[Role Playing Game] Memento Mori

[Talavaj]

[Bump]
But if I think about it how's that going to work if the modulate will cover the holes?
Shit!

Lol, yeah. That is a bit of a problem. Well it shouldn't be much of a problem if you have simpler shadows where it won't really show. (because if you cover the entire hole with a shadow it should look correct)
It just may end up wonky on the edges of the shadows.

Also I see you did a lot of progress there. The bump looks even better with parallax.
I also love the hard shell model.

I'll update later, off to sleep for now !

 

[tobyfat50]

So yeah, believe it or not, the texture is the same for all materials. This is what happens when you have all those layers to blend together in different ways. I am actually surprised myself that the hard shell for example turned out like that.

I think the shadow maps actually work with fake parallax but you have 2 options; one is using a top layer shadow map as you said but you'll have to be satisfied with its flaws or make it look good somehow. The other would be adding shadow maps to all layers which actually looks like the real thing as the shadows are placed on the vertical surfaces as well. Almost got it right last night but the problem that remained was to add the priority planes wich I don't know how in neo dex so I'll have to use art tools. Almost certain that it will work as blending the modulate with the other layers worked for the parallax + bump. If it works I'll send you a sample.

One more thing, adding shadow maps to all those layers will reduce performance. Bummer!

 

[Talavaj]

Actually another problem is that stacked modulate is like guaranteed to end up pure black. At least that was what happened in my test for some reason.

Scratch that ! Actually it looks like it ends up just fine. It behaves a bit weird with priority 1 but it looks perfectly fine (as if it is in the bottom of the holes).
Without priority it doesn't work, as suspected.
That is using only one layer on top as a separate model.

And you are right, using blend is just so much nicer, and it doesn't even bug up with priority planes.
Here is a high quality parallax texture for a pedestrian crossing.

And here next to a flat texture without parallax. On a chunk model of a crossing on Broadway 500 S. A hyper large amount of detail using only 74 triangles, although a bit taxing at the texture size.

The current concept of the area is a large flat clearing with lots of ground foliage and a hole filled with rain water with a single car wreck sitting in the middle.
The area will be considerably dark however because of the four tall buildings sitting on every corner and a dense canopy of trees growing from their decrepit walls. There will be little but intense light shining through the few holes illuminating bright spots on the ground.
It is fairly safe and highly resourceful spot, with the 4 buildings around housing 7 Eleven, a pharmacy, food mart and a clothing store with an above ground level full of apartments to search.
It is the main hub of the first level and the very first spot to take shelter at.

And here is the crossing in real life, you can see the big and small tiling from the textures in there, obviously it is not super accurate as I'm limited by the texture palette. But yeah, a lot of the models and doodads I model nearly identical.
Even their placement is gonna be close to how they are in real life L.A ! (of course unless things changed since the google earth pics were taken) and of course in the mod most of the environment is going to be completely ruined.
I still ask myself whether I will keep the real life naming or make up my own, but considering it is just so blatantly obvious what things are modeled after (plus I keep mentioning it here often) I may as well keep it. Due to the nature of the mod I could see people getting offended shall I accidentally draw some actual connection to real people (which is not intended as the mod is purely a work of fiction) but I doubt anybody would ever take any legal actions against me for a mod somewhere on a wc3 modding site.

 

[tobyfat50]

Actually another problem is that stacked modulate is like guaranteed to end up pure black. At least that was what happened in my test for some reason.

I had this problem at the beginning as well, I'm glad you solved the problem.

but I doubt anybody would ever take any legal actions against me for a mod somewhere on a wc3 modding site.

I'll drink to that if it actually happens!

