By Jeremy Laird on May 15th, 2014 at 9:00 pm.
4K, 6-bit, 8-bit and 10-bit panels, G-Sync n’ FreeSync n’ Adaptive-Sync, 120Hz-plus refresh, DisplayPort 1.2 and 1.2a, backlight modulation, multi-stream vs single-stream and IPS vs PLS. The PC display market is completely out of control. But in a good way. Things are developing faster now than at any time I can remember since getting into this game. And I am incredibly, astonishingly, implausibly old. The Atari 2600 was still on sale (just) when I achieved something approaching sentience. I still haven’t truly recovered from the 2600’s piss-poor Pac-Man port. Anywho, the last week or so has seen some really interesting developments in the monitor market, including the announcement that AMD’s FreeSync tech is moving into the mainstream courtesy of official VESA status and the appearance of a cheap Samsung 4K monitor with 60Hz support. High time, then, to pull together the state of play in PC monitors into something we can all understand. Well, hopefully.
Before we start, a heads up on the sections below. I’ve stuck some headers in so that you can quick-jump to the bits that interest you most. I realise not everyone is an anal about colour depth as I am. Well, sometimes, at least.
Also, this is not meant to be a definitive guide to the basics of LCD monitors. Luckily, I’ve got that in the bag already. You can read it here. Instead, this is more of a primer to help you keep up with the latest developments
Driving 4K panels
First up, 4K. I’ve touched on 4K before, so I won’t get too granular with the basics other than to regurgitate the simple idea that it refers to screen resolutions with roughly 4,000 horizontal pixels. Totes amazeballs etc in terms of visual fidelity in games and elbowroom on the desktop. Problematical in a lot of other ways.
Prices for 4K panels have plummeted specutacularly since Asus wheeled out its 32-inch effort for £3,000 back in November
Obviously, you’re going to need one hell of a GPU to push out what amounts to four times the pixels of a common-or-garden 1080p panel. But more that that, it’s a problem driving 4K in simple 2D mode. DVI, even in high-bandwidth dual-link mode, lacks the necessary pixel-pumping minerals, as do most versions of HDMI.
DisplayPort 1.2 and 4K
In fact, even DisplayPort 1.1 doesn’t support 4K as a single stream. Put simply, you need a graphics card with DisplayPort 1.2 support to run a 4K monitor as a single stream rather than multi-stream mode, which is rather like running two monitors or desktops on a single panel and creates all kinds of issues.
Anyway, you’ll also need a monitor that supports DispalyPort 1.2 to do the single-stream thing and as it happens the affordable new Samsung U28D590 (a 28-inch 4K model currently available for pre-order here for £500) will do just that. It will also punch out 4K at a proper 60Hz.
DisplayPort 1.2a and Adaptive-Sync
However, what the Samsung U28D590 doesn’t have is support for DisplayPort 1.2a (yes, 1.2a), which isn’t a huge surprise, because 1.2a isn’t out yet. But it is a critical omission because with 1.2a comes support for what could be the killer gaming feature for the next few years, namely Adaptive-Sync.
It’s a new industry-wide standard from VESA and it’s basically AMD’s FreeSync tech made mainstream. FreeSync, of course, was AMD’s more open take on Nvidia’s proprietary G-Sync tech which involves dynamic refresh syncing between the GPU and display.
Nvidia’s G-Sync is truly a glory to behold. Just a pity it’s all so proprietary…
I’ve seen G-Sync in action and it’s genuinely revelatory. Kudos to Nvidia for that. Without G-Sync, Adaptive-Sync probably wouldn’t be happening. But an open standard that doesn’t restrict you to a particular GPU vendor is infinitely preferable. And now it’s happening. Yay.
High-refresh vs adaptive refresh
We’ve been waiting years for something to come along and improve upon good old 60Hz panels with v-sync enabled in terms of gaming smoothness. Suddenly two turn up at around the same time.
I give you super high refresh rates of 120Hz and beyond and dynamic syncing between the display and the GPU. But which is better and are they antagonistic or complimentary?
The first thing to grasp is that you need a very high performance GPU to get the full benefits of a high refresh display. That’s especially true if you are also running very high resolutions. It’s not much help having a 120Hz panel if your GPU is chugging along at 40fps or whatever.
In that sense, dynamically syncing the display with the GPU is my preference. It allows smooth rendering at much lower frame rates and that means more future proofing for your rig and less money spent.
