Michaelius
Supreme [H]ardness
- Joined
- Sep 8, 2003
- Messages
- 4,684
I don't know if anyone else is reading neogaf but if you have occullus rift drivers instaled then they will cause conflicts and problems with G-sync and need to be uninstalled.
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Ah, that's different. If they're pinning the overdrive to maximum and running one of those pixel un-stickers, even if they're technically within spec, it is sort of asking for trouble. I've never thought about this before but I speculate the driver current demands are higher and spikier when you're trying to change every single pixel from max one color to max another at a high rate.Not just turned on, but repeatedly cycling through some utility designed to "unstick" pixels or subpixels.
Good point. We would need to know how much overhead is required for GSync. There would be packet loss that could explain the anomalies if the bandwidth is being oversaturated. Maybe this is why we don't see other 144Hz 1440p panels in the wild?Looks like I won't get mine till the next shipment arrives, oh well.
I did want to give you guys a heads up on a theory for the issues at 144hz though, A while back, when the monitor was announced, people had doubts that Display Port 1.2 had the bandwidth required to keep the monitor going at maximum settings.
Display Port has a theoretical maximum bandwidth of 17.28 Gbits/s (or 2.16GB/s) while the maximum bandwidth required to run this monitor at 2560x1440 at 144fps with a 32bit color depth (not the monitor, the graphics setting) would be 1.98GB/s.
I have no idea as to if it may be hitting the cap, or if the cap is at a comfortable distance (although, if the drivers were stupid and it was sending audio for some dumb reason this would start to cut it very close.)
*for comparison purposes, at 120hz you require 1.65GB/s
EDIT: I just remembered, there maybe some overhead for G-Sync?
Cable length may be an issue. DP recommends a cable no longer than 3 metres to ensure quality of signal. Maybe the closer you approach the maximum bandwidth the shorter the cable you should be using. How long is the DP cable that comes with the PG278Q? Has anyone tried using a shorter cable?Since out of range topic came up I'll mention that my 120hz 27" 1080p samsung monitor has always had this issue. I used to have it on a long run from an amd 6990 along with an ACD 2560x1440. The acd never had any screen blank-blinks, out of range, or out of range on wakeup issues but the 120hz 1080p one did. There were both on long, heavy 24awg mini dp extension cables. When I switched to a nvidia 780ti , I couldn't even run both of them at the same time (the ACD was on a powered dvi to minidp adapter). Even running the 27" 1080p singly, I would often get screen blank blinks and out of range.. much, much more often and annoying than amd. I have to suspect that displayport signals or connectors are not that great in general, and that nvidia's displayport signals/outs might be even weaker. At least on my setup comparing the 6990 to the 780ti - they are very much weaker.
Ah, that's different. If they're pinning the overdrive to maximum and running one of those pixel un-stickers, even if they're technically within spec, it is sort of asking for trouble. I've never thought about this before but I speculate the driver current demands are higher and spikier when you're trying to change every single pixel from max one color to max another at a high rate.
I'm sorry, I was quoting your post to suggest a point about the PG278Q that the amount of bandwidth it is using at 144Hz requires a shorter cable. Should read:Cable length in my case is definitely an issue. I wasn't suggesting that it wasn't. What I was suggesting is that it made the difference in signal strength between my amd and nvidia card obvious since it worked well on the 6990 and fails a lot on the 780ti's dp out.
Thanks for the heads up.
Also be sure to mention whether you have Oculus Rift drivers installed. I've heard reports of them causing issues. I'll see when my panel arrives tomorrow if it affects me.While waiting for stock, I have been reading the [H], OCN, and Asus threads multiple times per day. If your monitor is not behaving, please identify how many and which model video cards, driver version, number of connected monitors, gpu overclock, cable manufacturer, cable length, and software settings. We might as well hunt for a common cause to some of the reported issues.
Would also be nice if people with no problems did the same thing but I suspect they are busy playing games.
Not just turned on, but repeatedly cycling through some utility designed to "unstick" pixels or subpixels.
But, you're correct - it should work as advertised, regardless of the specific use case. So, that shouldn't be a concern.
I'll suggest that the CPU analogy that someone posted is a bit off - yes, I have my processor overclocked at all times, but it's not pinned at 100% usage 24/7.
