Blur Busters Forums

Bishi wrote:I've always been sensitive to tearing but now even more so, with blur reduction I'm also noticing stutters more and I find that with this enabled I prefer the gaming experience with vsync on. Frame capping below hz helps a little with the input lag although I suppose this would depend on the game used.

Make sure you cap your framerate via the game engine, rather than via video drivers (e.g. NVInspector). Capping the framerate in the game avoids the enforced latency of external frame rate throttling.

Bishi wrote:I've been experimenting to get rid of the cross talk toward the top of the screen by using different resolutions and hz combinations with total pixels but haven't found a good one just yet. 1350 @ 1080p 120hz/128hz seems to be as good as any other. I use 2.5ms persistence.

Strobe crosstalk goes down the lower the refresh rate you go. The "height" of the crosstalk zone is half height at 72Hz as it is at 144Hz, meaning it fits more easily in the offscreen space. The best thing you can do is try a 100Hz or 120Hz rate. Although lower refresh rate strobing can have slightly more noticeable latency.

Bishi wrote:The crosstalk can be quite noticeable on some BF4 maps as they can be very bright compared to most FPS games, a great example of this is flying a chopper on one of the island maps or the test range.

This crosstalk pretty much disappears at ~100Hz/VT1500. That said, 100Hz might feel slightly more laggy than 128Hz, but only by a millisecond or two.

P.S. I've seen certain LightBoost monitors (e.g. ACER HN274HBbmiiid and the early XL2420T 120Hz monitors not capable of 144Hz) that had much worse strobe crosstalk than BENQ Blur Reduction V2.

I find 1080P @ 100Hz + 1502VT to be the sweet spot because even with a SLI of overclocked GTX670 on an overclocked i7 4770K, it's impossible to maintain a steady 120fps in a game like BF4 so rather than fluxtuate between 100-120 I'd rather pretty much always have my 100fps with the game capped at 99 like Chief told me so I get ride of most tearing without using Vsync.

Yes, 100Hz VSYNC ON + game-engine framerate capping at 99fps = fairly low latency VSYNC ON.

Battlefield 4 has so much engine latency, that the low-latency VSYNC ON technique (VSYNC ON + framerate cap below VSYNC ON) above only adds a mere few percent (only half a refresh cycle penalty!) to the whole input lag chain compared to VSYNC ON. If you're not aiming to win tournament money at competitions, this is a great sweet spot for motion image quality during strobing during BF4. The blurfree/tearfree/nearly stutterfree experience actually may improve gameplay scores of average/casual players (testimonials), and BENQ Blur Reduction V2 allows brighter strobing than LightBoost.

Chief Blur Buster wrote:Yes, 100Hz VSYNC ON + game-engine framerate capping at 99fps = fairly low latency VSYNC ON.

Battlefield 4 has so much engine latency, that the low-latency VSYNC ON technique (VSYNC ON + framerate cap below VSYNC ON) above only adds a mere few percent (only half a refresh cycle penalty!) to the whole input lag chain compared to VSYNC ON.

I thought the trick of capping the frame rate 1 fps under the refresh rate was only when you didn't want to use VSYNC ON but you say that it also helps even with vsync so that we should ALWAYS be doing it wether we use vsync or not?

A last question, what do YOU personnaly think about the use of (Adaptive Vsync)? It seemed like a godsend feature when it first appeared few years ago as the way I'd see it is that you'd get the best of both worlds. All the advantage of Vsync On when you could handle it but no frame halfing if you couldn't maintain it so wouldn't it be even better to always use Adaptive as even with an extremly good computer, it does happens sometime to drop a few frames under 100fps in BF4 as it's very demanding and I wouldn't want to drop from 100 to 50 from time to time

PanzerIV wrote:I thought the trick of capping the frame rate 1 fps under the refresh rate was only when you didn't want to use VSYNC ON but you say that it also helps even with vsync so that we should ALWAYS be doing it wether we use vsync or not?

There are benefits even for VSYNC ON.
The visual/latency behaviors are different, but there is a noticeable and appreciable latency reduction in this situation:
(1) You use VSYNC ON and then cap the framerate slightly underneath (even fractionally; e.g. fps_max 99.5);
(2) You use a GPU that can run at full throttle against the frame rate cap;

What happens is that as soon as the frame finishes rendering, it's almost immediately displayed without being piled into latency-inducing chains of buffers. The game engine is throttling the output of frames, so the chain of buffers (adds latency) never fills up because the frame rate is always lower than the refresh rate.

In fact, frame-rate capping has larger latency improvements during VSYNC ON than during VSYNC OFF. There was no latency improvements in certain games such as CS:GO (e.g. fps_max 300 versus fps_max 143 had the same non-GSYNC latency during VSYNC OFF). The chief benefit of the frame rate capping is better visuals (smoother motion) during VSYNC OFF.

PanzerIV wrote:A last question, what do YOU personnaly think about the use of (Adaptive Vsync)? It seemed like a godsend feature when it first appeared few years ago as the way I'd see it is that you'd get the best of both worlds.

