LANDNAV - Version 0.01a available now

Hey everybody, some of you may know I'm working on a game called Dog Go Run. Well, I found a programmer I can afford hire for that game -- which is awesome -- but before I dive into a large scope project like that with a new person I wanted to try out a smaller game idea I've had for a long time. This game can probably be built in about three months, and then after that I can polish it while working to grow some marketing hype before releasing to the wild.

Anyway, it's a simulation of the land navigation training I did in the military. It's going to be 95% authentic. I tweak the rules of the scenario slightly to allow a little more player meta-game strategy and a little less uncertainty, but still, it is going to be a very demanding game. In some ways, almost diametrically opposed to mainstream video game design. So it will be interesting to see what kind of audience appreciates it.

I got some work to do before I get the programmer started, so in the meantime I am working on the few art assets we will need. There is only a single player character, his gear, and a 40km area in the backwoods of southeast united states. For the environment I will likely use a majority of purchased art.

The first thing we have to test is the actual land navigation using compass and map. There is no GPS, because GPS is for small crying babies. If you are going to learn to do land nav, you are going to learn to do it right. Or just quit, okay? I said this is military training. No whining!

So far I have made an accurate compass as well as notepad and pencil.


The last tool player needs is a map, which will be coming a bit later. The map has to be to scale with the 3d terrain. There is a few ways to do that, but I will figure out the best way once I return home in a few weeks. As authenticity is important here, I have to make sure I am using the correct measurement systems so that player can use the exact same methods as in real world. That makes things a little more complicated, but it's all doable even for a bozo like me. Just takes some patience.

 

(using compass to shoot an azimuth)

Whenever I play a new game, I write down the things I don't like about it. One thing that really bugs me is when a game does not teach you how to play it. That's not clever, nor does it make an enjoyable challenge. And a lot of times, once you do go through the pains of learning how the game works, then you discover it's not very nuanced or challenging. It's like when a girl leads you on and... ah well, you get the point. Lies!

I hope to do something better with this game. The fact is, learning how to conduct land navigation will take hours of step by step, hand-holding instruction. A robust tutorial will be essential, because how many gamers know how to read a military map or triangulate their position by shooting an azimuth?

But following a tutorial in this case shouldn't be dull. You are learning a real life skill. And you are going to have to apply that skill in a demanding test. Everything you need to know will be taught in a crystal clear way, and you will have plenty of repetition to practice what you are taught. But applying those same skills under time crunch and with many complex factors to gamble is a legit challenge!

For me, that's where the fun is. You make so many decisions per second, you constantly reevaluate those decisions, and none of them are ever certain. It is a real struggle of what you know, what you think you know, and what you want to believe. It can be a very personal challenge. You might only be walking, but the game is a very intense struggle of decision making that has very little downtime.

Some players will naturally be attracted to the game solely for the theme and authenticity. Among those, only the most hardcore will stick with it when they become hopelessly lost and have no recourse but to manually correct their mistakes. That is a small audience for sure, but I don't want to grow the audience by compromising the design. If this game is anything less than an authentic simulation, it will become a boring walking simulator. So my challenge then is to sell a really strong hook. Strong enough to motivate an audience to overcome some legit challenge.

The goal of this game is not market success. Primarily this game is a stepping stone towards my larger project, but still I want to take on that challenge of "how do I hook consumers to play something out-of-the-ordinary and then motivate them to overcome a greater challenge than most games offer?"

 

Sounds interesting 'Firewatch' uses a navigation system, but its hook is great narrative and artstyle. I was getting into it but lost interest somewhere as I kept tracking from one place to another without figuring out the goal.

 

I should play Firewatch. I watched their GDC presentations and enjoyed them a lot. For me though, story is never enough of a hook -- there has to be really engaging gameplay. That's the only reason I passed on playing it so far, despite the beautiful style and rave reviews.

 

That's true, I guess players all individually assess certain aspects differently. For example, I love playing games with puzzles and a great narrative. 'Fran Bow' which is a simple point and click game currently holds the title as my best loved game so far.

Designing a game, as you know and are no doubt more competent than me I'm sure, is a complex process. In my opinion, it is more difficult than programming, art design and sound combined.

Getting that sweet spot, where you can entice someone to play is so hard to get right.

I think your main issue is the same as me. You're going to get enticed with graphics before anything else. This is going to be a problem.

The rules in indie are simple, scope small, have a solid prototype, then run with it.

