But…But here it says SVOGI was developed by Scheidecker at Epic.
Quote from site:
The original research paper was made by Cyril Crassin.
+1 for what KingBadger3D said^^
But SVOGI was run on a single GTX 680 at 1080p. I would guess that Epic might look to reintroduce it (or something like it) when they feel it can run reasonably on a mid-level card (maybe the 860 or 960).
We chose to remove SVOGI in order to reduce our maintenance burden and allow us to iterate more quickly on other new and exciting features. It takes a considerable effort to maintain a complex system like SVOGI as the engine grows and evolves: every time we add a new feature or modify an existing feature we have to ensure that it works with all existing features.
For that same reason we do not plan to re-integrate SVOGI in to the main branch of UE4.
We are considering other options but SVOGI has tendrils throughout the rendering code so it is not a good candidate for a plugin right now. It would also be a considerable effort to get it working again with the latest code.
–
A lot of our efforts recently have focused on improving static lighting in UE4. That was motivated by bringing up the mobile renderer where static lighting is still largely required as well as looking in to solutions for games that want to target high frame rate (usually 60fps, higher for VR!) while also maintaining a high visual bar.
I think dynamic lighting is an important feature and workflow for many games and we will improve our dynamic lighting features over time. UE4 now has experimental support for Light Propagation Volumes as some have pointed out and I think that can be a good solutions for some games. We will work to improve LPV and we have other ideas we plan to explore to provide options for games that do not or cannot rely on Lightmass.
We chose to remove SVOGI in order to reduce our maintenance burden and allow us to iterate more quickly on other new and exciting features. It takes a considerable effort to maintain a complex system like SVOGI as the engine grows and evolves: every time we add a new feature or modify an existing feature we have to ensure that it works with all existing features.
For that same reason we do not plan to re-integrate SVOGI in to the main branch of UE4.
We are considering other options but SVOGI has tendrils throughout the rendering code so it is not a good candidate for a plugin right now. It would also be a considerable effort to get it working again with the latest code.
–
A lot of our efforts recently have focused on improving static lighting in UE4. That was motivated by bringing up the mobile renderer where static lighting is still largely required as well as looking in to solutions for games that want to target high frame rate (usually 60fps, higher for VR!) while also maintaining a high visual bar.
I think dynamic lighting is an important feature and workflow for many games and we will improve our dynamic lighting features over time. UE4 now has experimental support for Light Propagation Volumes as some have pointed out and I think that can be a good solutions for some games. We will work to improve LPV and we have other ideas we plan to explore to provide options for games that do not or cannot rely on Lightmass.
[/QUOTE]
Mobile is the new ‘Console’ now, downgrading for the masses.
1 Like
@Nick_P, Take a look at the papers ive posted and the idea’s im pushing. It’s not rocket science, If you have something better on the burner let us know but avoidance tactics with no detail is weak.
I understand that SVOGI is such a heavy tech. Although, I think if Epic would ever deliver us those tech demos (may be Samaritan though it’s #MadeWithUE3; Elemental & Infiltrator - both are good, but I find Elemental still one of the most impressive thing on market besides modern engines like Snowdrop or Luminous engine) then it’d be just spectacular to run that actual 2012 demo that once Alan Willard made a walkthrough on. Those statues, that spotlight sphere, etc. <omnomnom>
*I may be wrong about it (dropping the tech demos into marketplace) but I’ve heard about that… somewhere))
**And yes, SVOGI would probably cause an enormous perf hit on my rig (because even GTX680 is a bit outdated nowadays), but it still would be an interesting experience to try.
For that same reason we do not plan to re-integrate SVOGI in to the main branch of UE4.
[/QUOTE]
Its really sad to hear that . Honestly I wished to try SVOGI because i really really loved when Alan Willard demonstrated it. Till now i have always watched that walkthrough video over and over again just to see the beauty of it. I was totally amazed when i saw the light environment changing from day to night. It was just…spectacular.
Just wondering…instead of integrating it into the main branch of UE4, can Epic share the code for SVOGI (and maybe a note on how to actually integrate it) so that other members here can integrate it for themselves and upload it to Github? This way Epic dont have to worry about maintenance burden or anything else. Just a suggestion. 
