Most likely, it’s not doing much well, much at all even, the only thing good about it is that it works, which is why it’s still in service (ignoring the fact that I had to clean off it’s dust because I can’t use my desktop for some time)
So it was it’s last chance to shine I suppose…
I recently made a post about the testing I did on the Low Performance Reference laptop I have, with some suggestions about a low performance preset.
If anyone is interested to help, if you just disable the Microbe Ripple Effect option and disable MSAA/Antialiasing in the Graphics settings, do you get a good experience?
Or if you have your own suggestions for what settings you use or disable when playing on low end hardware, it would be great if you posted them.
How much could we afford to disable in such a low performance mode? What is the least needed for a “coherent” game of Thrive?
No other graphical settings seemed to have much of an effect on my low performance reference machine, except for the resolution scaling. But I am not sure if we could have the preset change the resolution, as that seems like it might be a step too far for a preset and should be left to the user to change. But perhaps that is not the case, do anyone else think that reducing the resolution to, say, 75% in a low performance preset would be acceptable?
I suppose it would be acceptable as long as you can still discern pili and flagella well even when close to eachother.
I think many modern games already do that (low graphics preset reduces render scale below 100%).
But like I said in the meeting, depending on the hardware the FSR overhead might be some high that setting render scale to like 80 or 90% actually reduces performance due to the FSR processing. So depending on the hardware and render scale we might need to have the preset go as low as like 50-60% (which starts to look really bad without FSR 2) if the hardware is so weak that the FSR processing overhead is very significant. That still has the problem that we don’t have the scale of testing required to determine what kind of hardware is really impacted by trying to use FSR.
I also disabled the Anisotropic Filter, Background Particles, Distortion to Microbe Backgrounds, Chromatic Aberration, and the Bloom Effect, and applied the low resolution background blur. My FPS went from 5-10 to 15-50, with occasional drops during saving and occasional high microbe amounts.
So it seems to increase the performance fourfold, proving the viability of the solution.
Fantastic! Did you use resolution scaling setting as well? Also, do you think you will use these settings in the future or are they not providing the visual experience you want for those extra fps?
On my low end machine, disabling those effects did not change anything significantly, maybe a 1fps difference at most. Does it help your machine to disable those? If you had the time, could you enable those effects and see if it makes much of a difference?
Somehow, my organism looked reddish with all of the graphical settings off in complete darkness. Also, yeah, it does not make the game look as “pretty”.
It definitely affects my old Nvidia GeForce 840M GPU memory usage, which went up from 400 MiB to over 650 MiB. Also, my FPS went back down to 5-6 FPS consistently, no longer hitting the 35 to 50 FPS range.
Could you disable all of those settings, and then turn them on individually and see which one(s) cause the fps to drop? That could be very helpful.
AKA, only have one setting activated at a time to see which one is the performance hog.
If the impact of the setting disablement is not consistent among the devices, should all the major ones be disabled under the performant setting or should the player be let to choose which to disable?
I think that the low performance preset would disable all of the settings that have been shown to reduce performance on at least one system; if there are settings which do not seem to ever reduce performance then I would not suggest having it disabled by default on the preset.
Also, on my i5-4300u Surface Pro 3, it seems like all of the settings that I tried on the i5-2520m system seem to be applicable, with the major difference that the chromatic aberration effect should now be disabled to get back a good 10% or so fps at the beginning of the game.
What is the cutoff point below which it’s not worth to disable a feature?
I will do it later today. I could not do it yesterday because I was busy updating everything in my computer before Windows 10 reaches EoL.
Are you planning to switch to linux by the way?
I don’t know. It seems very complicated as someone who is not good with using technology.
The largest without a doubt is enabling MSAA. Anisotropic filtering might have an effect on FPS (maybe 2?). I did not see any noticeable effect from background particles. I think the reason Bloom Effect, Chromatic Aberration, Distortion to Microbe Backgrounds, and Microbe Ripple can affect FPS is that many of those effects apply to cells and/or chunks, depending on their density when encountered.
I wonder if switching the models for cells and/or chunks to something less detailed would help the performance…
So pixelated mode could be coming to the game?