Do you watch the Extra Credits video? They're pretty much the gold standard.

 

It's a great channel. One of those channels I let play in the background while I am doing tedious/mindless work. Always pick up some good nuggets.

It probably looks like I am focused on the graphics first, but that's only because it's the only thing I can do right now. I've got design written out in detail and have been "playing the game on paper" daily, but the actual prototyping won't start until I begin working with the programmer. I can do a little of that myself, but it is so slow for me my time is really better spent on the things I can do well and quickly.
I have to go do some work in the next couple weeks, then we will begin prototyping probably after the holidays.

The main thing I try to do at this point is figure out what the right questions to ask are. Starting from, "how do we ensure it is fun?" I break that down into more specifics. It's only a place to start of course, once we are actually playing something there will be new questions I couldn't have forethought.

I also have to make sure there is a plan for how all the production will happen. I don't have to figure everything out but I need to have some idea where workflow might hit a hang up. So I do a bit of exploration into each area and try to discover any problems there might be. After all, in order to play the game we got to be able to make it first.

Since I am sitting around in the meantime, might as well get started on some art. Still, there is prioritization. The first art we need is actually essential for prototyping. Player has to use the compass, notepad, and map from square one -- these are things that cannot be substituted with primitives.

 

The first goal is to get our core mechanics functional. This means that using map and compass, we should be able to accurately travel from A to B using dead reckoning.

Dead reckoning means that you plot two points on your map, figure out the azimuth between those two points, and then follow that azimuth, keeping count of your pace for the correct distance, to arrive at the correct place. An azimuth is a degree vector. Your compass has 360 degrees, so an azimuth is some degree in relation to north.

Dead reckoning is the toughest method of land navigation. but its the most precise. It requires mathematical precision in your planning. It also requires complete confidence to carry out your plan without deviation!

Another method is called terrain association. This will be more familiar to most people. It means simply that you identify some prominent terrain feature and can confirm it on the map. If you can identify two terrain features, then you can shoot and azimuth to them and deduce your approximate position from the intersection of those two azimuths. Three terrain features and you can make a precise plot of your current position.

In our military training simulation, unfortunately, the confidence that can be gained from terrain association will be hard to find. We are in featureless backwoods of North Carolina - the long leaf pine forest nearby Fort Bragg. It is possible to discern some terrain features, but you gotta be really good at reading a map. Even still, without strong dead reckoning skills and too much reliance on terrain association, you can bet your ass will get lost!

I've used terrainparty to pull heightmap data from a 20km area around where I did my land navigation training. I plug that into world machine to generate a terrain. Right now I have not made any adjustments to the raw height data, but I go ahead and set up the world machine project because, after some playtesting, if I determine it is too hard then I can go back and edit the terrain a bit. Perhaps make some features more prominent so that player can catch a break every once in awhile. A hill that is easy to identify. A big cliff. A large flat open area. Things like this can help you confirm that you are going the right way -- but, double edge sword -- if you misread it can give you false confidence and get you into a lot of trouble.


The next thing we need is a map that accurately represents this terrain. For that there is several methods that could be used. I believe the best on is to render a texture programmatically. This texture draws a contour line of every ten meters of elevation change. This is the basis of a military topographic map. I got the idea for that here:
https://answers.unity.com/questions/788782/turn-a-terrain-into-a-contoured-map.html
If that doesn't work, there is some shaders I can make as well.
All I need is the accurate contour lines, and the rest can be filled in in photoshop.

 

Okay back in action. Today I have hit my head against a big brick wall many times and rendered it to a fine dust.

Doesn't look like much but it's a big step forward for me. I've figured out how to output a tiled terrain from World Machine. The entire play area will be a 20km x 20km terrain, but I need to break it into tiles for world streaming (performance). It does take a long time to build and export -- longer the more you split the terrain -- so for testing I do 5km tiles. The final terrain will be composed of 1km tiles though. (that's the initial plan anyway.)


Now that I got the terrain workflow figured out, I work on getting a small demo setup so that I can walk the programmer through and help him visualize how the final game is going to operate. Nothing will be functional yet without any scripting, but with some models in place I can give a better idea about how menus will operate and how the game world will be composed.

 

Good to see you still working on this buddy, I got side tracked with study, so I'm kinda done. The WIP section can be a lonely affair as it gets losts in a sea of threads, so it's hard to stay motivated. But keep going. You always have something interesting to read and btw have a great new year and start to 2020!

