Jump to content
The Dark Mod Forums

peter_spy

Member
  • Posts

    3201
  • Joined

  • Last visited

  • Days Won

    109

Posts posted by peter_spy

  1. Mine is still more like TDS tribute, but more subtle, with a certain pattern and a pulse table + shaderparm3 multiplier, so I boost in in more lit environment if I need to.

    35 minutes ago, STiFU said:

    However, we're all pretty happy with the current outline implementation.

    You have a straight on full-white color, which is as crude and in-your-face as you can get. And it enhances aliasing for straight lines. Still a lot of work needs to be done to make it look unobtrusive.

  2. 15 hours ago, Anderson said:

    candle warmly reminds me of the pleasant shades from Thief: Deadly Shadows.

    Except TDS had cubemap frob highlight and the outline is more like a Dishonored 2 thing. Although here it's as subtle as a fart joke. You don't have to be an art graduate to know that extreme color values work only in particular situations, and this is definitely not one of those. Setting outline alpha to something like .5 should help in making it a bit less jarring.

  3. You might want to take a look at more modern games like Deus Ex Human Revolution, where both the world and UI are driven by a certain art style and aesthetics, so there's no any particular divide between them, and looks more coherent this way. TDM assets are not coherent, and there's no one art style everyone has to adhere to, and so even UI is moddable. Well, was.

    • Like 1
    • Thanks 1
  4.  

    3 hours ago, stgatilov said:

    Yes, no way to customize.

    Why is that? I have a solution that looks way better, aesthetics-wise, with my assets, than what you came up with in 2.09. Is there no way to leave the option to customize this for someone who really knows what they're doing? Otherwise, that effectively locks me with 2.08.

  5. On 5/27/2021 at 12:37 AM, totallytubular said:

    I don't want to start a discussion over on their forum.

    I sound like an a**hole nit-picking their code.

    Actually, Robert might appreciate this, there's nothing wrong with honest code review ;)

  6. You'd have to go and ask him either on idtech4 discord or on the forums, he's Takeshi Covacs https://www.doom3world.com/index.php?forums/rbdoom-3-bfg.19/

    Given his huge experience in this, I doubt that he made a mistake in implementing the PBR model, and the artists working with him are using correct PBR data (textures/material setups) for reworked character and environment models. IIRC, they're using the metallic-roughness model (as opposed to specular-glossiness). Haven't popped in the idtech4 discord for a long time, but the screenshots they posted there looked correct / similar to other PBR games.

  7. 6 hours ago, Frost_Salamander said:

     I have to say, that is probably the most discouraging thing that anyone has said to me in recent memory.

    Don't let that discourage you at least from doing anything that works for you. Failure is normal, both at work and in personal projects. I failed to get my ideas across here more than several times, and I don't expect that everything I do or say will stick with others, even if it works for me brilliantly.

    That said, I think brevity helps, time is our most precious currency anyway. Second thing is at least trying to keep wide group as possible in mind. IMO TDM is hard to work with as it is, and not sure if you're noticed, but there are literally no art guys here, the only resident modeler here is Epi, while there is no shortage of developers, or people with IT background. If you take a look at other design threads, avoiding the pitfall of thinking like a dev is already a problem, and so TDM risks falling into even further obscurity of being for IT crowd only.

  8. IMO it's not really convenient, because for assets you typically already have safeguards like original Photoshop files with layers for textures, 3d scenes with models that have modifier stacks, etc. So you can get back to an earlier version of any assets whenever something goes wrong. Nobody cares about having tons of snapshots for textures or models, you just need the latest version. You don't need thorough reviews or careful integration with other people's work, diffing is meaningless because of file formats, etc.

    Learning the whole concept, framework, and tools is just tons of informational noise, in comparison to what you want to achieve.

  9. I'm on the phone, so I'll try to be brief, but in general I think it's an overkill, for multiple reasons.

    First off, I use git at work, and I had difficulties with understanding it back when I started. I do git KTs for junior colleagues nowadays, and they do find it difficult to grasp the concept; usually they need a week or two to get the hang of it. Remember we're talking about technically proficient folks, now imagine how much harder this would be for everyone else.

