Jump to content


Photo

Possible Future of DR?


  • Please log in to reply
31 replies to this topic

#26 Epifire

Epifire

    Member

  • Active Developer
  • PipPip
  • 393 posts

Posted 14 May 2017 - 05:58 PM

Well I'm looking at the blend you're talking about Neons and I'm just not sure how they could implement any kind of painting with patches. I could see it working if the terrain patches in question could be turned into models inside DR, but that sounds like a nightmare operation in and of itself. Something that would be kinda interesting would be a height gradient blend mode. Where vertex z coords register create a heightmap (high areas getting a white coverage). It would likely establish the zeroed point of the selection to a center of origin, which could make gradient confinement unpredictable.

 

The proposed method would be something like an additional selection conversion type with the DR right click (kinda like set to static option). It would have several functions to handle that mode. Being a convert to model and a gradient based vertex color option handled in one click conversion. You probably would have to enable that as a finishing step, and would also have to setup a two stage vertex blend material in order to see the blend occur at run time. I doubt painting vertex colors in DR would be an easy feat which is why I suggest range based gradient. Possibly could have an adjustable center of origin to control gradient intensity?

 

That's the most realistic based operation I could see being implemented. Tell me your thoughts on this one guys?


Edited by Epifire, 14 May 2017 - 06:00 PM.

Needing dem environment models?

 

Speak to the EPI!

 

Or rummage my online mesh repository instead...  https://www.mediafire.com/folder/7q1r8him1y63d/TDM%20Repository


#27 Obsttorte

Obsttorte

    Scripting guru, Mapper

  • Active Developer
  • PipPipPipPipPip
  • 5153 posts

Posted 15 May 2017 - 02:07 AM

It is not really realistic as it suffers the same issue. If I understand you correct you want to use blend on non-models. This won't work without changing the map format.

 

The only way to use blend is on models, and with the DR ASE Blend exporter there is actually a way to do this in DR for over three years now (see here). The only downside is that you have to write the material file for it. However, this is explained in the articles and there are example materials there, which you can copy and just have to change the image paths for the diffuse etc... Should be doable even for a non-coder.

 

So the only advantage to this would be a script which does this automatically. It performs the blend and sets up the material. This is doable. But again, you can't avoid using models (and I don't see why this is a problem actually).


  • Epifire likes this
FM's: Builder Roads, Old Habits, Old Habits Rebuild
WIP's: Several. Although after playing Thief 4 I really wanna make a city mission.
Mapping and Scripting: Apples and Peaches
Sculptris Models and Tutorials: Obsttortes Models
My wiki articles: Obstipedia
Let's Map TDM YouTube playlist: ObstlerTube
Texture Blending in DR: DR ASE Blend Exporter

End of shameless self promotion.

#28 Epifire

Epifire

    Member

  • Active Developer
  • PipPip
  • 393 posts

Posted 15 May 2017 - 04:13 AM

It is not really realistic as it suffers the same issue. If I understand you correct you want to use blend on non-models. This won't work without changing the map format.

 

My proposed solution is a script that would convert the patch to a model and then establish a generic modifier on the finished shape (with a script). What really makes me curious is could vertex color data be authored by a non-modeler app? Because you'd be applying black and white vertex color conversion from bounding measurements along a single axis. Otherwise you're stuck trying to implement a whole vertex painting UI and toolset which just seems like an absurd undertaking.

 

Running with that design concept you'd create the said patch and when finished with the shape you'd convert the patch with the new script. The only other thing that would have to be defined somehow is axis selection and an invert color option. Otherwise I think we're starting to see the same point.


Needing dem environment models?

 

Speak to the EPI!

 

Or rummage my online mesh repository instead...  https://www.mediafire.com/folder/7q1r8him1y63d/TDM%20Repository


#29 Obsttorte

Obsttorte

    Scripting guru, Mapper

  • Active Developer
  • PipPipPipPipPip
  • 5153 posts

Posted 15 May 2017 - 04:49 AM

My idea was, that at bounderies where patches with different materials applied meat blending is applied. I guess this would be relatively easy if it would be restricted to the usage of two materials, although more are generally possible. The major restriction with the approach I have in mind is that the vertices have to meet up. This is not an big issue when using patches, but mappers using it would to have to keep this in mind. Letting go of this restriction would require way more math.

 

However, implementing it would only make sense if I would assume that it would get used afterwards. People where already praising me for the DR Blend Exporter but I don't know if it got used at all.


