WEBVTT

00:00:00.000 --> 00:00:06.400
HDMI made the entire internet mad recently by introducing a confusing new branding scheme

00:00:06.400 --> 00:00:11.200
when the old one was serving everyone just fine. But hold on, there's more because it's about to

00:00:11.200 --> 00:00:18.080
get more confusing with the introduction of HDMI 2.1a. There's an A. So let's talk about what's new

00:00:18.080 --> 00:00:24.720
in the HDMI 2.1a spec. And as you may have gathered from the fact that it's called 2.1a instead of

00:00:24.720 --> 00:00:31.920
2.2, the list of changes is rather short. In fact, there's only one new feature and it's called

00:00:31.920 --> 00:00:36.640
source-based tone mapping. That might sound like some kind of gimmick you'd pay extra for at a

00:00:36.640 --> 00:00:42.160
tanning salon, but it actually has to do with the way devices handle HDR. Tone mapping is the

00:00:42.160 --> 00:00:48.160
process of taking an HDR signal and adjusting it to match the capabilities of whichever display you

00:00:48.480 --> 00:00:55.760
be watching that content on. You see, many HDR-capable TVs actually are not able to show an

00:00:55.760 --> 00:01:01.600
HDR image as it's originally mastered. Oftentimes, the original video is mastered at 1,000 nits of

00:01:01.600 --> 00:01:06.720
brightness or even higher, but most consumer displays, even many that are marketed as HDR,

00:01:07.280 --> 00:01:13.840
cannot get that right. Additionally, HDR content is often mastered for a wider color gamut than

00:01:13.840 --> 00:01:20.320
many TVs can show, meaning they can't display as many colors as are contained in the source material.

00:01:20.320 --> 00:01:25.920
Without tone mapping, the TV would clip the image. That is, it would simply discard the information

00:01:25.920 --> 00:01:31.760
it can't display from the video signal, and the image would instead have very incorrect colors,

00:01:31.760 --> 00:01:36.880
or just parts with no information, and portions that would be blown out to absolute crap so you

00:01:36.880 --> 00:01:41.520
can't even see them at all. So tone mapping effectively takes color and brightness information

00:01:41.520 --> 00:01:47.760
and maps it to a value that the TV can actually show. This allows the TV to approximate the original

00:01:47.760 --> 00:01:53.280
image accurately enough so that it looks okay on your screen instead of a distorted mess.

00:01:53.280 --> 00:01:58.000
So for example, a signal with 1,000 nits peak brightness can be reduced so that the brightest

00:01:58.000 --> 00:02:03.600
parts of the image are now, say, 400 nits. Everything goes in line with that.

00:02:04.240 --> 00:02:10.080
Now back to HMAT 2.18. Traditionally, tone mapping is handled by the display itself,

00:02:10.080 --> 00:02:16.400
but the source-based tone mapping in HDMI 2.1a offloads some of this responsibility from the

00:02:16.400 --> 00:02:22.160
display and moves it to the source, whether the source is a GPU, a streaming box like an NVIDIA

00:02:22.160 --> 00:02:30.000
Shield, or a Blu-ray player. But why exactly would we want this? Dolby Vision HDR actually already has

00:02:30.000 --> 00:02:35.920
a proprietary option where you can select either TV-led or player-led tone mapping,

00:02:35.920 --> 00:02:41.040
and tests have shown that tone mapping looks better when it's handled by the display itself,

00:02:41.040 --> 00:02:47.440
as the TV is adapting to its own characteristics, as well as data from its ambient light sensor if

00:02:47.440 --> 00:02:52.000
it has one of those built in. Having the source handle tone mapping means that it's doing the

00:02:52.000 --> 00:02:57.040
mapping based simply on what the TV reports its capabilities to be, meaning the adjustments the

00:02:57.040 --> 00:03:04.240
source makes may not be as accurate. So why do we want this? Well, it's because there is a big plus

00:03:04.240 --> 00:03:11.440
to source-based tone mapping, lower latency. Even if something might look a bit better if the TV

00:03:11.440 --> 00:03:17.280
is handling the tone mapping, this additional processing that the TV has to do adds latency,

00:03:17.280 --> 00:03:22.720
which can be a significant drawback for gamers who want as small of a delay between pressing a

00:03:22.720 --> 00:03:28.480
button on the controller and seeing an action happen on the screen as possible. So source-based

00:03:28.560 --> 00:03:35.280
tone mapping, or SBTM, is a sensible addition to HDMI's increasingly gamer-focused feature set.

00:03:35.280 --> 00:03:40.320
Additionally, source-based tone mapping does allow more of a plug-and-play experience for the user,

00:03:40.320 --> 00:03:44.480
as since it allows the source to read and understand the display's capability,

00:03:44.480 --> 00:03:50.400
the user won't have to spend as much time adjusting picture settings to get their HDR content to look

00:03:50.400 --> 00:03:56.640
right. For now, it sounds like SBTM support can be added to existing devices via firmware updates,

00:03:56.640 --> 00:04:01.680
but that'll be up to device manufacturers. So if you want the feature without spending

00:04:01.680 --> 00:04:07.920
extra on new equipment, your mileage may vary as AV manufacturers have a track record, let's say,

00:04:07.920 --> 00:04:14.560
of being very slow with updates. Additionally, HDMI 2.1a is going to suffer from the same stupid

00:04:14.560 --> 00:04:21.440
naming issues we saw with the original HDMI 2.1. All current HDMI devices will transition over to

00:04:21.440 --> 00:04:28.880
the HDMI 2.1a name, but they don't all have to support every new feature of HDMI 2.1 or 2.1a,

00:04:28.880 --> 00:04:33.920
so make sure you read the spec sheet to see if the features you care about are actually present.

00:04:34.480 --> 00:04:37.920
I mean, you never spend money without reading the fine print first, right?

00:04:39.040 --> 00:04:43.360
Right? So thanks for watching guys, if you liked this video, hit like, hit subscribe,

00:04:43.360 --> 00:04:48.080
and hit us up in the comment section with your ideas for topics that we should cover in the future.
