GAIA - AAA terrain generator, procedural texturing, planting and scene creation

I hear ya. For quite some time I was stuck on outdated systems, just barely able to play most things. And to this day I still identify with those people, so I actually have a 710GT, walmart special box that I use to test builds on as I go, ensuring it's at least able to keep up at the low end with a decent FPS (25-30). So far it's performed exceedingly well, that said... my development machine has a 1050 Ti GTX, so it's a different world. And I would expect that Adam likely uses something similar, so all the GPU compute modules are going to be faster, no doubt, but just how fast is going to widely vary from card to card across even the modern range of cards.
That said, I would expect that you'll still get modest generation times, based on what he's reported above, expect that to a player it might feel similar to a little hitch in disk IO often does with streaming anyway. This brings me back to the question of loading each of the stamps used in the generation, less the actual generation time as I'm sure that can be further optimized, but there is only so much you can do when loading 20 different stamps to generate a huge terrain. Of course, you can also counter this with some intelligent use of less stamps... using only 4 or 5 stamps is undoubtedly going to be faster than loading detailed control maps from disk.

However, I still come back to the point that I'm not so sure it's going to work for the largest variety of cases where large but not infinite terrains will often have hand crafted points of interest that need to be placed non-procedurally.
But if the session can track non procedural placements, and still track standard terrain tool actions like smoothing in areas, and hand painted forests, texturing, etc... if it can really track every action to replicate the terrain start to finish, then it might actually work and end up containing significantly less data in most use-cases.

And yet, I still want to convert the texturing to MegaSplat, so it would require the integrations to do all that at runtime too, meaning all the integration actions like that would also have to be maintained by the session. Starts sounding like an aweful lot of headache for Adam to go the full distance on that end, but it also wouldn't really surprise me if he did

 

Just for clarification, there was no dig towards his family. This is referring to Adam bringing up his family, mortgage, not once, but multiple times. I wished them well, which is more than how I'm being treated now. He brought this up a few times to me personally, and you can see for yourself where he's stated it publicly.

None of this reaction is even remotely professional. Basically, you're either forced to love this product, or bite your lip and say nothing. Great choices.

Every one of you complaining about this review, was freely able to rate this asset. You rated it high because you enjoy the tool and what you have received from it. I did not share this experience, but this is a double standard. You're trying to take that away from me. Take it, it's not important. These are opinions and if you were silenced and singled out yourself, this would be another story. This is authoritarian.

Everything I listed in my review is accurate. Streaming is not the only factor in this, but it was the factor that ended all interest in this tool. Adam and I talked about streaming. I was asking for more information in how he would address this, because it had been a decent length of time between the last talk. I was led to believe this would be implemented. I asked specifically if he would be using the scene management api, and he said yes, that's the plan and that he had not yet gone near it.

For those who wish to address the floating point problem without having to buy into another tool that is being promoted heavily, look into origin shifting.

Since this review, I've received threats of violence, stalking, nefarious activity with my connection, and hacking attempts into my accounts. This is for one review, which everyone who pays for a product on Unity has the right to do.

I stick to my original opinion and I respect yours. I only entered these forums to defend myself, as this was a clear personal attack as a result of an unfavorable review. Adam replied to the review. But after, he rallied supporters in this forum to silence the opinion. I'm not sure how much of a clearer picture you need..


I agree that I could have conducted myself better in the review, but I have very different impression of this publisher than you do. If this is worth threats, okay.

 

Damn, you just love assuming things and then using that as foundation for viewpoints and argument ...
Other user replies was their own, all you doing now is insulting each and every person other then @AdamGoodrich that replied implying they drones just following his instruction.

 

This is interesting. Currently I use Gena to do my runtime spawning loosely based on the speedtree spawning/delete example Adam had shared on youtube. I spawn really large sphere collider triggers to kinda create biomes and on entering and exit I control the spawning.

This helps me spawn fresh monsters and other prefabs in game and delete them when I leave the area. This is especially helpful because even if I manage to kill certain creatures off in an area, I would like to have them respawn over a time.

Basically I built myself something like Crux just with Gena

Is the above functionality what you have in mind @AdamGoodrich ? I am curious to know how you would approach this. Do you intend to generate all objects of a spawner rule at once like a session does ? Or do you intend to generate items around a player's vicinity just like I do with Gena ?