 

I had to go work for a bit. I hate doing that, but every once in awhile it's necessary .

Mostly I keep the WIP for myself. Lets me use community like a rubber ducky. Anybody getting something from it is big bonus but yeah, probably not realistic to expect much feedback.

Motivation for me is simple. If I don't stay super busy I get depressed. So I stay busy. The challenge is forcing myself away from work for awhile so I don't get all cross-eyed and cranky.

I am sure when you say "kinda done" what you mean is, "I took a tactical retreat so that I can overrun the enemy with a fierce surprise attack."

 

I got stuck writing my own third person controller with 2019.x because the standard template don't work!! Well, I'm sure I could power through, but I got other shizzle to be getting on with. It's really difficult to scope small, and I'm already thinking of starting another cyberpunk WIP?!

Take care of yourself dude. Ex vets are prone to mental health issues, as I'm sure you're aware of, and they like handle stuff by themselves. It's always good to talk to someone.

 

What's the actual goal in this game? Do you just have to reach a certain point on the map - the challenge coming from properly orienting yourself using these skills, or is there also a story - some kind of enemy hunting you or something like that as you try and traverse the terrain? 20km x 20km is pretty big for a game world - should make for a fascinating study of how players truly read their environment though.

 

@Deckard_89

Funny thing about 20kmx20km, which is in total 400 square kilometers.... I just realized that the terrain tiles alone make up 60gb's of data. So I did a little research. THe hugest AAA games to data don't come anywhere close to that big. Haha. Well, I'll have to size it down a bit!

I went ahead and reduced to a ten kilometer area. This is still larger than most AAA open worlds. But the goal in this game is simply land navigation, so it's not like I am compelled to fill every corner with some unique occurrence. I do, however, want to make lots of uniquely identifiable micro-ecosystems so you aren't seeing the same crap over and over. But this doesn't necessarily mean mroe enviroment art assets, it can just be a change in how they are composed and that can make a big difference in how an area feels and plays.

I still keep those extra surrounding terrain tiles though because I can reduce them down for background pieces.

The goal, which I will explain in fuller detail later once I start doing some real marketing, is to find a certain number of Control Points within a time limit. You are given a list of eight digit grid coordinates, and using a map and compass you must find each of these points which is marked by a small metal stake in the ground. The only way to find these metal stakes in the thick backwoods is to use map and compass. It won't be feasible to find them simply by wandering as there is many stakes and you could mistake the wrong one. This is common military training and also has been turned into a real life sport called "orienteering."

Besides the time hack, you also have to manage stamina. Thick brush, water, hills -- stuff that is a pain in the ass to walk through in real life -- all these things become obstacles that use your stamina. So you have to constantly make a calculation of stamina usage versus the time hack. If you push too hard too early, you won't be able to finish in time.

The story is pretty basic. You are in one of the US armies toughest courses. Only the top 1% are chosen to attend. Among those, most quit before it's over. From the remaining few only the top finishers are chosen. Land navigation is an essential skill, but in this course it's the test. If you find enough control points and finish in time, you'll pass. I also want to play with the idea of encouraging player through challenge (i.e. enticing them to not quit when they become hopelessly lost), because really just persevering is one of the big lessons land navigation has to offer. In the end I may gamify the design a bit to make it more mainstream accessible but at least to begin with I want to see what I can do to motivate players to buckle down and get through the pain the first time.

@iamthwee , thanks for looking out! Yeah I suffered from depression for many years. Still do to some degree but so long as I stay busy working towards my goals nothing gets out of hand. One great thing about this game project is that, in order to re-familiarize myself with the actual practice, I'll be driving down to the same area I did the training before and doing a bit of it again. Also I take my camera and gather lots of reference photos from the area. So research like this is a great way to get some of the "game dev" done away from the screen. I think going that little extra mile of commitment helps motivate to see a project through too. It's like a ritual you go through that says, "okay, I'm serious about this because I am traveling to do research."

 

Ah, right. I have actually done orienteering in real life as a kid, but on a much (much) smaller scale. In the North Yorkshire countryside in England, we did orienteering competitions during day and night, as part of class away trips. It's good fun actually (when you're not dying of dehydration anyway). I can see how a more "extreme" version employed in the military can be a good setting for a game, they'll definitely be a market for it on Steam I'm sure, as long as it's well made of course.

 

I hope so!

I don't have big expectations though as this the first game I am designing and handling all the marketing and publication of. Many mistakes will be made. But it's a game I want to play, and it's also going to use a lot of the same skills and techniques my larger projects will require. So mostly I come at this with the goal of educating myself.