Today I went to a friend to see his computer perform in max quality on the newest games and I realized how much crap my stuff actually is. I tend to do everything rushed and not put so much heart in to it because I want to finish this as soon as I can, one of the reasons my latest scenes are starting to look like crap. Why the barren turned out like crap, because I'm starting to put quantity over quality but it's so hard to do this alone...
Speaking of real life I came across this videos at my friens place which turned my reality upside down. At the end of the vizionary I realized how shit "poligon technology" *(don't hate me for saying this people) is, I had a feeling I was watching tetris. Of Course this was just a fleeting feeling at the moment.
It's hard to trust what this guys are claiming but it seems real.
Check this out if you didn't know about it before, might be old news, or not.

 

[BallisticTerrain]

I actually saw that first video a few days ago. I'm a little skeptical about it, as none of the big GPU manufacturers have even recognized it yet to my knowledge, and also it's going to be hard to 3D scan a dragon. Personally, I'm more excited about opensubdiv from Pixar. Watch these, it's really cool.

http://www.youtube.com/watch?v=xFZazwvYc5o

http://www.youtube.com/watch?v=Y-3L9BOTEtw

 

[Talavaj]

Unlimited detail (as they proposed) will never happen, voxel based visuals handle lighting and animation very poor.
They also can't have proper material shaders. Plus there are many more practical reasons against it.
Such as can you imagine how much more it would cost to model in that detail and how much longer it would take ?
Scanning of real life objects and converting them would work but it would only get you as far.

It is just messy altogether. Besides the entire idea is pointless because we are slowly getting to the extreme level of detail (that matters to the viewer relative from the distance the geometry is viewed from) using modern tessellation coupled with something like vector displacement.

Secondly a world where GPUs don't matter is not feasible, there is too much money in the business.
Nobody from "the big GPU manufacturers" would care because voxel based graphics could be rendered without the big GPUs, similarly MS would not acknowledge it either because of DX.
Both have gargantuan market share, it is just not gonna happen.

It is a nice university project I suppose, though It has no future in gaming IMO. Although the geoverse looks cool, I would love to see google earth in such quality. Also heh, seems like they themselves changed their mind about gaming applications.
They should just stick with geoverse or whatever similar stuff for geovisualization, I don't see it being useful for much else.

 

[tobyfat50]

I have to Agree, this got me hyped at first. It seems that they might of gave up on the gaming concept and went to what they might do best on the market. Although it seemed like a nice what if at first. I gues sticking to the shity poligons is still best for us artists, what would you want more. (The universe if possible please!)

Secondly a world where GPUs don't matter is not feasible, there is too much money in the business.
Nobody from "the big GPU manufacturers" would care because voxel based graphics could be rendered without the big GPUs, similarly MS would not acknowledge it either because of DX.
Both have gargantuan market share, it is just not gonna happen.

No wonder no gaming company goes near. Made some research on it and it seems to be true.
Bah, I'm bloody disappointed, next time I post something that intrigues me I should research about it before posting. *Mostly disappointed* I'm not very interested in global positioning that looks nice on "google" or space aged cell phones for that matter. Still, in the end every discovery might have its importance. You might not know what will pop out from one concept and it might be way different from it...

BallisticTerrain, yeah I already seen one of those videos when I came across the unlimited detail thing.

 

[Talavaj]

Once again, was too busy with work this week to make any significant progress. Most street doodads are done however and street tiles can start to be placed down.

Here is also a concept render of the main street crossing, although the level of deterioration is still too small.
It very likely will end up with more foliage, rubble and broken tiles.

 

[tobyfat50]

You're getting closer to the real thing. There's one thing I wanted to ask you for a long time. How do you fragment your models? Do you simply use boolean or something else. I have seen you make that shark in Zbrush but it's just recently that I started to use it myself.

 

[Talavaj]

With booleans, and more recently with Houdini FX. It has voronoi fracture.
It is much better with physics, fluids and animations than Max, or at least the default offering is much better than that of Max when it comes to those features. Though I wouldn't dare to use it for modeling lol.

The shark is made in mudbox actually. I got it mainly for the texture paint which IMO is a lot better than that in Zbrush.