 

Also guys let's move on. No need to feed the troll.

 

Cool and creative use of the tech!

At this point the intention is to keep the focus in Gaia in the broad brush strokes, and Gena on light weight and highly "contextual" spawning.

They are a sort of yin and yang type approach to the problem of procedural content generation and I think they are quite complimentary.

I am also considering releasing a pro pack or something at some point in the future where both approaches are more integrated into one solution. That might be a predicate for another exciting project that I might work on, and if I do it takes all of this tech in another very cool direction.

But at the moment am just focusing on getting Gaia 2.0 cracked.

 

Good if it can be integrated because it has also a solution about floating point problem already implemented.
About streaming, you should not use that word perhaps to advert Gaia 2, or be more precise indicating it is not open world streaming, many people can interpret it differently and it can be misleading.
I had the same issue with Terrain Composer advertising streaming, in fact it was not open world streaming, and it was not completed, and finally the author was saying the only way to do it was to integrate World Streamer plugin.

 

It was there on his release trailer and I honestly fell for it myself. Oh well, at least I learned how to properly do research before getting an asset on assetstore. The only asset that has stayed true to it's intent has been Gaia. Gaia in no way means to do engine work. No gimmicks. Just the way I prefer.

 

I was one of the first beta testers of Gaia and never did Adam promise streaming. I cannot fathom how you got that idea.

Also, regarding side projects, do you know that this is how Adam makes a living? Making and selling assets...and if he makes only one and spends years on one, he loses income.

You were angry that he wasn't working on what you wanted for your game, not sure where you got the idea, but still, making assets isn't about making your game, it is about creating something that will sell to many others. There already are several streaming system, of course, those will cost you more money, but they are made for streaming. They work fine with a terrain created with Gaia, because...as you know, Gaia terrains are just terrains and Gaia just a tool to texture and add objects.

I understand how unrealistic expectations not met can be difficult and that anger can make one act out impulsively, but your review is so unlike Adam. He spent time helping people one on one after Gaia was released. He listened to us, stayed up long hours, and really cares about our opinions.

Sadly, you picked a guy that most of us will gladly defend. It is fine to give an honest, real review. Criticism can help make a tool better. But yours had little basis in reality. You personally attacked the developer and most of it based on misunderstandings on your part.

oops, posted this before reading Adam say he wanted to move on. Leaving it anyway. Do with it as you will.

 

Wow ... two pages of posts really hot .. ..
I have one question ..
Can someone suggest me the ideal setting for a terrain to use in VR ?

 

Since Unitys terrain system doesnt seem to be regarded as a modern, optimised, high performance beast, and is due to be replaced by Unity at some future point, I wouldnt be surprised if the ideal VR terrain solution didnt involve Unity terrain at all.

 

I don't think Unity will replace that in a near future ..i can give you two reasons. First is that any new system has to keep backward compatibilty .. think all the assets or pojects that use current terrain, Second, and it's not nasty, there are many assets created to ehnahce unity terrain ..and Unity make money selling those assets ....

 

They can do what they did with the particle system - keep the old one for a good while for compatibility reasons but create a new one that co-exists and isnt bogged down by legacy compatibility stuff.

I'm not predicting how long we have to wait for a new system, I just know they have acknowledged that the existing one is long in the tooth and its been over 2 and a half years since they surveyed users about what they would like from a terrain system.

As for asset store issues, I really dont think that stops them creating new systems that render some assets obsolete - the most obvious recent example was the announcement of visual scripting at some point in Unity 2017 edition.

Anyway as far as VR goes, given the new terrain system doesnt exist yet I suppose my point is that people keen to get the very best VR experiences are probably avoiding the terrain system, eg using meshes instead where practical.

 

mmm ..I have seen games with awesome terrain created with Unity using VR ... from my experience, that is very limited, and in my humble opinion big limitation is not the terrain system, but vegetation rendering. There are excellent assets, and I would include the new coming Gaia 2.0, that work on top of unity terrain adding more and more horsepower .. and few others are going to be on asset store in the next few months .. so a new terrain systems should include even all those features from asset world terrain ....anyway ...if you take a look to Unity official road map you will see Terrain in the section "Research" and same is for visual scripting ( I am not a fan of visual scripting ..).. so we will see in the next years ...
So ..let move to my question :
Any suggestion about how create a beautiful terrain in VR ? ...settings, assets, vegetation, post-processing .. sky, etc etc ...any best practice to share ..


