It will delay for a week thou, Im sorry but something came up that will need urgent attention from me regarding all the activies being done, just for you guys to understand, in case anyone havent reading the Marketplace forums about copyright, check this: https://forums.unrealengine.com/unre…tected-content
The asset is fine, no issues with it, but I have other works at the Studio that need just to stop to be rethinked… main reason for the delay.
Version 1.2 is under test and I hope to release it before the HowTo’s video series in few days. This is what is planned for it:
The fourth item basically you can tell how the system will behave, creating cloudscape states, interpolating between them, telling how long the states and the transitions are. This enable a control similar of that offered by TrueSky, with advantages that the states are reusable, you can model the cloudscape behavior covering any timespan you want: from a minute to a year. Just stay tunned, that I will release a video prior submission.
Wow, that sounds like it’s going to be incredible!
Also fixing over last submission:
a) moon not projecting cloud shadows - found Jan 10th 2018;
b) layer 1 mask texture default value missing - found Jan 9th 2018;
c) not possible to change the Z axis for the blueprint BP_CloudscapeViewer - found Dec 27th 2017;
d) moon is transparent and you can see the stars through it - found Jan 16th 2018;
e) a workaround to make the clouds be rendered on top of other transparencies setting Render After DOF to false is not needed anymore.
On your roadmap I don’t see anything about the ability to add multiple suns/moons and/or planets orbits… I’d like to be able to add planets created with a planet generator … if you added support to add multiple orbits for those… like even having 2 suns and 3 moons and a couple of planets at different distances, some looking huge other smaller in the sky and moving at different speeds.
I see that ability to set how many hours a day is made of is set to happen in June 2018 … I hope it will get implemented sooner… I need some planets having days of more than 24 hours here on Earth for the game I am developing.
I didnt put the ability to put multiple suns/moons in the roadmap because it increases complexity a lot. I put it in the howto series because people want to know how, so I will guide the implementation in the video. The point on controlling the orbits individually is not that difficult, since the function to calculate the orbits are inside the BP_TheMoon and BP_TheSun, these blueprints can have different functions to calculate the orbits and be used setting the instances accordingly.
The complexity resides in the main raymarch material and the main blueprint, already expensives, where it will need to add more parameters to handle the additional directional lights and the ability to use the combination of each contributing light (like when two or more different suns are present in the sky) to the scattering effect at the clouds… it might also happen that too many suns may cause almost no shadowing at the cloud body reducing a lot the quality, since the overall beauty comes just for the shadowing part. The sky sphere material and its blueprint will also need to receive changes, thou quite easy.
As we can have a duplicate material just for this experiment, following the videos to add simultaneously another sun + moon is quite enough for the audience create their own material and be able to add more in the future using it as reference.
Overall, the roadmap is already filled with neat stuff and Id like to have a complete system before moving to specific implementations and in this case probably replicating more blueprints.
I am working against the clock to see if I could update the package on the Multiplayer support for the version 1.2 this January, thou Im not that much expert on this (doing a course with this topic) we have ourselves needing this for our game and two other customers that would like to have it aswell. Lets see if this goes well or if QA will dismiss it.
Actually the version 1.2 is scheduled to be submitted to MP at January 17th, we got both versions (with and without multiplayer) in test, the deadline is this for now.
This is the integration of Cloudscape Seasons with NeOcean asset from Jonathan Light:and here the ocean set with translucency:
Be sure to check the video little details in order to have them work together!
PS: Videos removed due to NeOcean infriging Copyright and being removed from MP.
It’s so beautiful! 
Hi there. Just for you to know, I didn’t forget you. We are a lot busy fixing some issues in the new version that we “need” to release with multiplayer enabled and also the new feature that we are calling “Timeline Orchestrator” where you will be able to show during a set amount of time (minutes, hours, days, weeks, months, years - you can rescale the time, so one year can be fit in 1 hour if you want, or a day in 10 min) different cloudscape states, that will be automatically blended in a smooth way and with duration also customizable, to create fully dynamic cloud evolution, from clean sky to shallow clouds that can evolve to cloudy states up to storms, and all these states can be triggered by events if you want to, or stay pre-programmed by functions. We still are doing everything in blueprints, not a single plug-in for this!
But besides releasing this new version, we will also release new documentation, some written with pictures, videos with audio, include them also into the HowTo video series, so it is quite a lot of work, but it needs to be done now, before the product becomes too complex and people will start to think it is too hard to use. All this in preparation for the other new features that will smash anything else in the market: storms with lightning strikes (all volumetric) and tornadoes that can go as high as it is visually possible. A new postprocess for rain will be included, with different rain stages to fit the specific scene needs.
