Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[zacharo]

Hi,

I would like to ask if anyone experiences differences between FPS when watching above ground and underground parts of fortress.When I reach a certain point of the game (lets say 40 dwarfes, embark 2x3), I am having a huge FPS drops while watching above ground levels: while managing underground levels, I am having solid 100 FPS, but immediately when I zoom to first above ground level (F1), FPS can drop to 60 or less and it is worse with bigger fort. (And always get back to 100 when I zoom back below).

I read wiki a discussions here or on Reddit and nowhere found anyone mentioning, that the different z-levels can have different FPS. I am aware of problems with growing fort (FPS death, pathfinding, items count etc), but I dont understand how can be the FPS afflicted switching the view of different Z-levels. Shouldn't the game compute these things in the same way, no matter what Z-level I actually watch? Am I doing something wrong?

I played game on dual i5 and quad i7, on Arch linux and Windows 7. Tried to set different print mode or other settings like weather and stuff, with no success.

[Pirate Santa]

If you're using the TWBT multi layer view that will most definitely affect your FPS.
(If you can see multiple layers with a slight fog over the lower ones multi layer is on)

[Libash_Thunderhead]

Even if you don't use twbt, it is still slow watchcing a plant infested level (usually the ground level).

[SimRobert2001]

Yes, i do as well.

[Loci]

immediately when I zoom to first above ground level (F1), FPS can drop to 60 or less and it is worse with bigger fort. (And always get back to 100 when I zoom back below).

Trees. The code that displays trees is... highly inefficient. Chop down the trees (or embark somewhere without trees) to avoid the tree-display lag. Note that an embark with "sparse" trees will get slower and slower as it turns into a lush, FPS-choking jungle.

[Dunamisdeos]

Never knew that. I'll have to pay attention next time.

[PatrikLundell]

I think you should pave over (most of) the surface to avoid the junglification. Strip mining would achieve that as well, but will apparently cause trouble with the path finding code. A roof at 1 z level will block trees from maturing, I think, and you could embark where there are no trees at all in most of the embark tiles. Surface clay seems to block tree growth as well.

[KingKaol]

Trees. The code that displays trees is... highly inefficient. Chop down the trees (or embark somewhere without trees) to avoid the tree-display lag. Note that an embark with "sparse" trees will get slower and slower as it turns into a lush, FPS-choking jungle.

I get lower FPS even when viewing above ground biomes without trees (tundra).

[PatrikLundell]

If TWBT is used the depth fade effect will still soak FPS. Also, trees are the worst offenders, but all kinds of vegetation can have an effect. In a recent fortress I had a significant FPS dip every autumn as the local grasses produced seeds (I assume. their display was changed, anyway). The embark had a single tree originally, and the effect was noticeable even with that all trees cut down.

[zacharo]

Well, i decided to create a new fortress in a Rocky waste (fireclay ground, no trees or other vegetation) and the improvements wasnt very noticable, FPS kept dropping below 60 quite soon when watching the above ground.

Then I dumped that fort and created new one in moderate vegetation location (some trees, some plants and other greenery) and decided to try a new thing -- I didnt build anything except roads and floors on the above groud (total flat) level, because previously i've been building towers, castles and other stuff (mainly to avoid sun sickness and for above ground crops). I believe, that this helped quite a lot in general (underground FPS can be even 90 for a quite long period of time, above ground drops below 60 only in siege or wood gathering. And i dont think i changed anything else, have same fort design, digged the same amount of rock.

Still the difference between underground and above ground thing is really annoying and 30 FPS is quite a big deal for me personally. While DF is awesome and I wish the authors the best, I really dont understand the development decisions. If I understood correctly, there are going to be some new features like e.g. taverns and libraries, which is cool, but what good it is, if the game gets unplayable after a few years of running a fort and adding new stuff surely wont help with this. And this problem with viewing  the different parts of the game area looks more like problem with really inefficient code (rendering?), which should't be that hard to fix (and the impact would be great I believe). While the performance issues related to game complexity and scaling might not be solved even with paralel computation or better hardware, little consolidation of the code shouldnt be an unattainable goal.

[DeDeRon]

I really dont understand the development decisions.

easy: adding features is fun, debugging isn't.

it makes me sad to see so much effort wasted - fighting fps death is more challenging than the game itself. instead of exploring the game you have to come up with strategies to *avoid* parts of the game. and you are right - it will get worse with every release.