Ah, that's different. If they're pinning the overdrive to maximum and running one of those pixel un-stickers, even if they're technically within spec, it is sort of asking for trouble. I've never thought about this before but I speculate the driver current demands are higher and spikier when you're trying to change every single pixel from max one color to max another at a high rate.
1. You don't even understand how unsticking utilities work. The only area getting worked is about 3 5x5 grids. Some of which overlap. That's <75 pixels out of the entire screen.
2. No, the only way my analogy is remotely off is that I am in no way doing what I am not supposed to be with this monitor, whereas when overclocking a CPU you're definitely doing something outside of the standard use cases. Overclocking isn't covered by warranty. Playing around with stock monitor settings is well within spec. Furthermore, CPU's are (or should be) designed such that they can withstand 100% usage at their stock speeds 24/7. Otherwise, they're faulty. This is the same with the monitor. Whether I'm flashing every pixel on the screen a billion times to unstick one pixel (which is stupid all hell and I am not), or playing a game, it should display the image nonetheless.
If you have a device that has time-based use/load restrictions, those need to be explicitly printed on the package, because it's not the same product.
Furthermore, the analogy is doubly true in that I am stability testing ASUS' own overclock after a rather catastrophic failure of the first sample... supposing I was pixel unsticking the entire screen anyway (which I am not...). What do you do to test overclocks? 100% intensive usage for several hours. How am I (hypothetically, because I am not..) doing anything different? I believe I stated the time duration. Mind you that wasn't even as intensive as it could have been, since the area I was flexing was about 75 pixels in size. In theory while gaming you're not actually doing different from the worst case usage scenario you mentioned anyway. Suppose you're panning the camera like crazy in an FPS at 144Hz. What happens to each individual pixel? IT'S CHANGING COLORS LIKE A MADMAN.
So...
Not that I'm actually doing that as I mentioned above, but you realize that could actually happen in a game either way right? I mean I'll have you know that my problems didn't happen during my attempted pixel unsticking. They happened during completely normal play.
Do you also have those faint, thick white bars about 25% and 75% down the screen actually present on yours when you're viewing that white image in person? I see them on your image and those aren't normal.
Question:
I got the rog swift in today. I have a ati radeon 6990.
Have it setup, but a few issues
1: The only refresh rates I can set are 60, 85 and 100. Thats all windows shows as a option and pressing the monitors button to change has no effect. It was my understanding even the ATI cards would be able to utilize the 144hz...?
2: I cant enable ULMB- Seems to be greyed out- Is this a nvidia only option?
Any ideas?
Whenever I use ULMB the textures seem to be a bit grainy. Is that normal?
but this is a huge bummer.
I'm not sure what you mean?all these doltz that buy these for use with an AMD card...wtf people? Go buy a Benq.
Because this is the only 1440p monitor that runs at 144Hz. All the BenQ monitors are 1080p.all these doltz that buy these for use with an AMD card...wtf people? Go buy a Benq.
This is the only 2560 monitor that offers 120hz+, so I'm not sure who the real dolt is here.. I'm not wasting my computer's power at 1080p.all these doltz that buy these for use with an AMD card...wtf people? Go buy a Benq.
I was giving an explanation as to why AMD owners would want to purchase this monitor. At no point did ASUS ever say 144Hz would not work with AMD cards. In fact, in their Q&A they say themselves that the monitor would act like a "normal" 144Hz monitor on AMD hardware.I believe he means Gsync only works with Nvidia and others reported it won't run at 144 Hz with an AMD card.
http://rog.asus.com/339252014/gaming-monitors/faq-rog-swift-pg278q-27-inch-g-sync-gaming-monitor/FAQ: ROG Swift PG278Q 27-inch G-Sync Gaming Monitor said:What happens if I plug the PG278Q into a non-Nvidia GPU?
It will act as a fast gaming monitor* supporting WQHD resolution up to 144Hz with 1ms (GTG) response time just without the G-Sync ability.
*A graphics card with DP 1.2 support are required to select refresh rates above 60Hz.
I believe he means Gsync only works with Nvidia and others reported it won't run at 144 Hz with an AMD card.
Post your card(s), cable type/length, driver, monitor settings, game settings. Asus said they are aware of "out of range" and is working with Nvidia for a driver fix.