I would like to see DLSS in URP
For instance, I use URP for programming VR games that are highly demanding so I can’t make them with HDRP.
We don’t plan to support DLSS in URP at this point and it’s not in our roadmap for at least the 23 versions. We would love to support different techniques in URP but our first priority at the moment for URP is functionality completeness with techniques that can scale across different hardware.
Hi!
Are you targeting PC VR or standalone VR (eg: Quest)?
For PCVR, HDRP is actually pretty strong at handling demanding graphics since it offloads a lot of the work to GPU, leverage DLSS or TAA Upscale, and things like cache shadows to save perf. CPU is usually the bottleneck.
Check this one for example:
We could really do with more guidance on making this happen. I’ve seen https://docs.unity3d.com/Packages/com.unity.render-pipelines.high-definition@7.1/manual/VR-in-HDRP.html but that’s pretty light on details. What the ArchVizPro team do is amazing but they obviously have a lot of experience and knowledge on the precise settings needed. It’s a black art that’s hard to get info on. (Also it really is just for modern high end PCs. Even with their excellent demos I can’t hit 90fps on a GTX1080ti PC, or RTX2080 laptop)
As a heads up, when I recently tried to suggest on reddit that HDRP can be used in VR I was massively downvoted. The common understanding seems to be that “HDRP is not made for VR, its made for film production and high level graphic games”
I will take a look to that.
Thanks!
Oh look. More techniques that scale across hardware. Will we have a proper upscaler now?
https://community.amd.com/t5/gaming/amd-fidelityfx-super-resolution-2-1-out-now-even-more-fsr-2/ba-p/544170
IMHO, upscaling features should be independent of render pipeline.
‘HDRP-only’ features should be limited to things that are fundamentally different in HDRP, such as lighting.
Postprocessing, upscaling, and antialiasing features (TAA) shouldn’t have to be tied to a render pipeline, nor implemented multiple times to support multiple pipelines.
(Yes, some features require specific inputs such as motion vectors or a depth texture - but those now all exist in URP and the standard pipeline, don’t they?)
Could not agree more. This is the sort of idiotic thing that has me very frustrated with unity in recent years. The scriptable pipelines caused a massive userbase split, and they continue to push them apart with pipeline-specific additions like this.
It’s NOT ok for basic level functionality like this to be selectively available. Unity management really needs a twitter-grade shake up.
my case is using built-in rendering to drive larger than 4K screen…