Blur Busters Forums

Softhe wrote: ↑

28 Mar 2022, 11:52

Figured out how to force display scaling here's how my cru looks for anyone want to do it https://i.imgur.com/SoUwyoc.png

I'm sticking to Nvidia Control Panel because I'm failing to change timings even when deleting every resolution (Standard, Established, Detailed), deleting every single hardware block, sticking to DisplayID 1.3 and 2.0 and putting them on top to force priority, I hope that ToastyX will fix this issue in the near future.

I'm currently using a custom 1920x1080p resolution with 1200VT and 2000HT, I'm wondering if I could get different results with CRU...

lyrill wrote: ↑

28 Mar 2022, 14:02

Softhe wrote: ↑

28 Mar 2022, 11:52

why does your hz all dip below 390.00? and why do they vary a bit too among diff res

I play cs only and in that game i get 500-600fps

Softhe wrote: ↑

28 Mar 2022, 17:37

lyrill wrote: ↑

28 Mar 2022, 14:02

Softhe wrote: ↑

28 Mar 2022, 11:52

why does your hz all dip below 390.00? and why do they vary a bit too among diff res

I play cs only and in that game i get 500-600fps

ok. but my question remains unanswered.

lyrill wrote: ↑

28 Mar 2022, 21:55

Softhe wrote: ↑

28 Mar 2022, 17:37

lyrill wrote: ↑

28 Mar 2022, 14:02

Softhe wrote: ↑

28 Mar 2022, 11:52

why does your hz all dip below 390.00? and why do they vary a bit too among diff res

I play cs only and in that game i get 500-600fps

ok. but my question remains unanswered.

Becuse i copy the 240hz profile timings

axaro1 wrote: ↑

28 Mar 2022, 16:27

I'm sticking to Nvidia Control Panel because I'm failing to change timings even when deleting every resolution (Standard, Established, Detailed), deleting every single hardware block, sticking to DisplayID 1.3 and 2.0 and putting them on top to force priority, I hope that ToastyX will fix this issue in the near future.

I'm currently using a custom 1920x1080p resolution with 1200VT and 2000HT, I'm wondering if I could get different results with CRU...

Follow up to this post, inspired by a pro (Overwatch EU contender) who is using this monitor at 240hz I tried 240hz but with 1950VT (2000 is out of range) with 2000 HT. Crosstalk is pretty much non existent except for top 2 rows (where there is some minor crosstalk)



240hz with the same pixel clock as 390hz makes 240hz also lagless (due to QFT), I guess that I'll stick to these settings until I'll get an upgrade over my Ryzen 5 3600.

• Edit 1: 936Mhz is Pixel Clock limit, the monitor seems highly flexible in terms of VT control, I'll also try and test games where I can actually output 390fps (such as splitgate) to see if I'll prefer 240hz over 390hz even in ideal scenarios. So far 240hz with 1950VT feels great in Overwatch (but I also can always output 250+ fps consistently, which isn't the case for Apex), I need to test some more.
  
  If I compare perfect scenarios such as 3510pps(390*9) Smoothfrog with 390hz VRB Extreme against 3600pps with 240hz (240*15) VRB Extreme then 390hz strobing looks almost the same if not slightly better, same for unstrobed performance where 390hz seems to have an edge in clarity.

• Edit 2: I tried capping 190fps with 240hz which is the fps drops range that I get in Apex and strobing induced double images are not a lot better compared to getting the same fps but during 390hz strobing, I might have to create a custom 300hz resolution for Apex the day I'll manage to get 300fps consistently (due to the engine hard cap at 300fps).
  
  I'm getting the impression that strobing induced double images are quite a lot worse than crosstalk artifacts, they just destroy motion clarity...
  
  Crosstalk is a minor issue compared to that, 390hz with 390fps in Splitgate still feels better than crosstalk-less 240hz, both in terms of clarity and smoothness.

Tomorrow I'll test and compare scores 390hz strobing vs 240hz crosstalk-less strobing in Kovaak, numbers don't lie, let's see what I play better with uncapped fps.

• Edit 3: My scores are better with 390hz strobing at 390fps, especially with flick based scenarios and tile frenzy mini, 1w6ts I had the exact same peak with 390 vs 240hz.
  