 

[Wazzz]

you wouldn't happen to be a hammerfall fan, would you?

 

[Talavaj]

As in the metal band ? Not really, I'm more of a "whatever sounds good" kind of fan. Though I do prefer older stuff.

By the way, would you want a "listen mode" (shamelessly ripped out of the last of us) or would you prefer not having one given it is a predominantly survival mod ?
It would be a feature that can detect enemies through opaque environment (much like the detect life/dead in skyrim), you could only use it while crouching still and it would temporarily "ping" enemies around you if they were behind an opaque environment.
I thought it would be quite a sensible feature given there is a lot of emphasis on stealth and going stealth blindly may not be very useful but I'm worried whether it would be too "cheap" or wash away the tension.

 

[BallisticTerrain]

I wouldn't want a listen mode. It didn't make sense in the last of us and it won't make sense here. It would seem way to fake and easily locating enemies like that isn't really ideal for a survival horror game. I hardly ever used listen the last of us except once, wasn't really all that useful. Things like that are for stealth games like dishonored and splinter cell were there is a way to explain it. (Dishonored mark of the outsider with void gaze, splinter cell the night/sonar vision goggles.)

There has to be a better way to force that cool mesh xray thing into the mod.

Zbrush advertisement begins now.
Also, I am deeply and unnecessarily offended you choose mudbox over zbrush. Zbrush is much more feature rich and you can actually build a model from the ground up in it, and do everything except for animation. With several of my recent models I haven't even touched 3ds Max except for a few UV maps, just used Zbrush and XNormal. That added to the fact Zbrush can export in tons of file formats makes it much more worth it. I tried the student version of mudbox before I got Zbrush.... it has an easier to learn interface, but that is literally it. My biggest complaint with mudbox is how scewy models are when you import them from other programs. I could not get one max model in mudbox without an error, plus, the scuplting has a tendency to seriously screw over the topology. Zbrush does not have this problem, and even has three different ways to fix it. (Retop Brush, Dynamesh, ZRemesher) Zbrush even destroys mudbox in the convenience department, allowing you to send models back and forth between max and Zbrush with a single button click. Literally, all you do is press it. You don't even have to press a "yes I do actually want to do this" button.
Zbrush advertisement ends now.

 

[Drunken_Jackal]

Yeah, listen mode should be removed. As well as things such as looking behind walls and the such. It removes some of the stealth essential in a survival game/mod. The player should just rely on their line of sight for this, along with avoiding the line of sight of the enemy.

 

[Talavaj]

xD ok, no listen mode then. Well the kind of explanation behind it in TLOU is that you can "hear" the enemies though indeed it was a very gamey feature in otherwise a fairly pragmatic game.
It is a bit difficult to make a proper stealth mechanic without some obvious elements though, like those displaying the FOV of the enemy or acquisition range or level of alertness while at the same time being as natural and minimalistic as possible.

Though you will be quite obviously tell what your situation is, as all things that generate threat in stealth are either audible, if you heard a sound other than hud sounds it is 99% likely that it reflected in noise level.
If you notice your character appear darker you are quite obviously in a shadow and as you can guess not moving and being in crouch keeps both the noise and sight levels low.
One thing that might be less "cheap" way of detecting enemies is that monsters will emit sounds all the time, (which is after all scarier than not seeing enemies at all) but you still won't be able to detect them properly I guess ? as I reckon wc3 doesn't really use 3D sounds it only makes sounds louder when they are further away ? or does it like use actual 3d sound ?

Also, Zbrush is definitely a more complete package than mudbox. But I already have a workflow in max and I don't need a self sufficient program. I only really got mudbox because of the texture paint. (and the poser tools I kinda like those too)
Though the main advantage of Zbrush IMO (because the others are really kinda cosmetic as you can make production quality sculpts with either) is the technology it uses to draw the mesh with which is a lot like voxel allowing for meshes to be displayed on machines that would melt down when viewing the same level of detail in mudbox which uses simple old polies and as such is generally lot less performance efficient.