If the game does do good though, I'd love to make some more maps for it. I think going through some idyllic english countryside would be really nice. Probably more appropriate for beginners as you can use terrain association a lot more, which is more intuitive than dead reckoning (though easier to get overconfident and make mistakes).

 

Got a few things figured out today.

One, speedTree isn't going to be viable. There is no reliable way to make alternate models of the same tree derive from the same atlas, and it's just kind of stupid to have five different variants of a pine tree all calling a separate material. But I found that with a little practice I can make trees just about as fast in Maya anyway. It's a little more work but being able to package 90% of my vegetation into a single atlas is going to be worthwhile. I'll only be working with a handful of species so the workload should be pretty manageable.

I also discovered that I will need a couple customized shaders for my trees. Nothing special, I just want something more simple than the speedtree and other default shaders coming in unity, but also need support for instancing and some vertex wind animation. I think I can manage to do this myself with Amplify Shader Editor.

As there is nothing to look at in the game besides the terrain and flora, it needs to be pretty lively and not ugly. There has to be dynamic wind and the leaves need to have some subsurface scattering so they don't have this ugly shaded look.

So far I have completed my first pass of the terrain, completed the players tools (except map), and figured out how I will make the bulk of the art assets -- the vegetation. That should cover most of the straight-up art stuff.

 

I really like the color palette and tone of this. It is coming on nicely. Are you going for a more realistic style?

 

yeah I just do it realistic style. Use up a bunch of enviro assets I've collected. I always liked the feel of the original ghost recons so I think I have that in the back of my mind a bit. I chose to make this late fall time because that will make it slightly easier to read terrain and also require less leaf cards . So double whammy there.

 

@BIGTIMEMASTER Ah, the original Ghost Recon. A genuine attempt at a playable, authentic military sim, before it became about following floating, ethereal objective markers and posing with beards and tattoos... That you've bought with real money.

 

haha, yeah...... a darn shame. But, perhaps, opportunity.

 

Got some environment stuff made. A few variations of long leaf pine and some wiregrass.

Soon I should be able to start getting gameplay mechanics implemented with the programmer. Until then I just keep working on art.

 

Some more arts. Several new species of trees and plants. Mostly I am modifying pre-made speedtrees. Everything is 50% off right now so that's great for me. I had to make a few things from scratch like the wiregrass and long leaf pine needles.

Although there is no game mechanics yet, I try to imagine what it's going to be like playing it. I try to think of some subtle clues I can give to help or hinder the player. For instance, maybe I want player to feel clever by identifying easier travel routes on the map. So I put a distinctive thorn bush on a lot of the lower slopes and draws. With experience, you might learn to identify areas like this on the map, and then favor higher, flatter areas.

Of course, the more convoluted your route becomes, the easier it can be to get lost. So you don't always want to go the most convenient way. It's always a tactical choice to make.

I don't go too far with anything at this point, but I am also thinking about art fundamentals like color and composition. Right now I am mostly making some basic rulesets with which to spawn vegetation -- the final art pass to make everything as picturesque as possible won't come until the very end. But good art takes a lot of iteration so I do make notes and have fun doing some little experiments. Just have to be careful not to waste too much time tinkering. Thats an easy thing to do especially when it's something you enjoy.

 

Wow this is looking very professional indeed.

Just a few little critiques, don't forget you can fool around with Quixel. Basically, you just need to sign up to unreal (you don't have to install the engine) just use the credentials to log onto the website and download fbx's and textures.

I used it to create this, also tried it in unity.

Link

You have two options, either use a mesh object for the grass, this is a bit taxing on the CPU GPU or go for a plane with texture. E.g

Spoiler

Another good tree resource is.

http://www.evolved-software.com/treeit/treelibrary

Works well with adjustable LOD I think. Would be more active but I gots exams to prep for lol!

P.S would love to see some post processing effects, it's lacking that 'pop factor' I feel.

E.g

Spoiler

https://assetstore.unity.com/packages/tools/particles-effects/aura-volumetric-lighting-111664

 

@iamthwee -- yeah I agree with you. I haven't done much with lighting or post yet but I dropped in Enviro sky and weather and the out-of-box lighting isn't bad. I've checked out TreeIt but it doesn't compare to Speedtree. If I was on a tighter budget I'd probably use it, but I already have a subscription to Speedtree. For the most part though I am purchasing trees and then making minor edits.