...May be one day someone will create an asset specific to create terrain for VR ...

 

Some conversations about terrain systems, performance and future upgrades probably include vegetation as part of what they mean by terrain system. So yes I suppose it can be possible to get the wrong end of the stick when talking about the performance of terrain systems if we dont go into full technical detail when making points.

I suppose my point about suspect Unity terrain performance comes from the fact that over the years I've repeatedly seen people going on about converting terrains to meshes, and various criticisms about both functionality and performance of the unity terrain system. Including on threads such as this one which deal with excellent addons to existing terrain system. I'm assuming there are multiple reasons why some people turn terrain into standard mesh object(s) for a particular project, and that at least one of the reasons involves performance.

And since VR and your question have performance issues at their heart, please believe me when I say that my responses are designed to be very much on-topic and not a distraction from your question.

The roadmap is a little out of date compared to the few details about Unity 2017 that were mentioned in the GDC keynote. I dont recall terrain being mentioned at GDC but visual scripting certainly was, so I look forward to the roadmap being updated and/or more detail about Unity 2017 to be announced at Unite or whenever. I certainly have no particular reason or hunch that a new terrain system is coming real soon, and really I mentioned it mostly to demonstrate that the current one is acknowledged to be a bit of a relic that will ideally be replaced.

I'm sure this will be an ongoing topic and I do intend to give you more of the answers you are probably looking for once I have ascertained some more practical knowledge myself. I've got a Vive but I havent had a chance to do anything with it for some time.

But this reminds me, what sort of hardware level of VR are you talking about? Obviously there is quite a large difference between mobile VR and PC VR for example. And even at the high end I would have to be a lot more frugal when deciding everything from polycount to post-processing options if I am targeting the minimum recommended PC spec for Oculus or Vive as opposed to a project where I control the hardware and can specify 1080Ti GPU's if I want to show off and push the limits.

 

I can share a little bit of what I am learning and am attempting to reach the <11ms loop target for the Vive. Some of this may sound pretty basic to some of you.

Convert terrains to meshes. Use the lowest LOD you can get away with. Keep all textures within a single render pass. Limit the number of split vertices and triplanar textures. Choose geometry over normal maps, parallax and tessellation, even if it means a sparser scene.

Choose a biome with less foliage. Bake everything except for one directional light. Use blob shadows for dynamic objects within the vicinity of the player. Limit the field of view of the horizon through design and try to focus the player on their immediate surroundings. Utilize billboards, sprites and infinite blankets for distant objects and trees wherever possible.

Consider attaching weather effects to the player within a small surrounding volume moving through world space. For time of day, consider a soft shadow replacement such as NGSS with shadows on medium in a small range around the player. Water can have issues with reflections and may require a creative approach to make it work within a scene.

Manage aliasing through design, avoiding stairs, white lines and invisible seams in geometry. Build near the origin to avoid floating point errors and then combine things into fewer meshes before moving them further away. Try to keep objects at their original 1:1 scale if possible and use real world measurements when sizing and creating them.

Pick one post processing camera effect or a suite that uses a single pass for everything.

Lean heavily on audio for interaction and for guiding the player's attention. Use sound to create the illusion of there being more depth in the scene than is being rendered. Split or downmix stereo files to mono for positional placement.

A single plant, tree or animal can have an impact. Try to help the player focus their attention and orient themselves within the overwhelm of sensation they may be experiencing.

In my opinion, the best practice for VR as a solo developer is to design around the limitations of the medium from the ground up, rather than trying to solve all of the very complicated problems that there are right now alone.

 

Yep. I remember reading this and have a fairly strong opinion on it

Yep this is some downside. You would need to do it un-touched. However have been thinking about that and was recently talking with @antoripa about opening up the session management system to other developers. They could then plug their systems into the session as it plays back. This would mean his procedural roads and rivers could be built into the playback process. Very cool eh!!

Yep I tend to agree. In my opinion its all about the artist as they are the ones that add the magic touch. I created the stamping system purely so that their vision could come to life.