Also the models that you import into mudbox bring up errors all the time because the software requires a very good topology for proper subdivisions (it will absolutely hate you if you don't use quads). So yeah for that reason it is pretty obvious that I wouldn't recommend it to people without a proper 3d modelling software and a considerable experience with polygonal modeling.

 

[tobyfat50]

Actually I've been switching from Zbrush to Sculptris. It's a really simple version with only some basic tools and yet The paint tools, geometri and uv paint are all I need, not to mention that it's a lot gentler on my machine, for everything else there's 3ds max. Obviously, I'm making my stuff on a less performance computer but stuff like this calls for a new one.

 

[Talavaj]

Yeah, I'm actually thinking of incorporating mudbox to sculpt the streets. I have spent exhausting amount of effort to figure out how to make them the way I want but I ended up with a "meh" feeling that makes me think it is not what I wanted it to be (which is really disheartening).
I want people to think "what the actual fuck!" when they first get to the game and I personally am not feeling it as of yet which makes me redesign a lot of what I made before.
If you remember the building model I have shown before or the concept room, none of those assets actually made it any further because I deemed them to be sucky xD.

I also want to edit the OP (because some things listed in there are no longer true and different features were added) but I feel just changing a bits of it now will be pointless.
I want to wipe it out entirely and replace with a completely new description which will be more compact, easier to navigate through and nicer. (less text sauce more plain information)

So ywah ! You will like it hopefully ! It is neat though, although the many setbacks waste a lot of time I feel they helping the overall quality. I have been researching a lot and went through and absurd amount of reference gathering such as watching walkthroughs/playing things like TLOU/Crysis 3/I am alive and taking screens of every inch of those and of the google earth (lol!) to get to the point I am at now.

Which is this. As you can see by the date created, the whole brainstorming process started nearly a year before this project was released in this forum.

 

[RiotZ]

Do you think with the assets you've added to the engine, you could make a better, visually more complex multiplayer version of WC3?

 

[muzzel]

Nice screens, can i rent your modeller guy?

 

[Talavaj]

Nice screens, can i rent your modeller guy?

Afraid not, my modeller guy (me) has a ton of things to do as it is.

As for multiplayer, I don't want to dabble into it personally. Though my stone & sword pack is developed with multiplayer application in mind.

 

[tobyfat50]

Can't wait for this to blow my jaw man, I know that you still got a lot of awesome stuff there that you didn't show us and probably won't. I, myself only posted less than 1% from my entire projects as I don't want to spoil too much of it and want to keep it as much as a surprise as I can till the release date. And yes, I also find myself remaking stuff time and time again as I don't feel that it will makes enough of a impact.

 

[BallisticTerrain]

I just realized that my new computer does not have warcraft 3 installed, or frozen throne, or neodex.
I should probably take care of that. I'm working on a different project right now that doesn't have to do with warcraft, but like I've said before, if anyone needs me as a spare modeler, just ask.

 

[Talavaj]

Yeah, the most of the stuff I show is merely the concept art. I don't want to show the actual content until I'm 100% sure it is final.

Also I found the game xD you mentioned before. It is called Enslaved: Odyssey to the West.
It is a bit too fantasy, set in a very post-apo America. The level design is great but a bit too deteriorated.
I'm aiming for something in between this and the last of us with less open areas to conserve space because of how limited the wc3 maps are in size.
TLOU, with more foliage that is about it.