Pac-Man failed to point up or down. Potentially damaging to the formative gaming mind…
Of course, dynamic syncing at higher frame rates is even better. But I reckon the biggest benefit for most of us is in smoothing things out at lower frame rates. So, I’d prioritise dynamic syncing over high refresh if forced to make a choice.
PLS vs IPS (and VA)
This one is quick and easy. If you see PLS in the spec list of a monitor, that’s a good thing. It’s essentially Samsung’s take on IPS, which has emerged as the favoured LCD panel tech for high performance PC monitors.
IPS isn’t the best by every metric. TN panels have faster response. But overall, it’s the best compromise. A few other companies are also doing their own take on IPS, including AU Optronic’s AHVA (Advanced Hyper-Viewing Angle) tech, just to confuse things a bit more. Sorry. But with any luck, you’ll also see something like ‘IPS-type display’ which will help guide you.
Without getting bogged down in a basic guide to display types (again, go here for the basics), a few VA monitors can still be had. Just be aware that VA is the slowest panel type in terms of pixel response. Overdrive technlogy can offset that, but usually at the price of nasty inverse-ghosting side effects.
We’ve done this in detail before, but the issue here involves the manner in which LCD monitor backlights are modulated for brightness. Most monitors do this by turning the backlight on and off. Leave the backlight on all the time for full brightness. Switch it on and off rapidly and various frequencies to achieve a scale of brightness.
For those of you who habitually place a fan betwixt self and display, flicker-free tech will be quite the boon
In theory, if the light is switch fast enough, the human simply see something dimmer. But some monitor manufacturers claim this kind of modulation can induce flicker. A bit like the rainbow effect and DLP projectors (which I am personally quite sensitive to), the extent to which you will notice this varies from person to person. I’ve never detected flicker on an LCD panel, myself.
Anyway, you can now buy affordable flicker-free monitors like the BenQ GW2265HM. So if you think this is something that bothers you, here’s a list of flicker-free panels.
Display colour depth
Finally, while we’re talking Samsung U28D590, the official specs claim 1 billion colours, which in turn infers 10-bit-per-channel colour depth. And yet this is a TN panel. Is this possible? What does any of it mean?
I’m going to slightly contradict myself here and dig a little deeper into basic principles. In really simple terms, colours on a digital display are generated using three primary-coloured subpixels – red, green and blue, hence ‘RGB’. And the total number of available colours is a factor of how many intensity levels a display can achieve with each of those primaries. Combine those differing intensity levels in the three primaries and you have your full colour palette. That’s how colour perception works in this context.
Now those ‘intensity levels’ or brightness levels for each subpixel are achieved in discrete steps, with completely off at one end and max brightness at the other. In theory, there is no limit to how many steps you can take between off and on. The better the display, the smaller each step and the greater the overall number of steps per primary.
Now, a 6-bit display can do 64 intensity steps per primary (6 bits have a maximum of 2 to the 6th power = 64 binary values). So that’s 64x64x64=262,144 colours. 8-bit panels offer 256 intensity levels per colour channel. Do the maths and you get 16.7 million colours.
Stand approximately 600ft away and this collection of RGB subpixels will look like a white box. Possibly…
10-bit pushes that up to 1,024 intensity levels per channel and a little over a billion colours overall. Where things get complicated is the use of dithering to simulate greater colour depth. The idea here is to rapidly oscillate a given pixel between two colour states fast enough that the human eye is fooled into seeing something in between.
Nice in theory and suddenly allows your 6-bit panel to render at 8-bit levels. Except dithering is never quite as good as native colour depth and almost always introduces visual noise. Look closely at some colours in a dithered display and you’ll see them ‘fizzing’ away as they hop about between intensity levels.
Moral of the story? True colour depth matters. I haven’t seen the new Samsung TN panel and its alleged 8-bit native capability, but early reports suggest it’s very impressive. Here’s hoping.
The elevator pitch
In the meantime, what should you conclude from all this? It’s complicated, but I’d say that if you can hold out for a while, it’s worth waiting for all this DisplayPort 1.2a and Adaptive-Sync stuff to shake out.
I generally tend to upgrade displays less frequently than any other component. There’s a good chance that if you hang on for a month or six, you’ll end up with both a pixel count and refresh-tech support that will see you good for several years to some.
4K and DisplayPort 1.2a.