Some pics and random words:

Travel at night with poor moonlight. This requires solid dead reckoning skill, but there is one big benefit to travel at night. The control points you must find will have a glowing chemlite attached to them which can be seen from hundreds of meters away. This means that if you can just get close the final bit can be much easier.




A thick, nasty tangle you definitely don't want to crawl through. Wherever water drains (usually a terrain feature called a draw) you'll find tangles like this. Some people will walk 10km's out of the way to avoid going through something like this. But that's not always practical. Sometimes, you gotta bust through. It's not uncommon for soldiers to lose boots and become completely lost and exhausted in the process -- victims of the legendary "draw monster."

The weather does what it does, and you are on a time hack. If you make few mistakes, perhaps you can afford to wait out some bad weather. If you make too many mistakes, you'll have to hoof it all night and no matter the weather. Here is some nice fog which is made by Enviro Sky and Weather. Right now you may think it's pretty, but in the game it will become a big problem. How will you read the terrain?
One unique thing about this game is that the day and weather cycle will drive meta-strategy in an intuitive way.

 

Yes that does indeed 'look' better. The fog adds that natural white fading you see in modern paintings to distinguish foreground and distance. But you have to make sure it doesn't interfere with gameplay.

Yup I use treeit + quixel (tree and leaf atlases) and it work pretty darn well, admittedly I don't have the luxury of speedtree. With treeit you can also quick edit vertices to shape the trees and the branches.

Don't forget, there's a useful game document ->

https://forum.unity.com/threads/game-design-document-template.240038/

I'd use it to make out your game from start to finish. And remember when you've cut some, you gotta cut some more. General rule of thumb.

Looking great though! Do you have any ideas for NPC's or forest creatures etc yet?

 

Adding in some wildlife like birds, squirrels, and maybe deer would be highest order of polish not done until the end. Making the models is easy enough, but then there is animating them, giving them logic, etc etc. Although some life in the woods would go a long way in making the experience feel more rich, it has no effect on gameplay whatsoever so it goes on my "chopping board."

One NPC game mechanic that may make it in is "cheating." There is a few old dirt roads in the area which make travel much easier, but of course it means immediate disqualification to get caught walking on or near the roads. But in order to make that part of the game, then I need some NPC instructors randomly spawning and with basic AI logic.

Another kind of cheating is swapping control point stamps with other candidates. Say you run into another soldier who has one of the same grid points as you. You could ask them what the stamp was there and then you could fudge it on your own. The risk here is that it could be wrong. But the payoff could be massive time saved if it's right. A simple gamble mechanic.

I think it can add some fun strategy and character, so as long as the rest of the mechanics don't cause too many problems I will try to get that in there.

First order of business though is getting the tools working and getting the world streaming with reasonable performance so that we can test out gameplay.

 

Here's some video of walking around in the woods. Nothing happens, no reason for it to be four minutes long. I was trying to find the control point marker I put somewhere in there, but I got lost.

Wind animations on those little twiggy trees is kinda creepy.

I'm about to get started on actual game mechanics so the art will be taking backseat for a little bit. I am happy with the basic look of the art but I have to write down a bunch of notes about workflow. I need an easy way to adjust colors on vegetation, and if the workflow isn't carefully organized it's not so easy to do that.

Right now the forest kind of feels like it it's natural and here and there you find a pleasant little area, but the colors aren't quite right -- the bright green and the orange don't match and too many strong bark color variations. Feels like annoying christmas decorations in a shopping mall or something. Needs more harmony! Anyway, I won't be fixing this right away but it's on the radar.

 

I really like the feel. Sure there are a few issues but striving for photo realism is always a trap inside games especially when you add NPCs etc as it is super difficult to get them to interact properly and realistically. It's known as the uncanny effect, which I'm sure you've heard of.

Second, you're probably like me, in that you will want to model, texture and animate each creature by hand. There is something rewarding about that but ultimately it is a lot of work. I was looking on the asset store and saw this ->

https://assetstore.unity.com/packages/3d/characters/animals/forest-animals-pack-4990#content

It can take a lot to change our mindsets, we think it is cheating, or it looks like an asset flip, but don't underestimate how much time it can save. How much do you value your time? Is 40 euros worth it?

I think you've got a really good concept here. I've not seen a 'Fall' type game like this, so it would do really well on steam.