  So what I tried to do was lower fps with 390hz to see the point where strobing induced double images starts to mess with clarity, from my results (tested with ~50fps decrements). 300 to 390fps I play better with 390hz, at sub 300, especially when getting 200 and 250fps I match 240hz 240fps scores (although the latter looks clearer). Sub 200 fps clarity is pretty much fcked with 390hz strobing, same with 240hz strobing but to a lesser degree

Softhe wrote: ↑

25 Mar 2022, 15:01

Here are some CSGO settings i think are great
Overdrive Extreme
VRB Extreme
Both of these are fine at normal also

color settings;
r gain; 55
g gain; 50
b gain; 50

r bias; 51
g bias; 49
b bias; 49

r saturate; 100
g saturate; 80
b saturate; 80
y saturate; 55
m saturate; 50
c saturate; 80

Whenever I open CS:GO my VRB gets set to 'Off' do you experience this issue too?

disregard - :p

Someone asked for this in pm so Im posting it here for anyone who's trying to force display scaling in CRU here's a video.
https://youtu.be/IZEM6Qhw0N0

Softhe wrote: ↑

31 Mar 2022, 14:23

For it to work for me I had to remove the "Default extension block" too. I imagined that was where the displayport resolutions were hiding but it seems I might've been incorrect.

However I've noticed that when I wipe the EDID and add my own resolutions in CRU it seems to break the overdrive for me at 360hz and 390hz(these were the only 2 I tried so I couldn't tell you the cutoff, at 240hz the overdrive setting still works).

Breaks as in the OSD toggle does nothing and randomly toggles to "OFF", which makes 390hz pretty much unusable. Switching to a lower refreshrate to toggle the overdrive and then going back to 390hz will only work for a short while, not even minutes in some cases.

I was wondering if this is also the case for you, or if it's a quirk with my monitor/CRU profile.

axaro1 wrote: ↑

29 Mar 2022, 16:13

I tested 1950VT 240hz and I can't say that it looked as good to me as what you're describing, using both NVCP and CRU.

However I also tried 1956VT for 240hz and, to me, it looked significantly worse than 1950 VT which surprised me.

axaro1 wrote: ↑

29 Mar 2022, 16:13

936Mhz is Pixel Clock limit

I can push the Pixel Clock to 975Mhz as evidenced by my screenshot from earlier in the thread. The limiting value for my unit when pushing the VT seems to be the Horizontal Scanrate at 469.380 kHz, anything higher like for example 469.381 kHz results in "No Signal". I'm unsure which one of these values are important(more important?) for QFT as the QFT guide just tells you to check either radio-button.

I'm not sure if it's possible to work with the same granularity when using NVCP, in CRU I just increased the Horizontal Scanrate which NVCP doesn't let you do. But atleast NVCP doesn't seem to break the overdrive.

I would be interested in seeing some crosstalk examples if you've got the time/energy/drive to shoot and post em!

tlp wrote: ↑

01 Apr 2022, 01:15

I can push the Pixel Clock to 975Mhz as evidenced by my screenshot from earlier in the thread. The limiting value for my unit when pushing the VT seems to be the Horizontal Scanrate at 469.380 kHz, anything higher like for example 469.381 kHz results in "No Signal". I'm unsure which one of these values are important(more important?) for QFT as the QFT guide just tells you to check either radio-button.

I'm not sure if it's possible to work with the same granularity when using NVCP, in CRU I just increased the Horizontal Scanrate which NVCP doesn't let you do. But atleast NVCP doesn't seem to break the overdrive.

I would be interested in seeing some crosstalk examples if you've got the time/energy/drive to shoot and post em!

NVCP timing approach surely is different to what CRU uses.

I didn't find a limit to Pixel Clock when tweaking HT, I just tested my previous 390hz timings with 1200VT and 2000HT but replaced 2000HT with 3000HT approaching 1400Mhz pixel clock and it didn't go out of range.

My Pixel Clock limit is based on 2000HT with Higher VT Achievable -3VT(to fix possible tearline issues)

tlp wrote: ↑

01 Apr 2022, 01:15

I tested 1950VT 240hz and I can't say that it looked as good to me as what you're describing, using both NVCP and CRU.

However I also tried 1956VT for 240hz and, to me, it looked significantly worse than 1950 VT which surprised me.

Anyway, here's the Crosstalk test at 240hz with 1950VT: