Search Unity

  1. Unity 2018.3 is now released.
    Dismiss Notice
  2. The Unity Pro & Visual Studio Professional Bundle gives you the tools you need to develop faster & collaborate more efficiently. Learn more.
    Dismiss Notice
  3. Want more efficiency in your development work? Sign up to receive weekly tech and creative know-how from Unity experts.
    Dismiss Notice
  4. Build games and experiences that can load instantly and without install. Explore the Project Tiny Preview today!
    Dismiss Notice
  5. Nominations have been announced for this years Unity Awards. Celebrate the wonderful projects made by your peers this year and get voting! Vote here!
    Dismiss Notice
  6. Want to provide direct feedback to the Unity team? Join the Unity Advisory Panel.
    Dismiss Notice
  7. Improve your Unity skills with a certified instructor in a private, interactive classroom. Watch the overview now.
    Dismiss Notice

Occlusion Probes - Where are they?

Discussion in 'General Discussion' started by jjejj87, Oct 14, 2018.

  1. jjejj87

    jjejj87

    Joined:
    Feb 2, 2013
    Posts:
    226
    So, I've waited long enough with the hopes of the Unity team providing more information about Occulusion Probes. The book of the dead video was posted 9 months ago, and this very thread about 5 months ago and I feel it is time that occulusion probes story gets a revision. Before I begin, I'd like to express that I mean no disrespect to any specific individuals.

    1. Unity marketing video misleading developers
    I think since the Unity 5 release video, I have noticed that the video material released by Unity has increased both in quality and frequency. I am happy that the main engine that I love is getting the media it deserves but there have been many cases where the videos would often suggest one thing but in reality, something slightly different.
    Here is the Unity 5 official trailer

    Here is the blacksmith trailer


    For the Unity 5 trailer, the first thing that we see is "Realtime Global Illumination". And as we all know it, it isn't really realtime and it requires precomputation. It also had the nightmarish CPU bake which really just made GI not a real option. Realistically, the light probes system were the only things close to realtime global illumination, and as we all know it, it is very limited and visually incorrect. The video made me feel like something close SEGI but in reality well very different. Still, this could have been my assumption and feel that I picked up after watching the video.

    For the blacksmith trailer, again, like the Unity5 release trailer, it had many new things that made it look super awesome. I did not fail to read up on it as I suspected something similar to the Unity5 could happen with just a video. After reading couple blog posts, I was convinced that this time, it was really happening. Unfortunately, it didn't happen exactly. First of all, the grass and all cool foliage shaders and workflow was never implemented to the vanilla build. The wrinkle shaders (tension map based normal manipulation) also never made it to the vanilla build. The atmospheric scatter as well. They all worked, and with reasonable effort, they could also be re-implemented to my project as well. Or so I thought, the features of blacksmith, started conflicting with common 3rd party assets, and as the Unity version went up, some of the things just stop working.

    Then there is the Book of the Dead, which by the way looks really impressive. However, as the existence of this thread points out, occulusion probes were part of it, but will not be part of vanilla. There is also the issue where the "assets of the trailer is released" but not the actual trailer files themselves. At first, I find it "similar enough" but the more I dig in, the more I find the discrepancy between the trailer and the released material. Sure the graphical assets are there, but not the whole thing. The trailer suggests "the following can be done with Unity!" sort of message but I the released assets tell "the following can be done with Unity but it requires many other things not included or supported by Unity". A little cheesy, but I think enough comments were made regarding this particular feeling.

    Why am I saying this? Well, I want to acknowledge that Unity is a business, and the marketing/media division is doing their best to oversell the product without definitively misleading consumers, for the sake of hype I guess. I don't relatively find this strange or disturbing or ill intended. It is sort of expected tbh. And let's be honest, when we advertise our games made with Unity, we take creative license and try our best to make things look best. It really is just good job done by the marketing division.


