling around high frame rates and “high refresh rates”, so much so the terms are often used interchangeably. Each term refers to a similar visual concept, but different tasks performed by different hardware devices.
Movement depicted on a display refers to the number of consecutive still images or frames and the rate at which new frames appear per a SMPTE time-coded increment, which is one second of clock time expressed as frames per second, or fps. The greater the number of frames, the higher the frame rate.
Refresh rate refers to the number of times per second the display creates a new viewable image and is expressed in Hertz (Hz.).
To illustrate the difference between frame rate and refresh rate, merely pause active content. Frame rate halts, however, refresh continues. Simply put: Frame rate is in content, refresh rate is in hardware.
Though widely known, for context I will briefly make the addition that a manufacturer’s marketing information pertaining to refresh rate may not (read: rarely, if ever) match the engineering specifications of a television. Sadly, “biggest numberitis” is a malady that still persists in the marketing world and the cure eludes television manufacturers. Motion Rate, MotionFlow, and TruMotion refer to the jargon chicanery TV makers use to describe features such as black frame insertion or framerate interpolation, and the associated numbers inflate the actual refresh rate of the display. It’s a minefield rife with danger in today's HDMI 2.1 world, as the 120Hz goes-faster-model description on the box may refer to an actual 60Hz operating panel inside.
HDMI Forum Variable Refresh Rate
It will probably enter the lexicon as VRR, but the HDMI Forum prefix is key for HDMI 2.1 products. Variable Refresh Rate (VRR) is nothing new to the PC-based gaming community, having begun life – and eternal conquest – as graphics card maker AMD with FreeSync battles Nvidia and G-Sync for format domination. PC users found the VRR format available to them was dependent on who manufactured the graphics card inside their PC. As such, a wide belief extends this to TV support for either FreeSync or G-Sync to properly support VRR game consoles. Just throwing this out as a free life lesson, never make something thought to be proprietary into a tattoo. HDMI Forum VRR (or if you prefer, HDMI VRR) is the consequential standardization of this process, with the HDMI group establishing their own version of the VESA Adaptive Sync protocol. Next Gen consoles Xbox Series X supports HDMI VRR in conjunction with FreeSync VRR, and April 25th Sony announced VRR support for HDMI 2.1 VRR compatible TVs for PS5, a presumed alignment with HDMI Forum VRR.
For ardent gamers, VRR is a critical feature to image clarity. Next Gen consoles can send 120fps information to a Television that is capable of displaying that frame rate (see Refresh Rate above…trust, but verify!), so crucially for gaming, the match-up seems idyllic: the gaming device will send video frames to the display at a rate it can efficiently process them. In reality, the in-game frame rate fluctuates perpetually as plot scenarios evolve. In an F1 driving simulation, a sequence with the car being prepared in the garage parses out frames less frantically than a 320kph (biggest numberitis shamefully in use) pass in the Belgian countryside.
An HDMI VRR-featured television allows the frames to be displayed at the rate they are sent, not at a fixed rate, adjusting the display’s refresh rate in real-time. Sudden changes in frame rate cause image tearing on displays with static refresh rates. HDMI VRR dynamically adjusts refresh rates to match rapidly changing frame rates for a consistent, smoothly transitioning image.
A VRR-capable display will vary refresh rates anywhere from 40-120Hz.
Xbox Series X and PS5 both are HDMI VRR capable, so the display will require this feature designation as will any device the signal traffics through, such as an AVR. Ultra High Speed cables are necessary at all device connection points.
While an exciting feature aimed at gamers, be it known VRR technologies are yet to be deemed perfected. In some instances, they can affect input lag and may not completely eliminate all screen tearing artifacts.
Auto Low Latency Mode (ALLM)
With some display manufacturers, getting into game mode was like trying to get into Studio 54 back in the day…you had to know somebody. Deep dives into normally bypassed parts of the TV menu were necessary to access and activate Game mode, often temporarily disabling settings directly relative to image fidelity on non-gaming sources. For HDMI 2.1 feature set televisions, ALLM signals the display to switch into Game mode as well as activate all the necessary settings to reduce delay and lag. It essentially optimizes pixel processing for best latency or best pixel processing. It won’t contribute to your gaming experience, only properly prepare for it. ALLM isn’t activated every time a game console is powered up, as it is capable of detecting game play, media streaming, or disc use. Not all televisions with HDMI 2.1 ports necessarily have the feature, as some Sony models launched required updates to enable ALLM.
Quick Frame Transport
Quick Frame Transport (QFT) is designed to improve the transfer rate of video game data from console output to display input. Display latency is measured as the time necessary for a frame of data at the source GPU to be displayed on the television screen. The data path is longer than simply the HDMI cable, as it includes output circuitry in the source device, TV processing, and screen refresh. QFT accelerates the transport time between the source HDMI 2.1 output port to the display’s HDMI 2.1 input port at a higher rate than normal. In gaming terms, the delay between pressing a controller button and discernible movement on the screen is reduced, with any incremental reduction in latency considered quantitative in the gaming world.
HDMI Quick Media Switching (QMS)
Using HDMI VRR, HDMI 2.1a Quick Media Switching eliminates the momentary screen blackout when HDMI sources are switched.
The black interval comes into play when refresh rates differ between video sources making employ of the HDMI VRR feature an elegant solution. One caveat, however, is the black screen will still occur if resolutions are different between sources.