Now is the time to flesh out your narrative and mechanics. At this stage I tend to fall back to blocks and simple grid uv textures to avoid getting caught up in the endless cycle of graphics. Where are the waypoints? What are the objectives, how do you break up the forest?

^All these questions need answering.

Great job, keep up the good work.

 

The programmer needs some time to play around with the large world streaming. We will be working with 100 square kilometers! I don't think there are many games that large. But in most games, having a large world can be a liability. You don't want empty dead space. In this game, for the gameplay to work and not get boring we will need a prerequisite amount of space. Too small and player will easily memorize things, which will spoil the challenge.

Since getting a world so large to run well is the biggest obstacle to development, we are going to work on that first. Everything else is bare bones simple.

I have done some more environment work. Two new biomes: Long leaf pine forest and fire-killed forest. The long leaf pine forest isn't perfectly accurate, but I can't find any way to model long leaf pine saplings without millions of tri's. As such, I just use an ordinary pine sapling.

I've tried to simplify down the color pallette, and also work on streamlining my workflow. I want to keep the amount of files I deal with low. I also want to keep trees together in atlasses, not for performance (using instanced indirect which, as I understand, calls directly from GPU and thus negates need for atlasses), but because this means if I want to tweak the colors on a variety of trees I can do them all at once, rather than individually.

There will be a lot of polish to go into the environments. Ground scatter, unique little micro-terrain diorama pieces to scatter by hand. But all that goes at the end. Because it takes a lot of time to optimize and organize the assets. For now I focus on how the environment is going to shape gameplay, and secondarily on very basic art fundamental stuff like the colors and lighting.

Right now, the only functional tool is the compass. Well, partially functional. It always rotates to point north. So already I try to do a bit of playtesting. I want to know, how will it feel to play the game? So I put a few Control Point stakes around, plot a course using a real life notebook and the scene view, then I jump into play mode and try to find the stakes. Is it fun, or frustrating? When does it become boring? Those are things I try to experience as if I am playing this like it's a game I downloaded from steam.

 

WOOHOO!

A bit of programmer magic has happened. Feast your eyes!

What is happening here? I will explain.

The video is probably pretty obvious to you programmers. This is basic world streaming. We have a total of 100 square kilometers, but it won't do to have all of that in memory all at once. We only want to call the terrain meshes which player can see. For now, we are starting with 1km terrain tiles. I don't expect player would usually be able to see that far, but as long as performance is not an issue this is a fine place to start.

We still have to test this out with the vegetation studio instanced items, but so far everything is working well.

The script "terrainHolder" is pretty smart. It is reading the naming convention of the terrain tiles and placing them accordingly. This means I can export directly from World Machine and each of the hundred tiles will automatically get placed in the exact right spot. That is going to make it easy for me to iteratively test out many terrain styles.

 

Not a full work day but got some art's in. Another biome which, although it's on flat ground here, will be relegated to some of the higher hills in the north. I'm going to get a little inaccurate and base the northern part of the map from the Blue Ridge mountains, and the southern will be the long leaf pine forest. Then the north area, because it will have more pronounced terrain features, may be a better place to start beginners out at as it will be easier to read topography there.

 

Little note here about ground:

I had been using a tessellation shader for the ground just to avoid that CG flat ground look (which I think looks a bit outdated). But the thing with tessellation+displacement is that for it to really look good, you have to crank the values through the roof. Which isn't really practical. So what you end up with is ground that just has slightly wobbliness to it. That does still look more natural than totally flat ground, but I found something a little better I think.

I make some little disc that have like five or six triangles and peak up just a tiny bit in the middle. I put a transparent texture on these. Like scattered leaves or twigs. Then I just populate these over the terrain like grass. I can randomize their scale and rotations. It makes the ground look way more real. And then I can turn off the tessellation completely.

I can also copy/paste the spawning parameters for say, a tree, and then put disc with the appropriate type of fallen leaf to spawn beneath and nearby the tree. This adds a great touch of realistic detail.

 

Looking good, yeah I pretty came to the same conclusions with tessellation displacement. If you fire up Quixel mixer etc which is free to use (if you have an unreal acc) you can see it in its full glory - but we're talking about micro details which isn't made for games.

I would have thought your method, with a reasonable normal map would suffice.

A few more things . . .

I think you have pretty much nailed the environment now (some of those custom trees need properly uv mapping but I think you know that).

If I had any further critiques regarding the environment it would prolly be the lack of puddles or mud / marsh conditions - but again this is somewhere in the mid polish phase.