It is a very good feature and a favourite among some LightBoost users. It may actually be better in BF4 than the framerate capping method, but I haven't tested how much latency adaptive VSYNC adds. Provided it simply "steers" the tearline off the edge of the screen (without dynamically adding any buffers to the buffer chain), adaptive VSYNC should be competitive in latency to the VSYNC ON + framerate cap method. Software can adjust where the tearline occurs and even steer the tearline offscreen, via precision timing (using knowledge of the horizontal scanrate, in number of scanlines per second). Forum member Aahigh (game developer) used an early form of adaptive VSYNC almost twenty years ago in the San Francisco Rush 2049 coin-op arcade machine.

However, BF4 adds so much latency itself, it is even hard to feel ordinary VSYNC ON latency. It's easier to feel in a game like Quake Live or CS:GO.

You keep mentioning zero crosstalk at 100hz and less but it is still very noticeable to me, although far less at 0.5ms persistence than 2.5ms, always for the top few rows on full screen motion test. Shame we can't set the strobe phase negatively and push it off the top

So for me at least there is very little difference in crosstalk between 80hz and 128hz so I'm sticking with 120hz for now @ 2.5ms persistence.

Almost makes me think im doing something wrong when i'm still seeing this crosstalk. Instant mode on, AMA high?

Hi! I just sold my 2xGTX670 to upgrade for a 2xR9 280X at 0$ which is why I've done it. I just remembered yesterday that I didn't have anymore my custom resolution with a VT of 1502 and HT of 2080. It prolly explains why BF4 felt so dark, definitly more than before even though I kept the same BlurBuster's settings in the utility at 1.5ms

I've installed the (AMD's Pixel Clock Patcher) to remove the 330Mhz limit as it's only when you use the 120Hz with custom timings that it goes a little above 330, otherwise you'd be fine at 100Hz without patching. You don't have to do it everytime you reinstall a new videocard driver right?

Then I downloaded (Custom Resolution Utility v1.1.2) and created a detailed "custom" resolution and rebooted. However how the hell do I know if I'm using the custom resolution I've set?! O_o It was more simple with Nvidia. With Toasty, does it change the values inside Windows so that everytime I hop from 1920x1080 @ 100Hz to 120Hz it will automaticly use either my custom 100Hz or custom 120Hz res???
http://uppix.net/aeyTkF.jpg

PanzerIV wrote:It prolly explains why BF4 felt so dark, definitly more than before even though I kept the same BlurBuster's settings in the utility at 1.5ms

I've read an article which mentions that persistence varies with refresh rate when default timings are used. So those 1.5ms in the utility were actually less ms.

PanzerIV wrote:You don't have to do it everytime you reinstall a new videocard driver right?

Yes, you have . Every time I uninstall the Nvidia drivers and install the new ones, I have to patch the clock limit again (and restore custom resolutions, and game profiles...).

PanzerIV wrote:With Toasty, does it change the values inside Windows so that everytime I hop from 1920x1080 @ 100Hz to 120Hz it will automaticly use either my custom 100Hz or custom 120Hz res???

There is an easy way to know if you're using custom timings: check what refresh rate your monitor reports. In my case, it reports 60Hz with every custom resolution. You can also check how the crosstalk is.

masterotaku wrote:Every time I uninstall the Nvidia drivers and install the new ones, I have to patch the clock limit again (and restore custom resolutions, and game profiles...).

There is an easy way to know if you're using custom timings: check what refresh rate your monitor reports. In my case, it reports 60Hz with every custom resolution. You can also check how the crosstalk is.

I can understand that you have to patch again the clock limit and restore custom resolutions if you uninstalled the driver, mostly if you also used a driving cleaning utility such as the nice one from Guru3D but what if you simply install the new driver OVER the current one without uninstalling first? Of course if I switch from Nvidia to AMD I do a total cleaning but when I was only installing a newer driver for a same card, I wasn't going over all this process simply for a small update if I had no weird behavior etc. Never had problems so far by just overwritting drivers. It gets annoying having to set again everything in the CCC or Nvidia's Panel x_x

Hmm if you say it's "simple" with your XL2411 I guess it should be also with my XL2420Z but I'm not sure if I understood correctly your "simple way" to check it. Where exactly you want me to look for the reported refresh rate. Inside the normal OSD, Windows or the monitor's service menu?? As for checking how the crosstalk is, it's somewhat harder. You meant by using the UFO's test?

PanzerIV wrote:what if you simply install the new driver OVER the current one without uninstalling first?

I don't know, because I always uninstall the old drivers before installling the new ones. A friend of mine installs them over the previous drivers, and custom resolutions remain. He has a Nvidia GPU, so I don't know if that also happens with AMD.

PanzerIV wrote:Where exactly you want me to look for the reported refresh rate. Inside the normal OSD, Windows or the monitor's service menu??

In the monitor OSD, at the right of "Current resolution". It's something like this (quoting myself from an old post):

Back then, in V1, it reported refresh rates a bit randomly, but since the upgrade, every resolution with custom timings reports 60Hz.

PanzerIV wrote:As for checking how the crosstalk is, it's somewhat harder. You meant by using the UFO's test?

Yes, that's an easy way.