Confirmed that Cloudscape Seasons works fine with the translucency mode for NeOcean (now set correctly) like in the video showing both 3rd person and Flight Modes (flying above clouds and under water too):
PS: Video removed due to NeOcean infriging Copyright and being removed from MP.
Tests with UE 4.19 preview 2 revels the asset will be already compatible! You can try it right now!
PS: already tested with preview 3 and works!
PS again: already tested with preview 4 and works!
Finished testing with Network replication and it will be part of the version 1.2 to be submitted in few days. Now just trying to finish the tests with the new feature “Timeline Orchestrator” . Hope to get it as part fo the update even that this means delay for few more days into February.
This video show the clouds perfectly sinchronizing when the client (2nd process executed) connects to the hosting game machine (1st process executed):
We had a fix candidate for the issue with blueprint nativization, but it is not working in 4.19 so we prefere to not include it now and study a bit more why such difference. Tests of other features keep going.
Version 1.2 almost there for submission. While the testing is being concluded, we are rewriting the demo maps for a more pleasant experience, with a new demo executable coming after and finally the so awaited HowTo’s video series.
Just a glimpse of the new Cumulus material setup:
Another fix will be included that has passed QA:
We are still working hard to make the new feature Evolution Orchestrator to come along this submission plus everything else we have said earlier for version 1.2.
Another “fix” or more like a workaround will be made available for the asset be successfully packaged with Vulkan (Vulkan is still as experimental feature). The “fix” actually was done because the current material graph is producing an unknown error in the shader compiler for Vulkan and which is not triggering compilation error producing a successful build, and in the end makes the whole cloudscape invisible in execution. Im discussing this issue closely with the Epic’s engine developer for this part and it might be possible the “fix” is not necessary… anyway a solution exists otherwise and the main material will be duplicated to M_CloudscapeVk and it should be used for while for those who want to test Vulkan.
Now, about the performance… unfortunately the performance it not ok “yet”. I was expecting a great improvement, but as it is for now there is a consistent 16-18 FPS drop on a specific test where the expected framerate was 76 FPS which represents a great loss. Lets hope it gets better before leaving experimental status!
We are talking here about Vulkan for desktop, not for mobile and I didn’t even try to get there, mainly because currently the shader compiler path is the same for both and the solution would still be too heavy for mobile despite the display having less resolution to play with. If you want to try, be my guest, but beware things will get hot!
So far, we have choosen a scene not too complex, with simple materials mostly (all green), FX 9370 3.6GHz and GTX 1080 not OC’ed and to be delivered for versions UE 4.17, 4.18 and 4.19 the remaining conditions:
Conditions: texture sizes 2048, Epic settings, Full HD, dynamic lighting, main light with movement each second to simulate 1 hour = 36 seconds (time pass 100 x faster), minimum SM5 feature set, Vulkan for Desktop
test a) Deferred Render, DX12, SM5, 160 raymarch steps : 69 FPS
test b) Deferred Render, DX12, SM5, 384 raymarch steps : 51 FPS
test c) Forward Render, DX12, SM5, 160 raymarch steps: 84 FPS
test d) Forward Render, DX12, SM5, 384 raymarch steps: 58 FPS
test e) Deferred Render, Vulkan, SM5, 160 raymarch steps : 46 FPS
test f) Deferred Render, Vulkan, SM5, 384 raymarch steps : 35 FPS
test g) Forward Render, Vulkan, SM5, 160 raymarch steps: 57 FPS
test h) Forward Render, Vulkan, SM5, 384 raymarch steps: 41 FPS
The amount of raymarch steps choosen 160 and 384 refers to minimum for gaming and minimum for cinematics quality respectively.
Notice that both APIs (DX12 and Vulkan) have better performance in Forward Render. Vulkan still experimental and the performance as said in the previous post, not ok.
We found another bug that no one reported, but we think it is urgent to fix because many people dont have another alternative for it: BP_SkySeasons (asset given for free to replace the engine BP_SkySphere). The issue is that during night time the frame rate is dropping by half and the faulting piece is this blueprint. More news to come soon, hopefully good news.
The new feature called Evolution Orchestrator is being made in blueprint and the performance was not good enough, so we got back to the issue with blueprint nativization and we managed to make it work! Now the performance is equivalent from a similar implementation we did in C++ and we just need now some time to rework the network replication because of the changes and we are good to go!
PS: all fixed. Now while we reapply all the tests I will finish the work with the new documentation and to update the FAQ, which will get pretty small now that so many things are working properly and no workarounds needed!!