@ryanjon2040, I can’t promise anything but I’d like to!
@KingBadger3D, Sorry, I’m not trying to avoid anything, I just can’t promise features we’re not actively working on. What I said is accurate: we are working to improve LPV and we have other ideas we’d like to explore (but aren’t because we are busy working on other features). We’ll be releasing updated versions of the engine on a regular basis so you will be able to see our progress and what we’re working on pretty much as we do it, including WIP features that aren’t quite ready like LPV.
That gives me hope!! 
You don’t need include third party code, just compiled libraries, headers and integration. NVIDIa shouldn’t have any problem with such integration as they are doing it for years as of now.
The question is how efficient is GI Works. As it also volume based solution.
Precomputed Radiance Transfer.
PRT is actually nice thing. It’s quite useless for closed and small spaces, as you are limited by probe density over map. But for open level it is good technique. You don’t need super accuracy, you just need lighting variance. And that is what PRT provide at very low cost. More over it works with time of day.
Yep, I also love it.
+1
That was my suggestion as well: Suggestion on SVOGI - Rendering - Epic Developer Community Forums
Providing a stable LPV implementation and something like SSDO seems to be a good solution for now.
Agreed. Some kind of dynamic GI that isn’t dependent on baking light is needed. Getting mobile tech caught up is nice and all, though I’d rather see innovation being driven forward with a focus strictly on Next Gen. Same thoughts regarding infinite sized seamless (multiplayer)worlds.
Is this thread still active or is there someplace better for discussion?
Seems like Octree might not be the best approach for SV Cone Tracing. I’m mainly an artist but always searching around to see what others are attempting, looking for papers etc. Just a few links that might be worth reading: Game'N'Motion: PlayStation 4 Partially Resident Textures & Sparse Voxel Octree Cone Tracing ( PS4 PRT SVOCT PSSL )
http://www.ign.com/boards/threads/so-could-the-x1s-secret-sauce-be-voxel-cone-ray-tracing.453362563/
Real-Time Voxels: View Centered Voxels
Also, aside from Voxel Cone Tracing thought this looked interesting on the subject of Tiled Resources/Partially Resident Textures. http://graphinesoftware.com/
http://forum.beyond3d.com/showthread.php?t=64333
Grasping at straws I know.
Yes id hate to see the focus on last gen console and mobile technology even though i understand the hype and all mobile gaming is a big fraud i have not seen a single big game that wasn’t pay to win.
Yea that sounds amazing, more procedural worlds like minecraft but with more advanced graphics.
Iv seen many great realtime global illumination techniques, so i hope just we can get something fully dynamic in unreal engine besides vxgi that only works an nvidia cards.
That was one hell of a Necro…
I can run vxgi on my AMD card.
Some users mentioned that cryengine is using a LPV system but thats actually wrong… They changed to SVOGI or SVOTI, not sure what the real name of the new system is.
I wrote it alreday a few times but i also don’t understand why epic’s effort is so “low” to integrate a nice dynmaic GI solution.
Cryengines SVOGI or SVOTI… seems to run very well even with “2 bounces” The results are really impressive…
Cryengine Interior Scene SVOGI "2 bounces"https://youtube.com/watch?v=hb4N8m5jAfE
Second Scene
https://www.youtube.com/watch?v=bnvmB9iztaM
[QUOTE=Adik;364887]
Some users mentioned that cryengine is using a LPV system but thats actually wrong… They changed to SVOGI or SVOTI, not sure what the real name of the new system is.
I wrote it alreday a few times but i also don’t understand why epic’s effort is so “low” to integrate a nice dynmaic GI solution.
Cryengines SVOGI or SVOTI… seems to run very well even with “2 bounces” The results are really impressive…
Cryengine Interior Scene SVOGI "2 bounces"CRYENGINE 3 [SVOGI] Baron Haussmann Interior Rendering Demo - GTX 780Ti - YouTube
Second Scenehttps://youtube.com/watch?v=bnvmB9iztaM[/QUOTE]
Third scene: Outdoor
Unless they changed something, the dynamic GI in Cryengine only works with sunlight