    2. But there is a problem.
    Here is my main concern. The games that we make and advertise gets consumed by the end user, the gamers. Unity engine's consumers, the developers, are not end users technically. See, the developers and gamers, the two crowd, look for very different things: developers developing the game and the gamers, playing the game. The way materials are presented since Unity 5, feels more targeted at "end consumer or gamer" instead of developers. Sure there are hour long videos on GDC and Unite where things get more "developer friendly" but even those, aside from other developer talks, are more about flash and bang. One fellow developer mistook the presentation at Unite as an Apple product release. I am just trying to make the point that the media direction feels "off".

    So, looking at how things work, I tried to stay objective and not just be a loud voice. But, throughout my development history using Unity, I have noticed this.

    Amazing looking trailers and presentation at events

    BUT

    Using temporary features to oversell Unity is becoming a repeated and almost expected phenomenon. (occulusion probes and etc)


    3. Then there's more.
    Now, to close my long post off. I love Unity. I love it because it was my first choice when I jumped into this industry and after several years of evaluation, it still is the platform I choose to develop. The recent development pace of the HDRP team in particular is just out of this world and I do feel like Unity is my "engine team" honestly. However, I am concerned about things like occulusion probes - not the feature itself but the way it was handled. There should have been a clear indication that it was a temporary, non-vanilla feature, only intended for demonstration. The fact that it didn't, and that users had to ask their way around on the forum, while it is not the biggest problem of all time, may be an indication that Unity, as a game engine company ,needs to re-evaluate how media/video presentations are done at higher management level.

    I can live with or without occulusion probes. Hell, I will spend time implementing it myself if it becomes that important. But that is not what I am trying to talk about. I am trying to say that maybe, just maybe, there are some directional corrections that us the developers suggest that the management needs to look at regarding videos and media presentation. Please remember that you are selling game engines to game developers.

    Hope this helps.
     
    AntonioModer likes this.
  2. LurkingNinjaDev

    LurkingNinjaDev

    Joined:
    Jan 20, 2015
    Posts:
    1,937
    Oh, I haven't noticed, time flies. This is the 'Unity shouldn't apply techniques in their demos unless it's possible in the engine with two mouse-clicks'-time of the month again. Sigh.
     
    Kiwasi likes this.
  3. hippocoder

    hippocoder

    Digital Ape Moderator

    Joined:
    Apr 11, 2010
    Posts:
    23,832
    I split this topic as it's probably deserving of it's own so the original thread is kept mostly focused. It's a good point though.
     
  4. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    11,164
    But that's not what's being said here. Towards the end of the post the point is re-stated in pretty clear terms, and I think it's perfectly reasonable:

    Yes, extending tools and writing custom functionality and effects is a perfectly normal part of game development. So, of course I expect their demos to do both. And I understand that lots of stuff is use-case specific, so it doesn't at all bother me that the scattering effect from a demo can't be dropped into my own game, and so on. Still, I think it's perfectly reasonable to want transparency about what's "out of the box" and what's highly customised, use-case-specific when demoing a product.

    As a developer nothing in the above cited videos bothers or mislead me. But I have certainly worked for managers and/or clients in the past who might see such a video and then come tell me to "put atmospheric scattering in the game" under the misguided belief that it's built in and easy because Unity does it for me, and that's no good for anyone.
     
  5. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    4,191
    I'm... not on the total opposite side of the fence, as I fully understand that as developers we have to, well, develop, but I think it'd be nicer if it was a little more clear. Like, even just in the Youtube description or something, saying how they built their own framework to handle certain things. In the grand scheme of things, it's honestly very little work.
     
    Kiwasi and angrypenguin like this.
  6. LurkingNinjaDev

    LurkingNinjaDev

    Joined:
    Jan 20, 2015
    Posts:
    1,937
    I think it was very clearly indicated multiple times by multiple Unity employee that

    - what the demos show is doable by a small team of professionals
    - shown features may or may not become vanilla features in shipped Unity Engine built-in function

    It is unrealistic to expect labels on every single frame which shown in the demo that "this pine log probably will be released in a package later" or "this feature probably never make it to the engine" (WTF are you guys talking about?).

    Not to mention that probably even Unity does not know if they can deliver a feature in the vanilla engine on a timely manner with a decent performance. Making a custom demo and put the same thing in a multi-platform environment are two very different beasts.

    But I think we discussed this very "issue" at least six times already.