On the flip side, with good tools that capture these touches, you could just capture delta changes, and regenerate and user end. Perhaps food for thought for later plugins.

Many times Its a mid range dev machine about 3 years old - quad core 3.5 ghz , 16gb ram, with a recent GTX 1070 and running off fast SDD. The 1070 is a new addition along with my VR setup, but even my old GTX 770 screamed along with the GPU compute.

I just started exploring this myself. I am aiming for desktop VR and will share what I learn.

I have had these exact conversations with Unity. They aim to solve the hard problems, and then people like myself fill in the gaps. Its an interesting dynamic that works well. Individuals like myself can run fast because we don't have any internal barriers to overcome.

 

Hi guys

A quick question. I'm pretty sure I know that there will be too many prefabs to cope with, but having watched the grass spawner video (gaia) I wondered if it was possible to add grass in a prefab style and the likely performance I may get if I was to try and spawn it in a similar percentage covarage to the usual grass. I guess you would have to spawn it with the object spawner rather than grass and I guess I'd also get 1 fps but struggling to work out how you would / could do it.

Thanks in advance for any suggestions,
Guin

 

LOVE this idea!!! Would be great it if were even accessible via API so game developers could maybe even code their own things they want to have saved in sessions. Would this be possible too when this is done? Theoretically?

 

I haven't played around with it personally but I'm pretty sure @turboscalpeur has spawned his grasses as mesh grasses instead of billboards and played around with it. I tagged him in case he doesn't see your post. That way he can share his experiences.

 

With an instancing system this would work well - otherwise abysmally because of the gameobject overhead. I got instancing working with Gena on 5.5 but didn't go ahead with it because you miss all the game object goodness. But for grass it would actually work very well.

 

Wooo.. feature creep in Gaia 2.0....

 

Yep ... I lke the idea too ...talking about the integration between Gaia and Sentieri the idea is that the road creation can be a step to run by the gaia session manager ...to me that's looks really fantastic as you can create a complete landscape running all add one that are in Gaia's galaxy ...I am sure that @AdamGoodrich will come up with a extraordinary open system

 

Thanks for sharing your experience

 

Hello @GuinUK, I'm not exactly sure what you are meaning; Are you expecting to use a kind of Billboards "Prefab System" outside of the Unity Terrain Inspector as a kind of Scattering System included in Gaia?

If you speak of spawning only Billboards, use Gaia, this is my Album dedicated for using Gaia in Unity:
https://www.flickr.com/photos/turboscalpeur/albums/72157659431189913

Or maybe, you just wanted to speak about Grass Models as .fbx for example and built in Prefabs, then using Gaia to spawn them?
If so, I would suggest you to use Gena in Gaia for spawning your Grass Prefabs, this is my Album I used always Gena for spawning Prefabs:
https://www.flickr.com/photos/turboscalpeur/albums/72157671583483592


About the fps, it's always the same thing, Grass=take care of your fps (and what about the Insects huh?? they are more in the earth than Grass ); so choosing optimised Models with good LODs and mixing with Billboards in the distance, then using a gpu instancing system for it and obviously the shaders, animations in addition can save you of the unbearable low Perfs.

Hope that help.

 

I got the idea from Adam. Respectfully Teila, I know you dislike the review, I'm cool with that, but to keep dragging me in here when I have no further interest in Gaia is counter-productive.

This forum's strengths lie in dedicated support. All the energy you're wasting on a review, is better applied to that, this forum really shines in that regard.

Yes I realize. I entered a financial transaction with him and he brought it up to remind me.

Yeah, I am obviously discontent. I'm disappointed, and I'm also what you might call a deplorable. People are getting tired of being bullied into submission for expressing opposing opinions.

I don't speak out much publicly but when I do, it's for a reason. I chose to express this via my review, and not in this forum but it got thrown into this forum anyway.

In the marketplace, there's a buy mode, and a sell mode. When you guys release your game are you going to single out your players for an unfavorable review? I think there's some reconsideration to be had there..

You don't have to like every customer, but they're coming to YOU for something. It's not the other way around. They don't have to support you. Because of that fact, it's the true case where one should bite their tongue, and not for holding an opinion that is unpopular.

