Glad to see more people are requesting this. Perhaps we’ll get some real attention from Epic regarding this one day.
perhaps if we keep this thread alive, like that other thread about the car physics
Yeah and the Foliage shading model they wanted nothing to do with for ages.
If you want Frostbite tesselation, take money in your hands and buy a Frostbite license
UE4 being freeware just can not deliver everything. Apart from this, they should really focus on realistic 3D waves and foam, like the Ocean System in Cryengine, not people who want to tesselate 8x8km by mouth feeding and then complain about performance…
UE4 is not freeware. you pay 5% of your gross revenue to them which means anyone using UE4 and Epic are partners the moment that person has any sort of revenue. if you’re a studio and get funded by a publisher, you already owe money to Epic.
but yeah, let’s all devaluate the requests from anyone else and throw in the feature we think they should really focus on.
here I go:
they should really focus on endless space worlds, like in No Man’s Sky, not people who want to make realistic water (which is doable with the current tools, of which there’s enough examples and community projects) by mouth feeding and then complain about performance…
you see this won’t lead us anywhere
we’re here to point the flaws of a supported feature. if you don’t agree at least don’t come here to mock others
btw you could also take money in your hands and buy a Cryengine license if you like the Ocean System in Cryengine. just saying
This is not about asking Epic to redirect their direction to where “we” want. It’s about asking them to not move on to the next feature while the current feature have a long way to go yet.
Almost every graphical feature in UE4 still has to either be optimized for better performance or has to be visually improved, but 99% of them will just remain as they are. That’s sad.
i was kidding, sorry. There is sunshine here today. On a rainy day I also would have misunderstood.
I like your request about endless procedural worlds
Will enter a request. I will do a joint request: endless procedural worlds with endless procedural 3D water. And I will add a topping: endless tesselation on the endless worlds.
Btw. you pay 5% of your gross revenue to them if you make more than $3000 per game and per calender quarter. Until then: freeware! FREEWARE! With the usual unfinished and unpolished features that come with…wait…wait…FREEWARE
But hands down, UE4 is actually working well, the vast majority of things works. The few bad eggs in the bucket just shine stronger. If tesselation is not working on 8x8km but on 8 times 1x1km, then just do that. Also always check in shipping built, not in PIE.
so yeah, if you’re a studio and you get a publisher to fund you then you already owe money to Epic.
despite you might perceive it to be the same, becoming partners with someone that allows you to use their tools is very different from using a Freeware product.
btw tessellation has an almost prohibitive cost if used on Landscape. even if the Tessellation multiplier is 0 which effectively gives you zero additional polygons. not just an 8x8km landscape, but any size.
for me it’s not even about Landscape (yet). using Tessellation (with displacement) on painted foliage that uses LODs makes me lose anything from 50% of my FPS to making my GPU crash (depending on the amount of foliage instances)
workarounds? please no. finding workarounds is nowadays widely accepted but doesn’t make it less of a bad practice.
workarounds will make some people happy because they can forget about the problem, but some will remain for which the workaround is not convenient or have a case where it just does not apply.
but as the workaround helps a good number of people, those that remain become a minority so it becomes less crucial which gives them the ok to ignore the issue.
all we’re asking is that they at least make the supported features work and perform properly.
If your game only makes $2999 per quarter, I think you still owe Epic nothing, no matter if you are funded by a publisher or not.
I do not think (or just have not seen it yet) that the $3000 rule changes just because you are backed during your development by a publisher.
So day-dreamingly speaking, if our game had been backed by Electronic Arts or Ubisoft, ya know , and they pumped $3.000.000 Dollar in our development, and then we totally sink the game and it only makes $2999, then Epic would not see a dime. We could then proudly say we used freeware!
If you have a publisher, you are probably better off getting a custom license anyway.
can we get back on topic?
I mentioned Frostbite just to show that is very possible to have performant tessellation on landscapes, in an effort to give examples to Epic in order to improve their current implementation which still works great on small meshes and non landscape actors. I love UE4 and it is only right to try and offer feedback on how it can be improved, and clearly many people in this thread also agree with this. I’m not complaining in any way, I simply don’t use tessellation on landscapes, however I did want to give this feedback, as it would be a really awesome improvement to their current implementation!
There is literally, and legally speaking also figuratively, no difference to using a freeware product like Blender.
Anyway, back on topic. I am slowly running out of ammo to keep the thread floating on top so that Epic gets aware of it.
Please keep this on topic, there is a legitimate concern here whether it affects you or not.
Arguing about something unrelated is not the way to get Epic to notice a thread. Why would they respond in the midst of some off-topic discussion? There is now so much useless info in this thread that the good info is hard to find. When I say off-topic I mean anything that doesn’t relate to this specific problem in UE4.
I’m not saying they won’t answer here, just saying that if you want a clear and concise answer the question needs to be the same.
@Maximum-Dev: Please try opening a feedback thread and/or AnsweHub bug report post for this with your images & data, you’re much more likely to get an official response there. Leave a link back to this page for reference though, it does have some good info. Hope you can get an answer!
Yes and please post the link to the Answer hub topic so we can all go there and support you on it.
Afaik it was already addressed on Answer hub several times btw but I suppose it won’t hurt to try again and again.
I’ve had to completely refuse from tessellation on landscapes. My performance drop was caused by dynamic shadows. As soon as i set scene light to dynamic, enable shadow casting,add far shadow cascade and simply enable tessellation in material parameters(leaving tessellation multiplier at zero), I get terrible performance drop in shadowcasting.
I’ve posted it on answer hub a while ago.
Please all just support the Answer Hub topic as posted by @Maximum-Dev above. We actually might get something done that way. Thank you.
Bumping this to say we could still use your support on the Landscape Tessellation issues Answer Hub topic.
Please also make use of the up vote system in place.
Thank you for your efforts!
So how many upvotes/posts do we really need?