Bucic

Clearing cache is almost always just another variable out of the way, not a sure remedy. I'm still on a holiday trip, so can't weigh in myself directly.

https://www.thewindowsclub.com/how-to-delete-directx-shader-cache-in-windows

And all this with forests not looking that good AT ALL. #bringbacktreeclusters As you can see daily from footage recorded in Ukraine tree clusters are often not traversable even for infantry units.

Feel free to include your modless track along with specs and DCS.log.

Already reported, both in F-5E and general DCS performance bugs. https://forum.dcs.world/topic/366466-f-5-old-not-remaster-performance-drop-in-the-latest-patch/

I plea to have this bugs triaged together. In specific conditions a single object in view sinks the entire game performance. trees - December 4th update, or maybe earlier as it was difficult to notice F-5E - December 24th update

The usefulness of tracks is obvious. But, for my own educational purposes, could you please confirm whether they also contain performance metrics of the recording machine? Like those we see in the extended metrics that are built into DCS?

Beautifully done!

It depends. Load Instant Action - Caucasus - Takr off mission. The AI there are definitelly remastered models, evident by their cannon bay covers open after shutdown.

Thanks for the effort! Do you confirm that subsequent sessions are free of the performance hit? Tracks are always good but I don't think they contain any original session performance metrics. EDIT: Guys, try clearing: Saved Games/fxo Saved games/metashaders2 Windows shader cache ... then running at least two sessions 5 minutes or longer.

Is this in line with your experience?

Please post a bug report with a mod less track if you can. I won't be able to update for another week. The sooner it gets reported, the better. Also please take a look at

Call Heatblur...

Cockpit shaking, LERX, wing tip vibrations... Continued 'aliving' of the F-5E experience. Merry Christmas, ED Team!

"The fans are thirsty. Either we go or you go!"

Just a word of advice. Even if the problems like those have been acknowledged but change somewhat with time it's worth noting that in the respective issue forum topics, with a fresh modless track and maybe even a fresh dxdiag report.

I don't think it's more serious than this https://forum.dcs.world/topic/366205-performance-trees-huge-performance-hit-especially-up-close/

Got the culprit yellow background but I still don't know where is the content (laser codes etc.) programmed with that huge red/blue font.

I've got a confirmation the bug has been confirmed for the current build/Remaster so I guess we just sit and wait now. I highly recommend the new Caucasus Free Flight instant action mission which lets you fly at 2 in the night, pitch dark with some stars. On that note, I'm going to check how such conditions look with and without my bloom mod.

So that would confirm the bug. I managed myself to trigger it even with LOD 0.3. Like I said before, as soon as "non-minimal" LOD is in view, the performance dives dramatically, with sawtooth frametimes. It's worth realising that as soon as you hit sawtooth frametime on a graph, FPS don't matter anymore. That characteristic is indicative of a serious performance problem where a game engine will not and cannot deliver any satisfactory end user experience as it turns itself into a heat virus, akin to Prime95 "benchmark".

Yeah, I cranked gamma up in my recent night flight video and I can see that as well. They just turn of the bulbs. BTW, what are those two blinding lights behind each pilot's ear? They seem to turn on and off by nav light knob as well.

Just checked it on my end. I was wrong. This IS how F-5E SAS works, at least in the sim The main purpose of simple implementations of such systems is to quickly dampen yaw oscillations. Apparently does that plus has an anticipateur implemented. BTW, here's how it looks when it's trully needed but turned off (both channels)