Learning this early will benefit your work in the long run. There's a balance to be had between your vision and customer satisfaction. Other Publishers demonstrate this trait brilliantly and these are the developers that I want to learn from.

No thanks, I'll opt for my own implementation rather than be coerced into buying a feature that should definitely already be in a tool for procedural worlds.

I do not expect slave work, I expect a complete product. That is it, and I get that from plenty of other great assets. I didn't with Gaia. I do not have unreasonable expectations. Multi-tile has been in his description since launch and it's still not here. From my point of view, my claims are not unfounded at all. A few pages back in this forum, Adam chose to seemingly trivialize another Publisher's work.

Anyway, this is my last post. You can continue to dwell on a review and beat me up, but I know you're all better than that. We're all here to create cool experiences. Be a little nicer to the next guy that has a differing opinion than you.

 

Nice. Just one question. The noise functions. are those to procedurally generate stamps with perlin or ridged noise ?

 

You are an adult, responsible for your own behavior. No one is dragging you anywhere. PM me if you want to argue with me, not here.

 

I agree, i would find anormal perfect reviews about a product or a game, when i need some plugin i always read the bad reviews to quickly get the limitations and downsides to see if it really suits my needs.

 

Nice feature creep! You know what else would be nice? The ability to draw inside the stamp to make adjustments. The longer you hold the left-mouse button down the "whiter" the stroke. The longer you hold the right-mouse button down the darker the stroke becomes. Hold the shift-key to undo changes under brush.

[ducks trying avoid to Adam's keyboard being thrown at him]

 

Yep ...I have just started ...I got VR sickness flying on my first terrain created with Gaia thinking to VR landscape
I will share my experience in the next days ..Many other device are going to be on the market and I think the VR isn gonna kick off a new step in what I like call "post-industral revolution" started in '80s. I put my money on that

That's make sense to me and it's what I expected ...nice to hear that

 

Read the god damn label.

The whole reason your ass is getting chewed on this is because you DID NOT DO A PROPER REVIEW. Your review is entirely personally motivated. Has nothing to do with the actual state of the product based on the features on the label. You can call it a review, you can defend your right to speak out, you even have the right to be an idiot like this, but you don't have the right to tell us we're in the wrong for disagreeing with you. I am a consumer, like many others. I am telling you, you didn't review the product. You failed. You berated it for not being what you want it to be. Get out of your bubble world and take a look at what you actually paid for.
Do you blame the kid at McDonalds taking your order when you order a Big Mac and don't get Chicken Nuggets?

It bothers me when people do this. You actually have no business reviewing a product if all you're going to do is hate on it for your failure to understand what it does, right now, before you purchase. If you bought it on hopes and dreams, that's entirely YOUR fault for making any assumptions about what it may or may not do in the future regardless of what the author promised.
You rated it 2 stars as a product, based on what? What was your ACTUAL criteria for removing those 3 stars? What was your logic? What did the product NOT do that it says it does on the label right now? Not your misconceptions. Not your idea of what it's "supposed to have" because you think it should. Not "what the author said it might contain some day". But what does it actually not deliver on, that it says it does? Show me where the active product marketing drives anything that it does not currently do please, I'd like to know where "tiling" became "streaming" too, misconceptions based on your assumption of what "tiling" means.
I'm genuinely interested as someone who reads most reviews for a product before purchase... in how you can justify a 2 star rating on a product that does what it says it does. As the so called community you are claiming to review for. I'm telling you that your review is harmful to a product that does what it claims. If I was a buyer who based entirely on reviews, your review might deter me from getting the product I actually need.
How does this benefit future buyers from knowing if Gaia does what it claims to do? Isn't your review actually counter to the intention of ensuring people know what they are buying? You can gripe all you want about it lacking streaming support, but it doesn't claim it has it... so... READ THE GOD DAMN LABEL.

Your intention of this review is not to help the community, it's just to vent in a way that you feel will get heard. You got heard, and now you don't like the result... so yes, go have a timeout in the corner for a while and pout that the community doesn't share your outrage about a product that does what it claims and nothing more.

Alright, I'm ready for your next below-the-belt shot about being some sort of mindless sheep who doesn't think for himself.

tl;dr
For those who are looking for a terrain engine that does what it says on the label, look no further. Gaia is what it says it is.

