Stutters since 1.6.7 update

Hi Charles,
Thanks very much again for your tips.
Unfortunately I don‘t have someone who can I test a similar graphics card and ram from. My Fps were already capped at 144.
Also I switch the game to full screen mode and installed the AMD Adrenalin software to test the Freesync and the Enhanced Syc options, but the issue persist. Not the whole time but still there.
So I don‘t have more things to try…
Best regards,
Carlos

 

I brought up Freesync as you have an AMD GPU and never mentioned the specific model of your monitor. The overwhelming majority of displays on the market are Freesync displays (with G-sync displays needing additional hardware modules to operate within spec). The VESA VRR spec is more or less the same thing as Freesync (and even NVIDIA often brands VESA VRR compatiblity as G-sync Compatible). I questioned the lack of driver software previously as I am unsure of how else a user would confirm proper operation of Freesync/G-sync/VRR without having driver software to confirm it. On my own display (LG CX OLED 48), simply enabling Freesync/VRR in the display menu is not enough to actually engage VRR, requiring me to enable variable sync within the NV control panel.

When using an AMD GPU, the Adrenaline control panel will have a similar settings screen for enabling Freesync.


Using in-game v-sync is not known to be a best practice for VRR and could possibly interrupt or prevent VRR from operating normally (which can cause microstutter). Each game implements v-sync in its own way and lots of games don't even make it clear to the end user which type of v-sync is being used (ie double or triple buffered). Sometimes, using the in-game vysnc can prevent VRR from working at all. Additionally, setting your driver framerate cap to something a few Hz lower than your panel's native refresh rate is recommended as it will prevent games from exceeding the variable sync spec window, which will cause its own stutters (this also applies to the lower range of a display's variable refresh rate range). If your display is 144Hz, then setting the driver cap around 140Hz makes sense to keep you in the proper VRR range. Again, many games have their own caps, but implementations can vary wildly so simply using a global cap is always recommended.

 

Failing all other troubleshooting steps, running the game with a performance overlay can help to pinpoint issues. Using such a monitor can provide you with a realtime frametime graph that will show hard evidence of each blip in frame delivery.

Its possible to watch the graph in the performance monitor and see if it aligns with dips in GPU utilization or CPU core usage. Sometimes the stutters have easily identifiable patterns that align with the hardware reporting if you are lucky. If the frametime graph is mostly solid but you still see stuttering, then you know the stutter is not rendering-related.

This was how I determined that AI pitcrews were wrecking performance on my own PC. Each time I circled near the pit straight when AI cars were in their window, I could observe severe imbalances in frame delivery and could confirm if on the frametime graph with certainty. Despite an average fps of 117Hz, the pitcrews caused short spikes to nearly 0 fps in between frames that were otherwise arriving ~8ms apart. These stutters weren't severe enough to alter my average fps readout, but the spikes on the graph were the proof in the pudding, so to speak. Before this, I knew my performance was wrong, but I couldn't pinpoint it exactly without help from the graph.

 

Hi Carlos,
I was thinking.... I built a PC for my son, last year because he had an old PC + a 32' 165HZ screen (freesync one). He plays Fortine, and had micro stutters bothering him. So again....here we are making tests to get rid off this micro stutt.
Computers science can be weird and unexpected . The solution i found to make micro stutt disapear, was to put the graphic whole chain (from windows 60hz to drivers and game) at 60 FPS, despite the screen was 165hz max, and it worked. (the screen reacted very well using lower hz)
Also, we tested different modes of freesync on screen settings, and I remember that it was not using freeesync option on screen and only v sync in game the result was perfect.
Not having a large amount of FPS was disturbing him, but he admitted than the sensation of fluidity was best at 60fps even if less reactive (for competitive FPS its not ideal).

Now he has a new ati graphic card and and no micro stutt (it very early in France, so i won't go in his room turnig on is PC, but i'll go take a look at his settings later)

All this for finally saying that despite it is counterintuitive, it cost you nothing to make this test...who knows.

Charles

 

Thank you Charles, I'll try this test and will tell you how it's going.
Best,
Carlos