I think it is important to take a step back now and concentrate on the mechanics and gameplay. The way I do this is cutting out all graphics and even textures so I can focus on simply 'gameplay.'

I don't know what methodology you go about using for making games, but I always find the 'agile' method pretty useful. But bear in mind all I do is hobbyist stuff lol.

Spoiler

Don't forget, the more realism you put into this, the more difficult it is going to be to add believable NPC's etc and dynamic moveable objects. If you follow the above methodology you avoid running into these problems because you prototyping the entire thing in one go rather than polishing one aspect of your game one bit at a time Good luck!

 

All good points @iamthwee.

Most of my planning I don't show here. Because it's relatively simple game and I don't want to give everything away. I studied AGILE briefly and took a few things away from it. Mainly that I've organized production into phases and each phases culminates with a playable build.

I also try to define a clear, concrete goal for each phase. And this goal is defined by an end-user experience.

There won't be any NPC's. At least, that is not planned at all. If there is NPC's, probably it would just be a dude sitting or standing and using basic talking animations I can buy from store.

In the case of this game, the graphics are essential part of the gameplay. I cannot really abstract them away, as reading the terrain and the forest is one of the fundamental things player has to do. This doesn't mean I needed to make more than a single biome with like 3 tree types though. I only keep making art because it's the best use of my time right now. I am hiring a programmer for this project, and he has only begun work a day ago.In the coming weeks you'll start seeing a lot more post about how the gameplay is coming along. Once we get our core mechanics implemented, I probably won't be making much art for awhile.

Our basic plan is to first make sure we can run a world this big. From there we put in the core mechanics and go through a series of playtest to validate each part. My final art pass isn't coming until the end. That is when I will get into nitty gritty details like water, roads, props, etc., and really focus on beautifying everything. For now my art is just focusing on feels, performance testing, and workflow development. I also have no experience as an enviro artist so simply iterating over and over is helping me to do a better job with it.

 

Contour lines created with world machine.This is the basis of players map.

 

Nothing particularly complicated here, but I figure i drop the WM devices here in case somebody in future is researching similar thing. I had one idea at the end -- any moron could think of this but who knows, maybe somebody even dumber than me is out there -- but simply plug that colorizer back to the scene view and then you get a nice visualization of your contour lines. This also helps you adjust spacing on them in a more precise way.



In other news, there is lots of tedious task to take care of when dealing with big world like this. So far our performance test is very good and the whole process has gone much easier than expected. There is still the final stress test to perform however -- that is all 100 square kilometers populated with vegetation and running in a build.

 

We've been doing a ton of testing to get the full 100km world to stream with good performance. Getting close but not quite there yet. For the sake of production expediency, I may compromise for a very occasional loading bar (like every 15-30 mins) to cover for a slight hiccup that occurs when enabling a new set of terrain tiles and their associated vegetation instances. But we will see, there is still some more testing to do.


It is important to note that this testing means doing many hours of tedious work only to discover another option that will not work. You see, game development isn't fun. It is nothing but misery and frustration.

I got some iteration done on the environments this weekend. I am trying to focus on composition and colors, so that where ever you go in the world it's going to be a nice scene. It is nice when things look nice.

Everything is generated procedural, though using Vegetation Studio Pro I do have pretty good low-level control on how vegetation is spawned. I'm pretty happy that, taking a walk through the currently 8km world (8km x 8km, meaning 16 square kilometers), I am finding lots of unique looking spots. This is important not just for the sake of visual variety, but it will help player get oriented when they get lost. "Oh, I've been here before!"

I'm trying to find the right palette to really make it feel like brisk autumn weather. At the same time, I don't want everything gray and brown because that does get dull. I've also adjusted the terrain a bit for more height variation. This gives some nicer views, but also will help make terrain association more viable for noobies.

I am not fussing over performance too much yet. But as a general rule, I try to keep things running at above 60fps on ultra settings. Because I've got a pretty good computer, that gives me some idea that I am working within the realm of possibility for lower-end hardware. It helps that the programmer has a little lower end machine too.

 

Looks so real! I honestly thought some were photographs at first, the forests actually seem dense. And those clouds! Are you using an asset from the store to generate those skies? Really good, keep going man.

 

Yes the sky is Enviro Sky and Weather. It's a great product. I especially like the lighting. But the weather effects leave something to be desired, and they tank performance. At least, out of the box for me they do. I am going to try and mess with them later, but if I can't get a better result I will try to find a better rain FX at least.

