I suspect it is the CPU usage. That would explain why you have the same problem when running at lower resolution, and, that the problem is worse close to gates. In the latest versions of Recharged I doubled the physics update frequency, which makes the physics more accurate and more responsive, but is also makes it more demanding on the CPU. It is strange that it runs better when you measure the GPU usage though. Let me know if you manage to figure out anything more.
I use an Intel i3-7100, released in 2017, 3.9GHz, 3MB L2 cache, dual core with HT (4 threads)
Over 95% of the time, all 4 threads are more than 70% idle. Only occasionally, one of the four threads drops to 40-50% idle.
After starting the program, the micro freezes (or frame drops?) during passing a gate occur only max. once per gate. Also, I have to pass the gate relatively fast. The issue is present in "forest" but I do not notice it in "dessert". There is no issue, when I pass the gate a second time.
It is possible that all this has to do with using an integrated graphics card (resides inside the CPU) or with some Intel driver in particular.
Edit: This post describes the situation during measuring the GPU utilization. Without measuring GPU utilization, forest is unplayable.
> Do you mean that you notice it during Timed race?
Yes, but only in the first round of the first timed race.
> Or, do you notice the micro freezes even when passing through a gate normally (Without times race turned on) ?
No.
PS: IMHO, this (micro freeze at the gates in the first round of the first race) is no big deal.