Edit:
@AdamGoodrich
I just felt compelled to edit and add this. Do not let such people or reviews deter you from keeping us in the loop on your thoughts and ideas. Most of us do recognize the difference between an idea and a planned feature. And even then, we understand things can and do change. I, for one, appreciate being in the loop every time a new idea hits the table. I'm not big on the authors who like to "surprise" us with big new releases and hidden features that get forgotten, lost, and untested. But, I also understand the fine line you walk to ensure not to set expectations that you may only be just starting to explore.

 

We had this problem before Gaia 1.0 in the beta tests

Stop looking at the shiny baubles and get back to work.

 

Since this has now entered legal territory and is against our forum rules, we must distance you from the thread. Reading over serveral posts I only see upset, so please pm author and request a refund, it's clear you're not using the product and hate it. Therefore you will not be able to post in here unless issues are resolved, for the benefit of other customers and forum users.

 

Yep. The feature is already there as a mask - not very intuitive at all - this promotes it to being a first class citizen.

Actually I have been thinking of making a paintbrush mode... you position the stamp, but then paint it gradually in... i think it would be a cool way to add interesting features to your environments.

That won't make it into the first v2 release because it will delay it .. but perhaps in a later one.. I think it would be fun to use.

 

And here are some sample images quickly put together with the new noise based stamps... its a combination of ridged and billow based noise.

Environment is 10k x 10k (obviously we are only seeing a small part of it), lighting and skies is Enviro, and water is Hydroform. I also added an oil painting pixelation effect which i think looks quite nice.

 

This is maybe an idiot-question... but I am facing the need to create multiple terrains (a long story involving heightmaps of limited rez, GIS data, etc). Can I create a terrain using Gaia and then somehow "bake" it so it becomes a normal Unity terrain, then deactivate it temporarily and create another terrain using Gaia, rinse, repeat until I have several tiles of decorated terrain which I can then stitch together with one of the tools from the asset store? I am pretty new at this, so am prepared to be the village idiot. Gaia sure is fun, but also slow on big terrains (my hardware is not feeble but not bleeding edge either). Divvying them up would be a win for me. Reading back in this long discussion it seems that I am not alone in this

 

Um yep. Gaia 2.0 is not far away either and its already handling multi tile quite nicely. PM me if you want to know more.

 

Thanks Adam, I have been digging around a bit in Unity forums but have not yet found a way to PM other users... If Gaia 2.0 is coming any day now then I think I will just wait patiently. Sure don't want to spend hours and painful hours trying to reinvent a wheel that you are about to release

I'm very interested in modelling (to a reasonable, not obsessive accuracy) real-world regions; my proposed workflow would be generic terrains+imported-heightmap+gaia-decorations. I'm more interested in getting the job done than in tinkering with Unity, so am inclined to purchase tools at the store rather than study for months to acquire the skills to roll my own water and terrain fx... I would be interested in your recommendations for terrain-decorating tools that are easy to use w/o a lot of fussing or coding, so I can focus on the GIS import issues (which are quite challenging enough!).

Curious to know, can I easily feed Gaia my own vegetation assets (say, different mix of trees and plants for different bioregions) and have the spawner distribute them just like the built-in assets? I have a rather nice collection of rainforest and tropical plants and would like to let Gaia loose on a "tropical island" terrain. Again probably a dumb question... a simple "yes" or "no" will suffice

 

To PM, click on the name above my picture, then select start conversation. The release is progressing well, but still a lot of work left to do, so it will be several weeks at the very least, probably closer to a month until its completed and nicely tested.

 

See Adam's post about PMing.