FM's: Builder Roads, Old Habits, Old Habits Rebuild
WIP's: Several. Although after playing Thief 4 I really wanna make a city mission.
Mapping and Scripting: Apples and Peaches
Sculptris Models and Tutorials: Obsttortes Models
My wiki articles: Obstipedia
Let's Map TDM YouTube playlist: ObstlerTube
Texture Blending in DR: DR ASE Blend Exporter

End of shameless self promotion.

#30 NeonsStyle

NeonsStyle

    Advanced Member

  • Member
  • PipPipPip
  • 1249 posts

Posted 17 July 2017 - 06:09 AM

The way I was thinking, was two separate patches with different textures. Verts are pulled up and pushed down into the two patches so they

merge together in part. ie some the bottom has bumps that clip into the top patch, and it has dips that dip into the bottom patch. Then you

need a script to calculate which verts are clipping into another patch, and blend those. I'm no coder, but it doesn't seem that difficult that

the game could use the vert height/depth as a method to blend them.

 

I'll read through that link Obsttorte, and check it out. In the map I'm making, it's a very old area, with walls

all cracked, with plaster mission, blended into brich and vice versa. If it was just one wall, I'd forget it, but

the whole place is like this, and it will add so much to the immersion of the level if we can find a way to do this. 

 

The other thing I'd like to see is us dropping the use of .ROQ files and adopt a more modern format of .mp4 for

video files. It would make it much easier for us to use video intros. I think it's time we dumped this antiquated format,

not only because it's difficult to make videos in it, but it's also crappy quality. 


I have a small YouTube channel making videos on a variety of games. Come and have look here:
 
https://www.youtube.com/c/NeonsStyleHD

 

 

 


#31 Obsttorte

Obsttorte

    Scripting guru, Mapper

  • Active Developer
  • PipPipPipPipPip
  • 5153 posts

Posted 17 July 2017 - 06:58 AM


I'm no coder, but it doesn't seem that difficult that the game could use the vert height/depth as a method to blend them.

We both have a very different interpretation of the term 'difficult'. ;)


I'll read through that link Obsttorte, and check it out. In the map I'm making, it's a very old area, with walls

all cracked, with plaster mission, blended into brich and vice versa. If it was just one wall, I'd forget it, but

the whole place is like this, and it will add so much to the immersion of the level if we can find a way to do this.

This can be done with the current setup. It requires some time been put in, of course, but should be worth it considering the result.


The other thing I'd like to see is us dropping the use of .ROQ files and adopt a more modern format of .mp4 for video files.

This is been worked on. Don't know the current status, though.


FM's: Builder Roads, Old Habits, Old Habits Rebuild
WIP's: Several. Although after playing Thief 4 I really wanna make a city mission.
Mapping and Scripting: Apples and Peaches
Sculptris Models and Tutorials: Obsttortes Models
My wiki articles: Obstipedia
Let's Map TDM YouTube playlist: ObstlerTube
Texture Blending in DR: DR ASE Blend Exporter

End of shameless self promotion.

#32 teh_saccade

teh_saccade

    Member

  • Member
  • PipPip
  • 482 posts

Posted 17 July 2017 - 12:20 PM

Hey - I've a better idea.

Why not just scrap everything and rebuild TDM in UE4..?

 

Neon/Gillette-the-best-a-man-can-get-logo-that-I-was-drawing-7-years-ago-for-synthwave-scene-that-is-now-popular-and-with-internet-tutorials-for-adobeCC - you're dealing with tech from a dozen years ago here.
It's like asking a bunch of shire-horses to pull a big-rig's trailer. Sure, it's possible, but it's gonna require as many horses as a semi.

Idk how many horses the TDM team has, but it appears to be dwindling.

It is possible to do many of the things that you have mentioned in this thread, but in a very round-about and fiddly, time-consuming way.
eg, the texture transitions might be done via transparent decals, the "painting" of grasses and like to be done through seed, which requires you to make your own decals, define them, get them working in DR, place them in your map, etc... You are basically a one-man games studio, on the backbone of what the TDM has achieved.

It's learning design tricks and techniques, such as, break a hard shadow by use of placing shadow casting foliage - no-one likes a hard line. Texture transition in mesh, choose complimentary textures (colour, pattern, form, etc...) or disguise the line with object, particle or distraction - or, realise that the player is probably going to spend about 2 seconds in that area and doesn't have the eye to collect such information as their mind is immersed in the mission, not the one area of z-fighting in a cutscene in Assassin's Creed 2 or why Ezio's sword is able to go through the ground when he assassinates someone, or why a body glitches and flies up into the air when you drop it on a rooftop of a certain angle, that - unfortunately - wasn't black-box tested on that particular module because it was added so late, in order to break up the monotony of certain city sections and facilitate a few missions/race events...