    Oh, and BTW, as a developer who works for hire, it's your job to show your customer, what you can or can't do in what time-frame and for what kind of money. I'm an enterprise developer, who used to work in the web business too, believe me, I know what kind of expectations can be surface in the clients' head. Kind of the same as to expect unity to label their demo: mission impossible.
     
  7. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    11,164
    Wait... have I screwed up? That's what I was trying to say. :confused:
     
  8. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    4,191
    I'll be honest, I'm kinda medicated and may have misinterpreted your entire post.
     
  9. Adam-Bailey

    Adam-Bailey

    Joined:
    Feb 17, 2015
    Posts:
    165
    Problem with demos doing things like that is it can be a little misleading. For an example an artist deciding what engine to build a portfolio piece in may see something like the book of the dead demo and have the artistic ability to create great assets but be unable to create anything that looks as good because the engine available to them simply doesn't have the demonstrated features.

    Unity is getting better in this regard, but one major advantage that Epic has in this space is that their UE4 demos work as advertised and users can pull them apart and learn from them.
     
    OCASM, jjejj87 and angrypenguin like this.
  10. LurkingNinjaDev

    LurkingNinjaDev

    Joined:
    Jan 20, 2015
    Posts:
    1,937
    Ok, last answer in this thread and I will unfollow it.

    I would be very surprised if one artist alone can pull of a Kite demo or an Infiltrator demo without the "small team of professionals" in a timely manner. Oh come on. Those demos were built by a team as well, they wrote code, they made shaders and special things to handle different things. I see no difference. Both engine is extendable, both teams did a great job, all of the demos require specialists of some kinds. So whatever.
     
  11. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    11,164
    I agree entirely... so it'd make sense for everyone to point out when they're using anything non-trivial which doesn't come in the box.

    I admit that stance makes it more of a truth in advertising discussion than a Unity specific one. Still, it's pretty common for products in other industries to include disclaimers with their advertising - "Accessories sold separately", stuff like that. Something like "Demo includes custom assets and code" would get the job done. It doesn't detract from the demo, but it would possibly help clarify things for the part of their audience Unity sells tutorials to.

    One thing that I find interesting is that certain visual effects tend to get jumped on for this. I don't notice it coming up in conversation when a 3D model isn't included, or a behaviour script. I wonder if it's because visual effects are easily mistaken for engine or renderer features rather than being seen as assets or content?
     
    OCASM likes this.
  12. hippocoder

    hippocoder

    Digital Ape Moderator

    Joined:
    Apr 11, 2010
    Posts:
    23,832
    IMHO

    1. lightprobes using tetrahedrons super bad for tight lighting. Because it's incredibly hard to automate and as an indie you do not want "artist control" for something that is very much automated by AAA games. Really, you want your artists diddling around with probes or making art? in our case we have to fight it endlessly to automate it, and it still doesn't work well due to the choice to use tetrahedrons. Which are fine for simple topology but break hard where you have walls and ceilings with gameplay on either side.

    2. LPPV is limited because it's a 3D texture, and still doesn't solve tight cases common in games. I don't mind it so much because sometimes you need lighting across huge objects, but in case like that - use a fill light. HDRP has a mode just for that soft kind of fill lighting now.

    3. Occlusion probes follow the idea of LPPV, and come with similar caveats. In other engines, occluders are used near the object, and can be automatically generated easily. Not saying they're bad in this case but limited to a specific scenario like LPPVs are.

    Purpose of this monologue?

    So Occlusion probes not being a native feature... well I can see them being niche myself, so it's not actually that vital to me if they go in or not. In fact I would probably argue they are a solution for a specific scenario.


    * Unity's lighting systems strongly avoid testing against scenarios with walls or roofs that have gameplay on both sides. Something I very much do test against because playability > whatever decision Unity makes
     
    AcidArrow likes this.
  13. jjejj87

    jjejj87

    Joined:
    Feb 2, 2013
    Posts:
    226
    I think you are missing the point of my original post. What you are saying is that "team size" and "custom solutions" are part of development. This is entirely correct and there is no arguement.

