An important question: HDRP or Built-in for commercial games.

¿Should we use the HDRP or the built-in for a commercial release?​

Sadly, I feel that only a few people (if anyone) is able to fully answer this question.​

​

I will present a bunch of general premises for a game released in late 2021 early 2022 and what I think is a common batch of needs for a game project with HDRP (There are more combinations, but I think this selection is pretty common and I suppose for what HDRP is intended to be use)

Note: Please read the responses before answering, they will bring more info and insights

Project Type:

• FPS / TPS or any Close range game view.
• Open World, medium size environments with Natural and Urban environments

Premises:

• Being a commercial game the devs are investing time and money so any payment below 100-200$ assets are ok, so if there is an AssetStore asset should be take into consideration (For example ASE is pretty common, so the ShaderGraph is nice but nothing essential).
• Realistic visuals "as good as possible"
• The performance budget is a game able to run in mid-high devices like PCs and consoles.
• No Build size limit
• Full Release: a range from late 2021 to mid 2022
• The developers have C# knowledge
• Enough general experience and knowledge (but not expert) in both render pipelines

Special Considerations:
Things that are important for achieving a good game

• Development Coste/Time needed (Time needed for similar results)
• Stability
• Complexity
• Risks
• Visual Results
• Performance
• Flexibility of results, for example day cycle or weather
  
• Distance view

Note: If needed I can edit the post to include, or modify things.

 

I can't answer right now I'm literally in the middle of both options

Advantages of Built-Int:

• Realtime GI*
• Tessellation * (In HDRP is limited)
• Better General compatibility with other assets
• Billboard system
• "Known development environment" Less development costs
• Stable tools like terrain with full vegetation
• Works on older devices (+ General compatibility)
• Known stability, as a technology has much more testing in released games
• Tons of assets in the asset store for the lack of features

Advantages of HDRP

• Better lighting
• VFX Graph
• Better visuals (Not 100% sure)
• Future proof
• New features with each update

Things that I'm not really sure:

• Performance

 

Pfff hard to decide...
I wouldnt want to be in your shoes
I think better idea is to start with built-in renderer because as far as I see, hdrp is still really problematic in terms of open world enviroments. No proper tree support, no proper grass support, no speedtree support, either I dont think terrain is super nice. But I heard that demo team is currently working on an open world demo which is I think will be based on hdrp. That is why I say start with built in.
On the other hand, there are millions of terrain and tree tools out there in the market which u can use for built in renderer. And it wont be problem to switch to hdrp when it has proper support...

 

Main difference is that in build-in(or custom) you can get what you want, but in HDRP you get only what hdrp-team do, nothing more.

 

If you want lighting and camera-simulation (!) to match real-world conditions as closely as possible then HDRP is for you, though sounds like landscapes are not its strong point (not tried).
If you do not want real-world accuracy or want to stylize the look of your game or need a consistent, accessible API to access the graphics pipeline in code then HDRP is not for you.
You can add some features via shader graph like custom materials, billboards etc.

 

You can pretty much match HDRP visual quality if you stuff your project with the right 3rd party assets. Vanilla built-in is pretty barebones. The risk/cost is researching and picking up said assets, then making sure they work well together.

The risk with HDRP is that you are almost entirely at the mercy of Unity's internal teams and their bizarre priorities. If you need any feature that still doesn't exist and has no HDRP 3rd party asset equivalent, you're in limbo. Case in point: it's almost 2021 and HDRP still doesn't support terrain grass.

The biggest advantages of HDRP are shader graph (easier for artists) and VFX graph. On built-in there's ASE, which runs loops around shader graph, but I'm not familiar with any asset which provides GPU particles equivalent to VFX graph.

 

I would go with HDRP.
But since you are starting you wont need graphics for about 1.5+ year anyway.

 

I think the biggest difference between HDRP and everything else is the camera relative rendering.

This feature alone is unique in Unity environment, and for an open world game, hard to ignore.
Just makes everything work better.

Everything else is pretty much do-able in Built-in although it will get ugly and hacky with all third party stuff.

 

It depends what you mean by 'commercial' game.

If you mean turning round asset store content for quick sales on steam, probably vanilla is the way to go, since you avoid a lot of compatibility issues. But I'd also recommend going into a casino and putting the money you'd spend on the asset store on a roulette table, as it has a better return.

If you're developing professionally (i.e. will make the content); URP is the go-to for most use-cases, since if you only have a small team and don't need AAA-visuals, then it's better to be able to keep the door open for switch, low-end PCs, or even mobile devices.

If you want to make a AA-ish FPS etc., and have a reasonable team, then HDRP is viable (see GTFO), but you'd expect some of the team members to argue for UE. HDRP is a good challenge from Unity against UE4, but the majority of developers with experience with high-end visuals are more experienced with UE4. HDRP, despite much criticism, is actually good at high-end visuals, arguable moreso than UE4 in some use-cases. But there's a comparative lack of tried-and-tested HDRP projects that prove this point, so the dilemma you should be having if you're going for high-end visuals on a modern engine, isn't really HDRP or other Unity pipelines, but HDRP or UE.

I can say from experience trying things out with HDRP it can do highly detailed AAA large environments, at competitive FPS, but you need to know what you're doing same as UE4 or any other engine (as a yardstick, if you turn shadows on for all lights or don't know what a trim sheet is, you probably don't know what you're doing).

 

I'm sitting here not even started grass in my hdrp project yet and I bought Microsplat to overhaul the terrain system. Are people saying grass on terrain doesn't even work in HDRP yet?! How could they even release it in that state if so?! lmao

 

Just to clarify a commercial game is a game in which you invest time and money and you want to publish on a store (like Steam) with a community, support, patches and you intend to obtain a gain from it, being a full featured game (Even if it's a EA).

The UE question I think is a nice questions to think about, but, this is not really the topic to discuss this, just to avoid letting it without response, I believe that the problem with "unreal choice" is that you need to hire a unreal team or "recycle" all the team to learn how to use unreal which can easily take 6 - 12 months of learning.

@KokkuHub has a great point and somehow this is what I'm doubting about:

And if the gains from HDRP are justified to invest the bigger effort/investment it requires.

The problem is that I have been listening promises from more than a year ago that HDRP will be production the "next month" and right now I don't feel that this an accurate assessment, there are only a few successful attempts of really skilled developers, but nothing you can rely on as a company/team (or at least not with enough confidence).