Faking Freesync Support??? – Pixio PXC277A 1440p 165Hz Gaming Monitor Review

As best as I can tell, this Pixio PXC277 Advanced doesn’t support Freesync. I mean, they say it does on their website and product listings, and in the on screen menu, and NVIDIA lets me enable adaptive sync, but the data from my Open Source Response Time Tool shows it does not adaptively refresh. Let me show you what I mean. This graph shows the light level over time during a response time transition – the transition doesn’t matter here, what does is those little waves after the overshoot spike. Those waves are each new frame being drawn – normally that’s because they are using a PWM backlight but what’s important here is that this shows us how long there is between new frames. At 165Hz, this monitor’s maximum refresh rate, that means there should be 6ms between frames. If we measure those little waves here we find each of them is 6ms apart. Perfect, just as expected.

Now for good measure, I set the monitor to run at 60Hz in Windows and ran the test again. This time, those waves are much slower – 16.7ms specifically – which is spot on for 60Hz. So we know that this is a good way for us to check what refresh rate the monitor is running at at any given time. It’s a reliable source of data. So now for the real test. I turned adaptive sync on in the menu, adaptive sync on in the NVIDIA control panel, and set OSRTT to limit the framerate of the Unreal Engine 4 game to 60 FPS. I launched the test – verifying it was indeed running at 60 FPS. This is all one take – although I will speed up the test section as this isn’t the important part. Now, looking at the results, specifically the raw data graphs, and manually measuring that wave and… Yep. 6ms. It’s not doing adaptive sync at all. Taking a closer look at that graph, you can see that no matter where you measure, a single wave takes 6ms, the exact same as when running at 165Hz natively (and with a 1000FPS cap in Unreal Engine). Because I know someone will spot it and ask, I did also re-test with GSYNC set to “Fullscreen & Windowed Mode” just in case and no, it made no difference. It still doesn’t adaptively refresh. For good measure, I also tested again at 100 FPS, just in case the adaptive sync range didn’t include 60FPS for some reason. Same again, no difference. Also, to be clear, this is using DisplayPort.

For the sake of absolute completeness, I also built up a system with an AMD GPU to run the test again and again here’s the full recording. Adaptive Sync is enabled in Radeon Settings – it lets me enable it so the display is reporting some level of support for the feature – I’ll again fast forward through the test but as you can see on the top right it’s locked at 60 FPS throughout the test. Now with the results open, looking at the graph and manually measuring the waves and, again, it’s right around 6ms, not the 16.7ms we should be seeing.

This isn’t the first time a reviewer has caught a problem like this, Samsung was caught doing the same thing in 2021, although that report didn’t get quite as much publicity as it really should have. I’m hoping that this is a technical oversight, and one that can be addressed with a firmware update, but regardless it’s not good to be – as far as I can tell anyway – faking compliance with one of the headline features of the monitor.

Now moving on from that mess… When it comes to the actual response times, I’m afraid this doesn’t get any better… Pixio, rather boldly, claims that this VA panel has not only a 1ms response time, but a 1ms grey-to-grey response time. See most manufacturers get away with calling every monitor a “1ms” monitor, because they quote the pointless MPRT or moving picture response time figure, where when using the backlight strobing mode the panel is only on for 1ms per frame, so while it’s a useless measurement, it’s a misrepresentation rather than an outright lie. If you actually test the thing though, even on the absolutely horrific “high” overdrive mode, and completely ignoring the insane overshoot, you still only get an average of 4.7ms. If you include the overshoot time – which considering just how bad that actually is, is pretty necessary here – you get an average of 10.3ms. That’s pretty far from the “1ms” claimed figure.

Here’s what that looks like in slow-mo. It does speed up the panel considerably in the mid-range – the darker shades still take an age to transition – but you are also left with considerably overshoot, or inverse ghosting. It’s often more distracting than regular ghosting as instead of just partially faded versions of previous frames giving you a blurred effect, this has inverted colours which are often shining brightly at you.

On the more reasonable, yet still weirdly overdriven in the mid range transitions, “middle” overdrive mode you get an average of 8.6ms, or 6ms if you ignore the overshoot time. This is likely the mode I would use, as while it does still have some overshoot on lighter shades, generally speaking it’s less noticeable as it helps improve the initial response time compared to the “low” setting. That mode is fine too, if a little slower with an average of more like 8.7ms or 8.1ms if you ignore the overshoot time.

The other measurement that OSRTT can do is latency – specifically the on-display latency as in how long it takes for a new frame to be processed and start to be drawn. A good result here is anything less than one frame of latency – and a really good result is where basically none of the results take longer than one frame, so in this case a good result would be an average of 3ms, with no results being higher than about 6ms. So, how did this do? Oh. That’s.. Awful. It averaged almost exactly two frames worth of delay, with no results being on the first frame – the lowest we got was a little over 7ms, and the worst was over 18ms. That’s pretty bad.