Because of how limiting the map size actually is, I started playing with an idea of making some parts of it instanced.
This would mean "super models" all stacked in one spot on the map and their render state turned off for all but the active one. They would share a couple of pathing templates depending on what the "dungeon" is. And content within the "dungeon" would generate randomly, these zones would be accessible "scavenging trips" that you could go into from your hideout by selecting the part of the town you want to scavenge into and you would be given information about what kind of threat level the area has, what kind of resources you can find there and about how much time and stats it would take to scavenge there.
When you choose one place it will generate a random instanced area of some template and the objective/situation would warry.
They would be set in areas not normally accessible in the open world such as floors in apartment buildings and the general objective would be to scavenge the floor while avoiding enemies if any. "Failed" scavenges could for example be that you would be loaded into middle of an encounter with a boss level creature that you would have to escape from or into an encounter where you are surrounded by enemies and would have to fight through.
Those are the ones I thought of so far and I'm free to whatever suggestions you people have for any suitable events that could occur !

I have also been watching TLOU bit by bit for inspiration on props and level design and stuff like that. The amount of detail in that game is just mind boggling. I wish I could come at least somewhere near that but it is truly daunting.

 

[BallisticTerrain]

Yeah, TLOU has incredible graphics for the PS3 and they obviously put tons of time into it. The environment is pretty interactive to. Its near impossible to do the same thing alone in your free time, as naughty dog had a big modeling team working full time to make all the art. Plus they had dedicated concept artists to work off.

 

[Talavaj]

-

I wouldn't say the graphics were amazing, at least not from a technical standpoint.
Though the fact makes it even more impressive because it did indeed look good.

The level design and direction though did it, and the mind boggling amount of texture/asset variation. I could swear I could not tell if there was any art reuse in the game.
Obviously, the multiplayer looks worse in comparison.

And that is precisely, the reason why some of my art pisses me off. Even though I often tell myself "I can model that" I still fall short for some reason.
Working on it though, I have been literally dissecting the MP maps and tightly waiting for the map-pack dlc to come out.
All fine though, I do believe there is potential for MM to look somewhat close.
Though I guess you will be the judges after the gameplay trailer is out. Hoping to gather a bunch of feedback too, both on the presentation and the mechanics.

 

[tobyfat50]

That looks awesome!

So you're making a trailer then? I need to get my popcorn ready.

 

[RiotZ]

I didn't really notice the asset usage in TLOU but, I was really, really impressed with the lighting and mostly the level design.

 

[Talavaj]

So you're making a trailer then? I need to get my popcorn ready.

A short gameplay footage ya. I wanted a trailer, but it has a little point, especially because I do not have enough content to make a proper trailer. I may make one nearing the first release later though.

Also, weekend has come finally ! and I can work on some stuff.
I'm finalizing the protagonist for the show-off, his attire was changed slightly and he was fitted with extra detail accessory fit for survival, he also wears glasses now.
They reflect according to how he moves, lol (one of those things nobody will notice) the same way that the water animation I showed earlier works.
And also have been brainstorming over the equipment, since there are virtually no stats (though the perk grid has similar effects but they are all under the hood) most of the equipment will only expand on depth.
But it is fun thinking up weird makeshift stuff like barbed wire armguards, dust scarves, makeshift pouches attached with belt onto the thigh (that expand the hot slots) and similar stuff.
A lot of of the detail equipment stuff is inspired after TLOU multiplayer too.
Been also checking out Survarium which looks better than ever. One of the levels in there has a piece of design that literally blew my mind. It is an anomaly that spilled out of a chemical tank and turned into a gooey formation that appears to have created a reef like environment inside where fishes swim around. (if I saw it right)

I like small immersive detail work, hoping people will notice it though.

Below is how the reflection looks like, I know the size and position of the glasses isn't quite right lol, though fitting geosets is easy thanks to how Manoo modeled the character I can model and fit attachments on an inanimate replica in max and they transfer perfectly when attached to the appropriate point on the model which allows for some very specific equipment attachments in game.
The texture animation is synced to body movement in every sequence and the wrap is distorted using vertex color brush so it appears more glassy and refractive.