    I did take a look at the tutorial, and it's way too long, you might as well replace some of its sections with knowledge base links from git and atlassian websites, as they're shorter and they portion the knowledge better.

    Last but not least, I had a chat with my senior devs at work on that, and they all agreed that there's no point in having a git repo for assets, more like Google drive directory, or that server setup Unreal engine uses for realtime collaboration.

    I get that you like git very much, but it really is a tool for very specific tasks, especially considering the big picture. Since you already have trouble in explaining the concept in the simple terms, no wonder you get crickets or pushback as a response.

  10. Missing assets and bugs were one thing, but they also overpromised on features and systems (see Crowbcat video). IIRC it was in last month or two before launch, so they knew it won't make it to the final game. I don't mind seeing what's up in two years or so.

    • Like 1
  11. This is not a specular hotspot, this is a cubemap reflection. I don't like mixing those up. I like how this is done automatically in PBR setups, where one gradually turns into another with respective roughness values. In non-pbr workflow, I prefer to control both separately. Besides, they don't even react to light in the same way.

  12. By the way, it was a huge surprise to me that Arkane actually used fully dynamic lighting system in Dishonored 2. At the same time, they did have some "in-house indirect lighting system + IBL cubemap network". Maybe that could also be a way to go for TDM?

    Here's the presentation: https://www.slideshare.net/UmbraSoftware/umbra-ignite-2015-jrmy-virga-dishonored-2-rendering-engine-architecture-overview-moving-to-multitasking

    CC: @cabalistic

     

  13. So with this solution you're favoring mappers who use simple lighting, while this will look wrong/more pronounced with more complex setups. If you want a sort of fake GI helper, that should rather be a parametrised entity placeable inside DR, instead of a global solution.

  14. From the content creator perspective, it's already hard, in this non-pbr framework, to get consistent and predictable shading across your models and materials. Introducing another variable onto the mix, that is based on eyeballing the results with assets that don't even use the complete diffuse/specular/normal workflow, looks like just shooting in the dark and making things even harder. But as an option toggle for the players that is off by default – sure, why not.

    • Like 1
  15. 14 minutes ago, nbohr1more said:

    Did you try the mod?

    The screens were taken with the gamma a little high but in motion this does not look glossy.

    Not yet as I'm still on 2.08, but that amount of fresnel on the statues would suggest they're highly reflective, so if they're not glossy, they just look wrong. That said, the static screens look rather random. The last example with the comparison looks kinda like fake GI or ambient light bounces. But I suspect it will look completely broken in my environment, where every surface actually does use a specularmap, and sometimes specular + cubemap. Would have to install 2.09 to be sure.

  16. Just tried the Outriders demo, as the game is made by Polish developer People Can Fly. Unfortunately, it plays like a fairly clunky Gears of War clone + generic loot shooter lore and mechanics, most likely a hard pass to me.

  17. 1 hour ago, nbohr1more said:

    All surfaces have some fresnel response.

    "Some" is the keyword here, you added so much fresnel to those stone statues they look like glossy surfaces, and they don't even seem to have a specular map or specular + cubemap combo. IMO there's no point in testing that on materials that are diffuse + normalmap only, if you want to have even slightly meaningful results (as relatively meaningful they can be for the non-pbr world, so to speak). It's "everything is shiny" first, then "everything has some fresnel response" as a consequence.

  18. @HMart Thanks, I do a lot of work with models and materials, so I was excited about this, but I think I received the message. It's like ThiefGen, covering your ears and singing to mute any voices you don't like. Since he ignored all the initial ideas and designed the whole solution in one go, I assumed he doesn't really know how the iterative development works, because it shows. It's even worse, if this was already consulted with mappers and iterated upon. So, if it's mostly doing stuff for three friends and their dog, my presence in this or any future threads like that is pointless. I'll stick to the tools I already have.

×
×
  • Create New...