 

I don't see any Tank trails.../

 

The tank trails are most definitely in the schedule

Just don't let me catch you walking down them!

 

Tankers don't walk...

 

Well we figured out the world streaming.

Tomorrow we'll get the tools working, and then we can do some orienteering.

 

Like it, reminds me a little bit of DayZ. I think in DayZ you can see how much fun it can be to navigate on an very large and realistic looking terrain, with countless of striking and interesting landmarks.

I have been trying to do this using procedural generation, but I'm still not satisfied with the results, very challenging.

 

A few years back I tried to do some land navigation in Arma. Because it has most the tools you need to do so. I think that's when I first spawned the idea that it could make a fun game in itself.

If you are using Vegetation Studio, it's a pretty powerful set up parameters to work from but also doesn't include crap you don't need either. I'm really impressed with the toolset, not just for it's performance but also for a pretty clean and flexible workflow. I really can't say there is much I would change about it if I were to redesign.

But still, to get to the point I'm at now (which is far from finished), it took like a week and a half of noodling all day. So just because it's procedural doesn't mean you don't gotta spend lots of time to make it look right -- it just gives better scale-ability.

 

Today I got the basis for our map built. There is no details beyond the terrain contours and the lat/long grid, but that is enough for us to dead reckon from.

I don't want to spend time filling in details on the map yet because I want to play the game a bit first. That will help me figure out good places to put different landmarks. For instance, I know that I'll want at least one radio tower or water tower somewhere pretty visible from a lot of places. This will be in an area I drop you in the early game. It will help beginners out having a highly visible landmark like that. I think I will put some of those "helpers" in one side of the world, and then the other side will be more untouched. This way I have an easier area for beginners, and as skill is developed I'll move player to the more difficult side.

I'll also want to cordon off the play area in a natural way. So I'll place in a paved road along the western border, a lake to the south, etc. We'll also have clearings, creeks, points of interest, and some stuff that won't go on the map but may help players out, like an old broken down truck or a hunters tree stand. You will be traversing back and forth all over this area I am calling "Camp Maccoy", so an occasional prop like that might just be the saving grace you need when you get terribly lost.


In other news, I think we ought to be ready to get our first play-testers in 2-3 weeks. I may put a build up for Feedback Friday, but since I won't have an actual tutorial ready, I think I may be asking too much commitment for that. Probably I'll need to record a video and offer step-by-step instructions what to do. If anybody here is willing to give me that kind of time, I'd love to get your feedback. Or if you already know how to use map and compass, you'll be all set and won't need my help .

 

waterproof case for the map. For immersion.

 

I've just made my first land nav exercise using the map and compass. I've been a little worried that maybe nobody will find the game fun. Or that even if they like the idea of it, they'll find it just too boring for a video game.

I'm not worried about that anymore. I gave myself one CP to find. I put it just randomly in the hills 4 clicks from my character. I decided on my route and set out. Things went fine for awhile, then I came to this big draw full of sticks and tangled-ass S***. I couldn't get through, so I had to backtrack out from there and go around. That got me a little disoriented. I tried to read the terrain but I wasn't in an area I could see very far. Then it started getting dark. Even though there is no time limit yet, I imagined there was so I pushed hard. In the dark, I got stuck in some draw even worse and got totally turned around.

Half an hour goes by I am sitting on edge of my seat and getting properly frustrated. I was near a rage quit, wandering around in the dark where I thought my CP would be, unable to find it. Then, I spot the faintest quiver of emissive light. There's my point! I had a legit rush. I haven't had a rush like that from gaming in a long time.

It's gonna be fun! It's definitely going to crush some people and be too hard for most, but those that get it are gonna be hooked. I don't think I've ever gotten a rush of emotion like that from a game.


There are some long stretches where you can more or less just walk in a straight line, but just as in real life I found these times to be a kind of relaxing, temporary respite from the stressful moments. This is only a single playtest, but within it I found a pretty wide range of experience and emotion. So I think that's a good sign.

The difficulty I faced does get me thinking more about non-boring and non-immersion killing ways to gradually introduce player. One thing we did in real life was you do the first couple training exercises with a buddy. (this doesn't actually make things easier). But in the game, having an NPC who maybe had done the course before, but med-dropped out and had to return, and sort of shows you how things are done could be a great addition of the tutorial.

 

I did land Nav in the Army, on foot and in a tank.

always had a good time. Never got lost.

Ft. Knox was better than Hood.