- FIXED - Related to DH Config Bug -> Huge FPS / Performance loss in 2.2.0-a vs older DH versions
All of the 3 tests are running on MC1.21 in the same MC Instance, same World and same DH config. All screenshots are taken after the DH tasks in the F3 menu have stabilized.
First picture I have DH 2.1.3-a-dev:
- Best performance so far
- DH Tasks stabilized in about a minute after loading the world
Second picture I have DH 2.2.0-a:
- Deteriorating FPS
- DH Tasks stabilized in about 3 minutes after loading the world
Third picture I have DH 2.2.1-a-dev (latest nightly build):
- Worst performance and FPS out of all of them
- DH Tasks stabilized in about a minute after loading the worlds
There no visible or noticeable visual difference between the rendering of DH 2.1.3, 2.2.0 and 2.2.1, yet the FPS drops are huge. I am terribly concerned about this, especially in low end devices like mine 15% GPU usage makes the difference between a stable 60fps and a stuttering 45fps. If you need any further information please let me know, I'm willing to help solve this issue as it is a deal breaker for me and other friends in low end devices.
27 Replies
2.1.3 and 2.2.1 are nightly builds, they change constantly. therefore it would be hard to conclude how performance was changed, without specifically knowing from what commit 2.1.3 and 2.2.1 are
and, are you sure you tested correctly? not a lot has changed between 2.2.0 and the latest version of 2.2.1. nothing that should noticibly affect performance anyway
The 2.2.1-a-dev is the latest one so far, I just downloaded it from https://gitlab.com/jeseibel/distant-horizons/-/jobs/7712923387
However I don't remember when the 2.1.3-a-dev is from.
GitLab
build: [1.21.1] (#7712923387) · Jobs · James Seibel / Distant Horiz...
This is a mod that adds a Level Of Detail (LOD) system to Minecraft. This implementation renders simplified chunks outside of the normal render distance allowing for an...
what is the date modified of the 2.1.3 file?
Yes, I'm sure. I am aware that nightly builds are unstable. So that might be the reason for 2.2.0-a to be performing better than 2.2.1-a-dev
imma test as well, give me a moment
thats not an old version. not a lot has changed between that version of 2.1.3 and 2.2.0 just like between 2.2.0 and 2.2.1, so im suprised youre finding such a large performance difference
Yeah me too. Please let me know the results from your testing
the 2.1.3 version is not downloadable anymore. gitlab deleted it because its too old
so ill just test 2.2.0 and 2.2.1
dont upload jars please, and im not gonna download, then run a jar from a random person on the internet
Uh, ok sorry. no problem
i do not see a large difference
seed: -3492584006216528006
Uhhh that's annoying because it could mean it is related to my hardware which is really impossible to fix... But Let me try too. I'll generate that seed in 2.2.0-a. Then re-start the instance, load and take the screenshot in 2.2.0-a. Close the instance, swap 2.2.0 with 2.2.1, load and take the screenshot and be back in a few minutes.
alright
I'm so confused
If anything, things actually inverted
Let me explain:
I created a new 1.21.1 instance, generated the world, tested 2.2.0, tested 2.2.1, downgraded to 1.21, tested 2.1.3.
So I see two possible options:
1.- one of the mods in my main instance is causing the FPS drops, meaning it has better compatibility with some DH versions than others, which is weird
2.- DH changed the way lods are rendered and saved into the sqlite database making worlds generated with 2.2.0 better optimized for 2.2.0+ while old sqlite databesed generated with previous DH versions perform worse
i dont think the sqlite format changed in any way that could impact fps, so my guess would be option one
Ok, then I'l test mod by mod until I find which is causing the problem. I really want to get to the bottom of this
sounds good
I found the problem. Finally.
So apparently, the 'cave culling height' of my main instance was set to -400. Meaning that it was basically turned off. That's why in a new instance with 60 as default, the fps were the same in all DH versions. I'm 100% sure I didn't put -400. So it must have been a bug with the conversion of DH Config from DH 2.1.3 to DH2.2.0 or to some buggy nightly build. Who knows.
I suppose now I should delete or at least rename this bug post as it was related to some buggy nightly that changed the DH Config to some weird values rather than to actual performance flaws in DH 2.2.0
-400? What in the world
Glad you found it though
Yup
Thanks for your help, I wouldn't have been able to do it alone.
No problem!