I also use a Time Sleuth as it measures on display latency a little differently and normally has better resolution since it controls when it outputs new frames. The only problem here is that it reported an astonishingly bad 30ms of on display latency. Now this is with a 1080p signal and over HDMI, so lets run OSRTT again using HDMI at 144Hz to make sure the HDMI port isn’t actually slower and.. Oh dear. It is. While the HDMI port can only run at 144Hz, meaning a frame is now 6.9ms long, we would expect at most a slight increase to around 14ms of on display latency, but what we get is a considerable increase to over 16ms! In fact, it’s now solidly 2.5 frames worth of latency, as only a handful of the 100 results fall within the second frame now. So if you were planning on gaming with a console on this monitor, you are going to have an even worse experience than if you were PC gaming over DisplayPort.

Um.. yeah so it’s almost a week later and I’ve got something you need to see. If you haven’t watched it already, Wendell from Level1Techs reviewed this very monitor and generally found it a good option. Long story short, we collectively tested our units over the course of a couple days and have some pretty confusing conclusions. So first off nothing I tried would make this adaptively refresh. This runs at the frequency you set in your driver or in Windows, and will not adaptively refresh. I’m pretty confident on that one. What I’m less confident about is what the hell the scaler is doing. You know how I said the Time Sleuth reported something ridiculous like 30ms of latency over HDMI? Yeah Wendell found that too. Except, if you disable Freesync and overdrive, it’s down to just one or two milliseconds. Yeah! That’s the same for DisplayPort too, here’s the OSRTT results with Freesync enabled in the on screen menu, now here it is with Freesync disabled in the menu. Just to make this clear, here that is in a bar chart. The maximum with Freesync off doesn’t even come close to the MINIMUM with Freesync on in the menu. To be clear, the HDMI port is still a touch slower, but that’s mostly thanks to the slower 144Hz refresh rate rather than the 165Hz you get over DisplayPort, so that’s all good.

This revelation changes how I feel about the monitor quite a lot – including what the gaming experience feels like. With Freesync on the monitor is just a constant disadvantage in anything even remotely competitive. With it off, it’s much more like a normal VA monitor experience – that is to say kinda naff. The slow response times – especially in darker areas – means motion is blurred, which makes hitting targets – be that enemies in FPS games or apexes in racing games – much harder. That’s not unique to this panel though, and at this price point that’s more to be expected. It’s not impossible to game on obviously, but a nice IPS is going to be a better experience for sure. Now back to the younger, dumber Andrew…

If you were concerned about the colours of the panel, you’ll be happy to know that despite it wildly missing it’s claimed 97% coverage of the DCI P3 spectrum with more like 88%, that’s still a great result and it looks great to the eye. As you might expect for a VA panel, the blacks are incredibly deep. The contrast ratio is around 4000:1, with a peak brightness slightly higher than quoted at 330 nits. The uniformity is better this time – check out my last Pixio monitor review to see what I’m talking about there – although the gamma curve is a little off. Accuracy wise it isn’t bad, it reported a DeltaE of just over 2, 2.07 specifically, with a couple higher results. It’s not perfect, but it’s a gaming display so that’s more than good enough.

Physically it’s a nice enough, if cheap, design. The stand offers exactly one axis of freedom, that being tilt, and the panel sits pretty low to your desk too, although it does offer a VESA mount on the back so any issues with the stand are easily resolved. The foot doesn’t take up much room and it’s remarkably stable too which is always good. The styling is a little gamer-y, but that’s fine. It still comes with the wonderfully labelled I/O ports, specifically one HDMI 2.0 port, two DisplayPort 1.2 ports and a USB port that’s only there for firmware updates. The on screen menu is controlled by a much nicer joystick style switch than the PX275C Prime, and the menu itself is nicer too. It’s still pretty basic, although you do get four total overdrive modes if you include off. You can use the “MPRT” mode, aka backlight strobing, but it disables Freesync – although as we’ve found out that’s no big deal – and it’s a great way to strain your eyes so I wouldn’t bother.

Hi again! Seeing as the new revelations myself and Wendell discovered changed what I think about the monitor, my conclusion changed too. Well actually it didn’t, don’t buy this. It is cheap, Pixio list it at $240 which puts it firmly in the “budget” category, but for just ten bucks more you can get an AOC CG27G3S which is equally a 1440p 165Hz curved VA monitor that while still likely to have a slow, smeary response time especially in darker areas, should have functioning adaptive sync support and likely much better tuning both to its firmware and to its overdrive modes. If Freesync worked on this I’d say it was a decent enough choice for the price, if a little buggy, but with as best as I can tell no functioning adaptive sync support – despite widely claiming to support “AMD Freesync Premium” – this isn’t worth your cash. Ok, I’ll hand it back to baby Andrew again.