Also, as seen on the glasses I began to use this "cell shaded" material to fake metallic/shiny fresnel. Using a grey/black gradient wrapped in a way that highlights the top and darkens the bottom edges.

 

[tobyfat50]

Omg, that is so scary it's awesome! The details you call small definitely have a big impact on the game. I think that I can make a very pretentious friend of mine like your mod and he hates Warcraft.

 

[FUJI]

Nice models they are very highly in qualities =))! Interesting project of yours ^^!

 

[Talavaj]

Hey hey, how are you doing everybody ? Finally, I have got a good feeling for once !
I made some neat changes to the chapter design and narrative, I'm not a story-in-games person but I feel good about this one.
That decision also resulted in some exciting changes that you may hear about soon enough !
p.s the demo prologue level is sexy as fuck. I mean, what ?

Also, hope you are liking the new logo !

 

[BallisticTerrain]

Ya, I love when I get an awesome feeling about my work. It's usually a good indication things are going good. Can't wait for the beta.

About the logo, it's better then the last one, but I think you should make one more characteristic of your project.

 

[Talavaj]

Yeah, things going pretty well. There may actually be a prologue which will be like a tutorial chapter.
Initially I intended the tutorial sequence to be within the first chapter but I just didn't have enough space to put it in. (because the backdrops are HUMONGOUS, bigger than the actual playable area)
I have some doubts about making it the actual first chapter area, there is a lot of effectively used space but I do not want to make it too long because I plan to wipe the player's character progression afterwards (because story related reasons).
It was intended as a tutorial area that would explore all the features then throw the player out after its completion. Lol, I don't know yet.

Current issue is with normals on the "super models" being absolutely effed up. It is not that apparent on smaller models but when you have a thing that represents a couple dozen meters it becomes jarringly obvious.
But it is not that much of a problem anyway, Ill probably bake a complete shadow map on them and just leave them unshaded. (besides if people also use ambient occlusion from the included ENB it will be perfectly fine)

Though it does show that the native smooth shading does actually do a lot. (although it is kinda too faint in my opinion)
Another annoying thing is that I have insufficient reference materials for things like destroyed objects.
eg. what exactly is below the street ? what exactly is the core of this shattered wall ? etc.

Anyways, I have very good feeling overall ! The level design, narrative and everything fits very well so far. (at least for the first couple chapters).

Also here is an example of the shading being awkward to say the least. The big dark area in front shouldn't be there as it is just a flat street.
This actual section of the model is out of the playable area hence the low-polyness.

And to give you the general idea of how huge the models actually are.
This is one half of the tutorial's chunk model, the yellow model is the protagonist.
Though, most of the model is not a playable area and backdrops extend beyond the map's extents.

 

[tobyfat50]

Always refreshing to check your thread. I'm glad that you had a inspirational moment and I know it feels great when you see your creation "more alive" than ever. I for one am out of inspirational energy. For now, in a experimental state again, I'm trying to make diffuse textures as convincing as can but I'm still having trouble finding a way to reconcile them with the bump and specular, too many variations I can choose from not one that I actually consider good.

The logo looks nice but something's still missing, doesn't quite match the stile, probably addings some gears or something could fix that?!

Are you having problems with disepiring models in the far out extents?

 

[Talavaj]

Initially I wanted to use a skysphere that would have the backdrop models in it. But because of the way sky works in the game I scrapped that idea.

The only disappearing issue I have is with walkable platforms which seem to disappear no matter how big the extents are, so I have to make a unique platform with full transparent material on top of the street (which is UBER pain in the ass to align right by the way)
Other than that the "super models" work fine and display correctly for the most part.
The street models themselves are fairly low poly though (only the holes add a lot of geometry) the buildings are a bit higher, but they lose detail on their way up.
Such as the apartments at the bottom have parallax and modeled railings, whereas the detail higher up is just a flat texture (usually print screened model lol).
There is a lot of similar optimization stuff, water for example is very poly dense only close to the playable area and the polies slowly disappear into the distance until it eventually becomes a flat texture because you won't see shit that far anyway.