Like we'd have to in ye olde dayef - resort to creative problem solving in order to overcome the limitations of our/the engine's intelligence and design.

Idk what the problem with .roq is - there's encoding and decoding for videos that works the same and fits in with the engine no problems. It's nothing but an additional stage to the pipeline.
From the looks of things, it requires a rewrite of the whole GUI for TDM in order to insert another format... not to mention resolution, the fact that .mp4 users will probably want the audio embedded and all that malarky.

To incorporate modern features would require porting TDM to a modern engine, I'd reckon.
This is how things were done, "in ye olde dayef" - back when games were more awesome than watching the return of lara croft's physically perfect, pre-rendered by 3-frames, "got the bounce" arse for 10 hours on some pointless mission to collect feathers for an arrow that you don't even need.

It took fkn months to get the fabric working for a cape and nUbisoft still can't get hair right - due to limitations in the engine (is anyone working on it..? no, because franchi$e is more important and the next engine might fix it, if there's dosh to fund its development).

Dunia engine won't even work properly if the machine renders faster than 60FPS, because - at the time - if you could run it at 60FPS you were probably about 7 years in the future when it came to processing power and it was to be released on schedule, cos bosses said so.

 

// http://bugs.thedarkm...iew.php?id=4493 <--- like this, except it also breaks AI and a whole bunch of other shit.

DR, compared to the "cutting edge" mapping tools available to general public is a lot more like, "handmade" than "click the buttons in this dialogue" - EVERYTHING must be done pretty much by hand (I don't even use the readables editor) and that, to me, is what makes it a special case - the skills required in order to produce any FM for a game such as TDM are far beyond that required to map using, eg, AnvilNext or some shite, where it's all modular and you have a hundred other people helping you - it's all solo and:

No-one's being paid for this.

My going rate is between £30-60 per part hour. That means, for the several hundred hours I've spent in DR and TDM, testing - It's probably akin to a year's wages - if anyone was willing and able to do that kinda stuff, likely they'd be getting paid for it and do this as a hobby. Dogs will hunt.

Half my taskbar is taken up with tools required to work specifically with TDM. There's a whole hard disc devoted to it.

Nb, I've not a single mission released in the list, despite having made several and ten times as many tests - I think perhaps only 2-3 people in this forum have ever played any mission I've made / in progress (idk if I will ever finish "the one", because I'm fed up with it not working and having to rebuild it every time there's a change that affects some mechanics or architecture or whatever.).

--------

One thing about TDM is that, up until a few years ago, its growth was sustainable and it had many active and enthusiastic team members.
In the past... year or so - TDM's profile and growth have started to outstrip its ability to keep up.

(remember the massive hissy fit I had when movers suddenly no longer caused damage..? It totally fucked entire sections of my levels - as did the no-draw surface tricks and many other things... I'd not receive replies to questions I'd ask because, usually, people wouldn't know the answers and I'd have to find them myself - using internet time machine for 404 sites a decade ago - It was possible to "workaround" them... Idk if you were here for that...)

In effect - the tools to create and the game itself change more quickly than it is possible to produce an FM of decent quality (IMO, only), that isn't merely a collection of prefabs, slung together into a mansion with some keys and some premise as to why you're running around performing a limited set of objectives.
Mind, probably doesn't help that I refuse to use architectural prefabs, cos I'm the kind of guy who recognises a typeface being used more than once in an modern advert that I used once, 5-6 years ago.

Check out the bugtracker: http://bugs.thedarkm...y_view_page.php

Why be expending so much effort on new features that will be bugged to hell for ages, when there are still so many residual bugs and glitches (and you've seen a few of my videos...), when there is still much QC to be performed on the "finished product"..?

------

A master adapts their tools to fit the situation, not try to alter the situation to fit their limited toolset. You can drill a hole with a hammer and nail, and sink a screw with a chisel. Don't need an electric screwdriver and powerdrill.

Many things are possible in TDM/DR - it is very robust and versatile - but it takes a fucking long time and a lot of effort and planning, and learning the game and how it works inside-out.

It's an exercise in problem solving and it can be bloody difficult at times, but therein lies the reward.

Despite never really finishing with TDM - it taught me so much about what's required (to analgise this to modern production work-flow), that it enabled a step up the ladder.

(also another reason as to why I'll probably never finish "the one")


Edited by teh_saccade, 17 July 2017 - 01:29 PM.





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users