Display Stream Compression 1.2
Companion to higher bandwidth requirements at the as-yet, fringe side of HDMI 2.1, specifically 8K and beyond, are new compression methods for data transport. Most common is Display Stream Compression (DSC). AVPro Edge carefully examined DSC when designing our first products and determined the criteria for DSC relied too heavily upon declaration that tangible differences to static images were negligible, heavily skewing the conclusion that it was an artifact free platform. As a result, AVPro Edge’s in-house engineering team designed ICT, our Invisible Compression Technology algorithm.
At first glance, DSC 1.2 appears to be very similar in concept to Software Defined Video-over-Ethernet (SDVoE) in especially using the approach of splitting a video frame into equally-sized slices along horizontal and vertical, and then processing these slices in parallel. Theoretically, it can be considered similar to interlaced analog video and scanning lines. In another facet, DSC 1.2 is said to be able to increase maximum color bit depth to 16-bit with YCbCr 4:2:0 and YCbCr 4:2:2 without the need for conversion to RGB first, taking on the persona of Color Space Conversion.
With bandwidths and resolutions to be addressed by this compression technology at 64Gbps and 10K respectively, much can change prior to implementation and many years may pass before it is needed.
HDMI Enhanced Audio Return Channel (eARC)
HDMI 2.1a features Enhanced Audio Return Channel (eARC) as the follow-up generation to ARC. The primary benefit with eARC is a robust boost in audio bandwidth and speed, from approximately 1Mbps to 38Mbps, supporting up to 32 audio channels including eight-channel, 24-bit / 192kHz uncompressed data streams. High bitrate audio formats from Blu-ray, 4K Blu-ray and streaming providers including Dolby TrueHD, DTS-HD Master Audio plus object-based formats Dolby Atmos and DTS:X are compatible with eARC. Simplified connectivity with reliable HDCP negotiation were design goals with the actual audio path clear of CEC pathways, avoiding unwanted control protocols. Audio to a Television originating from DBS, CATV, connected source devices, or internal streaming apps can be routed back via eARC to a surround system or to a sound bar through a single HDMI cable.
While connected devices are not specifically required to be HDMI 2.1a certified, certification practically assures eARC support. Manufacturers can produce devices compatible with both ARC and eARC however, eARC is not defined as backwards compatible with ARC.
Dynamic HDR
While HDMI 2.1 standardized the transport of dynamic HDR metadata over HDMI, it only formalized dynamic metadata interfaces already in use with HDMI 2.0 by Dolby Vision and HDR10+. Both formats function properly with HDMI 2.0 and HDMI 2.1a does not provide any additional capabilities. Standardization incorporates compliance testing to ensure Static and Dynamic HDR metadata can be exchanged through the HDMI interface.
All the benefits of static HDR10, HDR10+, and Dolby Vision remain unchanged in the HDMI 2.1 ecosystem.
Source-Based Tone Mapping
The “a” in HDMI 2.1a refers to the inclusion of Source Based Tone Mapping to the HDMI 2.1 specification, introduced on January 5th of this year, prior to the start of CES 2022. The HDMI Forum members, recognized that some streaming services were combining SDR, HDR, Dynamic HDR and graphics overlays simultaneously on one screen. In that mixed-bag environment, metadata is absent as it cannot be generated for a composite image. The source device optimizes the image suspending metadata, preventing the display device from attempting to tone map. SBTM is not a new HDR standard, rather, it is designed to partner with HDR10 and HLG and adjust luminance and color range per display capabilities when presented with this scenario.
Frame by frame, the source signals the display it is tone mapping the composite content to characteristics outlined by the display’s EDID. Signaling stops when the composite image ceases.
Another instance where SBTM is unique and beneficial is the enabling of “plug and play” HDR gaming without the need for pre-play calibration. From reading the EDID, the gaming device knows the display’s preferred color volume, and the gaming engine can appropriately tone map the content. No dynamic range in the display needs to be reserved for pixel values outside the specified color volume, applying full capability of the display to the content being played,
So, there can’t be a downside, correct? As with other unique HDMI 2.1 features, SBTM is a manufacturer-optional feature with no requirement to provide support. The same options apply to HDMI VRR, ALLM, and extended resolutions, bandwidths or 120Hz refresh support. Perhaps this is fodder for a future article, but rescinding listing of the HDMI version numerically and resorting to a provided feature list is a tedious imposition upon the consumer, and perhaps that might be the HDMI Forum’s ultimate intent.
Every square inch of carton real estate will be required to list ALL supported features. The HDMI Forum’s return argument, weak as it appears, is that this is always how its standards have worked and features that are optional give manufacturers flexibility in the functionality they offer.
Summary
Without question, the features scheduled to become active with HDMI 2.1 provides the consumer with beneficial, ease-of-use image enhancements in lockstep with a massive boost to the format’s bandwidth signal handling capabilities, when all devices and cabling are within the 2.1 ecosystem. HDMI 2.1a is considered a minor upgrade, adding source-based tone mapping to the HDMI 2.1 specification, yet significant of itself to warrant a designation change.
In the Link Training section, I commented upon the value of test gear and procedures designed to verify if specified products deliver their expected performance when installed into a new or existing system. That sentiment interleaves with the notion expressed at the end of Source-Based Tone Mapping that deciphering what features are on-board particular products by list only, sans any inclusive, governing numerical designation, will at times seem and ultimately prove futile, contributing to a collision of confusion we haven’t seen as an industry in quite some time.
As end-users yourselves, enjoy the new features of HDMI 2.1a, but as integrators, buckle up!
By Matt Murray, Chief Technology Officer
AVPro Edge
