As for me, we’re working on the other project right now and using URP, not FPS game. It seems more stable but it’s a different use case. Internally we gave up HDRP, at least this year and will come back and check it 2022.
Apart from stability and most common features implemented, proper documentation and samples (no, outdated code do pock around on Github is NOT a sample) is the key enabler in adopting HDRP in the production project.
Has been anything released with HDRP? I’d like to play something polished made in HDRP to see how it feels.
@sergiusz308 Same here, I am using URP (which is very, very good now) for a VR project. I initially tried out and did some testing on HDRP (about a year ago) and it was littered with problems so I gave up and switched to URP.
I was going to start on a small side project and thought about HDRP because looks (lighting mainly) wise it is so much better than URP. You can get close results with URP but you have to use third-party stuff. I was hoping HDRP is fully ready by now and everything works as they should -just don’t have the time to deal with added troubles of HDRP.
It seems like a good idea to wait little longer I guess.
Probably not what you’re looking for, since my game is a solo project, but I’ve been pretty happy with HDRP. I’m certainly not pushing the absolute limits of the pipeline, though. The main issues I’ve faced have been issues with trails VFX and particles causing random crashes, and random crashes on certain GPU hardware (like Intel HD Graphics 600 series). In general, though, things feel pretty stable. There’s a steam link in my signature if you want to see a trailer or try a demo of what I’ve made.
I’m not really sure what 3rd party stuff people feel they need. But again, one-man project, and I don’t have the industry experience to know what kinds of AAA features people might be expecting from HDRP which aren’t there yet.
Using HDRP for Buildmark ray tracing sandbox - good enough ). Currently the main problem beside myself is that texture streaming is not working with dynamically added content but this is the case for any rendering pipeline on Unity it seems/ray tracing denosier is far from perfect (but better than UnrealEngine one).
@dgoyette @Lex4art good to hear you’re happy with it, throughout 2019/2020 HDRP was a mess in many areas and it’s good to hear now from people that they can push forward with their projects using it.
I wouldn’t be that picky about it if my project would be a hobby one. For production though, then you got GDD to deliver, within set time-frame, the state of HDRP stack back then was not acceptable.
Once I finish current project I’d definitely give it a go once again, because it CAN look amazing. Thing is that Unity need to seriously catch up with resources and documentation that would enable people (outside of their dev team) to fully grasp it’s potential.
I believe @Onat-H and his team uses HDRP for Harold Halibut game.
Oddworld Soulstorm, probably. A while ago they had a blog post showing their migration to HDRP. They most likely modified it a lot too, no way you can ship a game that has been in dev for that long using vanilla HDRP.
this is not normal at all.
It has never happened to me for the last year and a half that I only work with HDRP.
HDRP is maturing nicely but the performance may be an issue in the latest versions.
why?
latest version… you mean it’s getting slower with new releases?
From what I’m seeing HDRP is ok for things that are not performance heavy like an adventure game.
The bigger your team and the bigger and longer your project is, the more disruptive updating Unity becomes. And whith both SRPs playing catch-up with basic features, you either update all the time or fork and add the features and fixes you need yourself. Specially if you’re targeting consoles.
yup. from 2019 to 2020 I have observed 10fps difference on the same scene in 3 different projects.
And as KokkuHub said, it is necessary to update if you want to have some important basic features functioning the way they should in the first place. i.e. improved decals, but also improvments on lightmaps, proxies, etc.
Now… if this was the first 6 months or a year HDRP was out, it would be understandable and acceptable.
But after almost 3 years… it is not.
Still I will continue using HDRP as it gets the best visual results and it is very important in order to compete with creators and teams using UE4 when visual fidelity and excellent lighting is a requirement. In the freelance and creative agency world you can’t afford offering subpar visuals. But when UE5 comes out later this year… I will most definitely re-evaluate my pipeline.
Well said! This was the exact though I had on URP also, but I am happy to say URP is now very robust. I chose URP because back then it was far more stable than HDRP. It seems HDRP is still lagging behind. Guess will have to wait another year or so.
P.s. Lately, I have been feeling the same with most things in Unity. Everything is in a perpetual preview (aka early access) state and after it is finally done, all the other systems have moved on to a new stage making this package obsolete (in a practical sense).
You seem to have a lot of production experience, I’d appreciate if you took a look at the quirk navigation list and update it with your findings. It’ll sure help me chose for the next prod and I see that other people take a peek also, it cold help many who need to make the same tricky choice.
https://docs.google.com/spreadsheets/d/1HwqJBj9J-2UFFxuxtSYL64hQ792gKrOD-Q7NLZqtEpM/edit?usp=sharing
What % slowdown is that?
Very good point. It seems that choosing the right version of unity and the packages has become in itself a full time job.
It has become a real problem but like everything in life there is a positive aspect to this: maybe it’s a sign that we need to change the way we produce games. Instead of running after the latest glitz and convenience (I do that) which will break or slow down everything else as we’ve seen in 2020.2+, we need to scope our projects to match the most stable and feature full version and after bench tests stick to it like the big teams do. In my case I love jobs or at least burst so I can take off <2017 from the list.
A good example, and huge inspiration, is the small team that made Dyson Sphere Program. They use 2018.LTS with packages from 2014 - ancient. You can tell they scoped their project by matching systems and visuals to what they had.
it is about 15-20% slowdown depending on the scene. Which is an important amount.
Very true. But it does not apply to everyone.
If you are an indie developer developing your next game, and you don’t do that, then you are doing something wrong.
If you are a studio that offers services to others, then chances are you can’t.
If you do it, you will not be able to compete with studios using Unreal.
You will not be able to take visually complex projects. You will not be able to satisfy your clients who say “but we saw Rayreace in a video from Unity” In which case, over time, you run the risk to appear outdated, less informed, and lose clients. If you run a service studio or small agency, your clients apart from delivering on time, expect you to be on the cutting edge for their projects. It is part of their marketing too.
I believe this is using HDRP. It was shown @ unity showcase/GDC.
I think this applies to just about everyone, including indies. Industry is very competitive and if anything it is even harder for indies because unlike big teams, they have either very low or no marketing marketing resources available to them. I know, if the game is crap, it does not matter how good it looks, but visual fidelity goes a long way for indies to increase the appeal of the game to prospective players (the clients of the indies).
Perhaps you would think so but it is not necessarily true. Many people mistake appeal with fidelity. Indeed appealing visuals are key. But photorealistic fidelity is not.
A game can be presented in many different ways. A photorealistic approach is not necessary for most games. The proof is in the fact that the vast majority of games are not photorealistic. They do not even fall into stylized realism, Often they offer an illustrative style and even cartoony or completely cubist, or plain graphic design. Not only it is cheaper, it is also faster and more creative.
In agencies, especially these involved in ArchViz or VFX, Marketing, Automotive, and Digital Production that game engines push their way recently, photorealistic rendering is pretty much a one-way road.
Yes, it looks like it. And if one looks carefully, it also showcases some of the weaknesses of HDRP.
i.e. materials can have a very “plastic” look if you do not fine tune each and every thing, And even then sometimes, you can’t avoid it. It appears to be an issue with how light interacts with shiny or semi-glossy surfaces. Volumetrics under certain circumstances, are not very flexible in terms of intensity tuning, lighting exposure has issues with high contrast bright areas. A bright sun can fill the sky and more.