The only problem limiting the scale is the pathetically short clipping distance of the game which appears to be capped at 10k no matter what I tried. Any idea if there is any way to bypass that ?

 

[tobyfat50]

So you're actually using a large far clipping then?! Sorry but I tried everything here as well, especially messing with the models and material properties. Have you tried a imported camera model? You can change more native properties like depth of field, Far clipping and Near clipping (Near C - Seems to be changeable but it doesn't work under a value of 10?) far beyond the acceptable limit, especially the field of view so much that it becomes like looking through a distorted mirror. These things can be scripted as well I think.
Hope I understood the question well!

 

[Talavaj]

I'd want as big far clipping as possible, yes. I currently use 10,000 and it doesn't seem to go past that using triggers. Lol, and when we are at it I would also like to turn off the hilariously unoptimized terrain (the way it can be done in the editor) that would be great but that probably cannot be done.
Most likely the maximum you can do is change the tileset to something like 2x2 pixel sized texture to minimize the impact on performance.

Pretty sure that the massive performance loss while in high drawing distance aren't the models themselves but the terrain tile, which is always at full resolution. (no texture filtering)
It is fine from the native camera angle but it is an overkill from any other camera angle. A 256x256 map has a terrain mesh that is like 131k polygons minimum (probably much higher if there are seams and material layers) and 65,536 tiles of unoptimized 64x64 pix textures all over the place which is not only insane load but also a mega brutal aliasing galore.

 

[tobyfat50]

For all that I think you need some kind of a hack, wish it would be possible with simple tricks. If you can't fix a problem, move on, it will cost you a lot of time maybe later you'll find answers doing somethin else.

Ok, Just let me try something!

 

[BallisticTerrain]

If only there was some way to completely remove the terrain mesh and instead use nothing but models.
I bet there is. Hmmmm.

 

[Rao Dao Zao]

If only there was some way to completely remove the terrain mesh and instead use nothing but models.
I bet there is. Hmmmm.

Use the alpha tile, raise all your walkable destructibles off the ground plane/lower the invisible terrain beneath them?

 

[RiotZ]

If only there was some way to completely remove the terrain mesh and instead use nothing but models.
I bet there is. Hmmmm.

Yup.

Use the alpha tile, raise all your walkable destructibles off the ground plane/lower the invisible terrain beneath them?

This.

It's a pain in the ass to work with alpha tile though, fair warning.

 

[Deleted member 219079]

oh god your signature xD

 

[Talavaj]

Use the alpha tile, raise all your walkable destructibles off the ground plane/lower the invisible terrain beneath them?

That is how the level is done at the moment, it however doesn't make any difference as the whole mesh is still rendered and there are thousands of polygons underneath everything that are completely useless and draining performance.

Also here is a concept of how I would like the water to look like. (the lens and the bokeh are added in photoshop for the picture) I have shown the mesh animated previously if you want to get the idea of how the water flows.
It is tessellated towards the camera and flat in the distance. It is overlaid with another layer which is an additive texture of the background for reflection which has rows of vertices animated using bones to distort the surface as if the reflection is refracted.
The lens flare would be a billboarded model like the one I made some time in the past and the bokeh glitters would be made using particle emitters.

Obviously, it won't look as good as the shopped image but the native bokeh and DOF of ENB may make it look similar.
Hows about that ? Or is it too fancy ?
This is of course, the background water. It would only work in controlled camera angle where the player cannot get any closer.

 

[tobyfat50]

That is how the level is done at the moment, it however doesn't make any difference as the whole mesh is still rendered and there are thousands of polygons underneath everything that are completely useless and draining performance.

That must really be a pain to handle!