    What I am saying is that an engine feature video should not be showing off what a team or custom solutions can do. Instead it should be showing off what the engine can do in its vanilla state. Imagine a paint brush company showing picasso's painting and saying "this is what our brush can do!". We laugh and understand that it is just marketing, correct?

    Well, for game engine features, it is hard to understand what is and what is not vanilla. It is hard to distinguish marketing from what is actually included. Thus, it is difficult to just laugh it off and mostly likely be misled. That was the point of my post. Hence my suggestion to management, a change in its approach to its media direction.

    The two are entirely different things.
     
  14. hippocoder

    hippocoder

    Digital Ape Moderator

    Joined:
    Apr 11, 2010
    Posts:
    23,832
    To be fair UE4's infiltrator was shown and still there is much in it that isn't part of the engine. But it can all be done in the engine. But they aren't one button features.

    What us users must do is understand that all games have niche situations (like my above post) and so long as the engine empowers us to achieve those demonstrated things, it's good enough.

    Because that is a tech demo. You can do this with the technology provided. It's showing a workflow that is achievable with the engine.

    Am I defending Unity's tech staff? F*** YES I am.

    Am I defending Unity's marketing dept? Only an idiot would defend the evil of any marketing dept and they wouldn't have it any other way... I'm sure they'd put a spin on it though :D
     
    Ryiah and Kiwasi like this.
  15. toto2003

    toto2003

    Joined:
    Sep 22, 2010
    Posts:
    490
    these reflections of the particles on the car, are they from realtime reflection probe or ssr post effect? they look pretty neat.
     
  16. LurkingNinjaDev

    LurkingNinjaDev

    Joined:
    Jan 20, 2015
    Posts:
    1,937
    I was thinking about this issue in this past week, regardless that I said I won't follow up. You all know, how it works...

    I think this whole issue is a little bit blown out of proportion because of the reasons we already discussed above.
    On the other hand Unity could do something about it. If Unity were making 'how it made' videos about these features, like Unreal did about the Infiltrator, it would help the situation very much.
    They (Unreal) explained how they set up the scenes, how they made the seemingly 3D explosions, how to set up the screen space reflection, etc.

    Maybe it would ease on the guys who want to build something similar to what Unity shows in these demos if Unity would take the time and explain how their team built it.

    I know, obvious.
     
    Adam-Bailey likes this.
  17. hippocoder

    hippocoder

    Digital Ape Moderator

    Joined:
    Apr 11, 2010
    Posts:
    23,832
    I see a lot of occlusion stuff sprinkled around the Lightmapping/probe API so it's probably going somewhere at least.
     
  18. thelebaron

    thelebaron

    Joined:
    Jun 2, 2013
    Posts:
    147
    I always do a mental sigh when I find out stuff isn't part of the out of the box package for these demos. I get it, its been said before Unity wants to demo what a small team can do with Unity's power and sure they totally have that right.

    But if you want to gain traction as an engine for use cases that are high fidelity, imho a tech demo should show whats feasible out of the box with minimal customization. To me its more impressive if you use the constraints of whats given and pull off amazing visuals than to go adding on custom hacks for a non game tech demo.
    I know BOTD is better than The Blacksmith in this case, but Ill still say my peace. The customizations make me think that same team could probably pull off something even more impressive with Unreal, considering there is more stuff built into the engine in its current state. The fact that Unity doesn't even make their own games with their engine kinda adds another layer to this.

    Alright thats enough whinging out of me.
     
  19. Adam-Bailey

    Adam-Bailey

    Joined:
    Feb 17, 2015
    Posts:
    165
    Relevant to this thread, massive kudos to Unity for the FPSSample project, exactly what many people have been asking for for years. Sure there is lots of custom stuff in there, but that's what will happen in any big project. The important thing with it is that it's all there for people to look at and pull apart, and there's no question that it can all work with stock Unity. :D
     
    OCASM likes this.
  20. hippocoder

    hippocoder

    Digital Ape Moderator

    Joined:
    Apr 11, 2010
    Posts:
    23,832
    It's something I asked for a few times, it's a logical and strong move. I guess Unity just needed time to evolve their internal demo team.