Ok, there are some things you can try like Terrain 1.0. It can generate terrain from height maps, I used this to make the wow terrains. It works by replacing the w3j file or something I don't remember, haven't used this for a long time. If it can do that maybe it can also completely remove the terrain or you could try that yourself but I don't see it working, will probably generate your a nice error.
The other thing would be Zépir's War3 Map Editor 1.0 wich I heard it can mess with the pathing and shadows also capable of drawing them on the minimap using a pen tool. Haven't tested this, maybe it doesn't even have a option to remove the terrain but it could be worth looking at.

As for the far clipping, try this sample map with cinematic cameras Camera Map. There are 2 camera models Camera_Mountain Scene01 with the far clipping 50000 and Camera F200 with field of view 200, you can add higher values if you want. Not sure about the far clipp as they even have near click options to but don't seem to work under 10 but the FOV works. You can export such cameras for 3ds max or create the in model editor. Open one in the editor to modify the values I'm actually using them in ep 2. The problem is that while activated this cameras are frozen so, I guess they were meant for cinematics only. To position one place a normal camera and it will play in that point but watch out, the facing is not the same as the regular cameras, it's something like from lateral of the camera which is annoying to figure (why I'm having a hard time with my cinematic). Maybe you can find some way, that's all I know!

Pff....

Sorry for having to use tusfiles! For some reason my slow internet doesn't work with pastebin at all, not even 1 kbit. Too lazy to move from tusfiles but I'll use 4shared or something eventually.

EDIT: The water is superb! There is no such thing as enough of something good! keep improving it!

 

[Systemfre1]

Surely the terrain has a file saved somewhere in the map MPQ.

Like how the lighting for maps has a file that you can set to nothing so the map ends up completely black.
Perhaps you can change the terrain to nothing?
But then terrain controls the pathing doesn't it? therefore it'd pretty much end up in a crash.

 

[tobyfat50]

Unfortunately That's what I think will happen, it seems objects stick to the terrain pathing. If there is no padding/no terrain, where will the objects go? Is it even possible to place them anywhere, nowhere within the editor? It would be awesome if it would work do.

 

[Systemfre1]

Indeed.
In the editor if you have an object with pathing, you still can't place a unit on there.
Only ingame do they stick to the object.
It has to be placed on the terrain :S so I guess thats that.
Would be cool if there was some sort of way around this, perhaps replacing it with a plane that has less polys?

 

[Talavaj]

It would only really require to change the render state of the terrain off like the editor does when you press ctrl+T, but I don't think there is any convenient way to do that in-game so yeah.

Also, the change of the logo does actually reflect the theme better now.
I think the heavy emphasis on the world's realism made me stray off the original path a little too much.
It was quite a bit ago that I realized that without being stylized the environment has no character at all and it is a bunch of objects that don't quite make sense or fit together at all.
The original listing was after all:

Litl - Memento Mori - announced sandbox survival horror in modern/surreal/religional setting with moral/mature/adult oriented themes.

and that is the kind of art direction I want to stay on.

Also, I'm home today and the weekend ! So some work is going to be done.

 

[Manoo]

oh it took some times from my last post . Any way I'm happy to see its still progressing. By the way nice logo, but I think it doesn't fit horror theme at all its more like fiery rpg stylish, it doesn't fit the name meaning, that's all :/. I recommend to reconsider your logo, but its my opinion .
By the way u should put less effort on trying to improve model graphics its already enough for wc3, because u won't gona finish it soon. Any way keep it going!

 

[Talavaj]

oh it took some times from my last post . Any way I'm happy to see its still progressing. By the way nice logo, but I think it doesn't fit horror theme at all its more like fiery rpg stylish, it doesn't fit the name meaning, that's all :/. I recommend to reconsider your logo, but its my opinion .

Lol, ok my gentlemen ! I heed your call and stylized the logotype further.
Don't worry though, no more graphix experimentation. I'm fairly sure I got the right direction now and it is going fairly well.
Actually so well I may make a chapter more than what I recently concluded.