To answer your question simply, yes you can use your own assets for grass, trees, and other terrain objects. The "built in" assets as you put it, are actually contributed or free packages of standard assets (turboscaleur's grass and some free unity SpeedTrees).
To clarify, Gaia does in fact create completely standard Unity terrains, such that as you create and "bake" it as you put it, you are creating a completely standard unity terrain you can further edit if you so choose. Gaia 2.0 will make editing of much larger terrains much more efficient in Unity, while making it still break down into smaller tile chunks and handle seamless chunk edges when 2 tiles are accurately placed next to each other.
Something you might consider for your "imported heightmaps" is to turn them into Gaia stamps, and simply stamp them to the terrain(s), making it very easy to reuse and regenerate terrain should you have the need, plus fitting into the Gaia workflow with minimal effort. @AdamGoodrich can correct me if I'm wrong here, as I haven't looked at the scanner, but I think you can feed it heightmaps directly to produce new stamps.
As for terrain "decorating", this is pretty much the primary focus of what Gaia does. You might think of texturing as a different process, but it is really just spawning a patch of texture based on a ruleset. Same as you'd do to place a tree, grass, or rocks. It is also worth noting as it comes up often... Gaia is the big brush, it'll let you procedurally place down stuff by rulesets with a reasonable level of granularity, but if you want more control over this process then look into it's sister companion tool also made by Adam, called GeNa which is a finer granular control over spawning.

Given the sound of what you want to do, and lacking any further information, I would personally recommend waiting for Gaia 2.0, and if budget isn't too tight, grab GeNa as a "full rainbow" spawner solution between the two.

 

Hear hear!!

 

Is there a further step to "bake" it? I notice that when I have created and stamped Terrain A, bits of it are parented to Gaia constructs in the hierarchy. The terrain is in one place, the spawned objects that decorate it are in another... if I want to create and stamp a new terrain, it appears that I have to move all those bits of the first terrain into some other parent, so I can delete all the Gaia-specific structure and be left with only a generic terrain and its children. Then I can deactivate Terrain A, start Gaia afresh and build Terrain B. I have done this manually, and it's a little tedious. Perhaps there is a cleaner way to "bake" the end result so I am ready to start over on the next one?

(btw, am I the only person who finds it quite frustrating that Unity won't let me rotate a terrain?)

That would be great. I have tried a cheap Stitcher tool but was not impressed with the result -- actually it was quite crude and in some cases generated nasty artifacts at the boundary. Disappointing. At present I have no way to stitch Terrain A and Terrain B together nicely.

Interesting. I'm not sure I would re-use them (the project I'm working on is a one-off) but that's worth some thought.

I think the biggest challenge for me at present is wrapping my head around appropriate scale -- how the scale of a terrain relates to the scale of, e.g. trees and the FP player and the textures and the scale of water surface fx and so on... and how to use external heightmaps (which are generally pretty coarse, alas) to create appropriately scaled landforms. I have not yet found a tutorial on this specific issue.... anyway, this is not a Gaia issue (other than in how it relates to the sizing of terrains), but if anyone knows of a good write-up on how to select terrain size and its relationship to real-world scale, I'd like to educate myself. There must be a sweet spot where the terrain is sufficiently fractal/detailed to be convincing to the player, yet not 1:1 with the real world (impossibly expensive); I'm not sure how to find it. Right now I'm seriously thinking about upsampling GIS heightmaps...

As y'all can tell, I am pretty much a novice with Unity; have done a few tutorial projects, and now I'm trying to take the great leap into tackling something I really want to achieve vs following a pre-created tute. I guess one of the dangers of writing code-free, pointyclicky tools like Gaia for a sophisticated task like terrain generation is that n00bs like me suddenly get into the game and use the friendly tools to venture way beyond our expertise level anyhow I apologise for my lack of Unity experience and appreciate the patience and good advice. If my level of discourse is too low for this forum I am willing to slink off and lurk until I'm better qualified to converse with my [virtual] elders.

 

Thanks for the reply. I think I will definitely look into the instancing idea. I really want to try and create a unique grass feel to my game not using billboards and this could work well. To be honest I'm just playing around with Gaia and Gena and seeing the different options I have in terms of textures, billboards and prefab .fbx models.

Both of these look amazing, but it is the latter I am most interested in. I was trying to work out the best method of using Gena to spawn them to look thick, lush and organic without hitting my fps to a snailspace. I'm quite new to the Unity scene and am trying to work out the limitations of what can be achieved.

Thanks for responding

 

Hehe - i am quite happy about this - the logic to make gaia work almost melted my head

We are all noobs at something. Just have some patience, be easy on yourself and enjoy the journey

 

If it's any consolation I find myself in exactly the same boat. I've run through 4-5 Unity tutorials, most of Gaia's tutorials and because I can click 3 buttons on Gaia, replaced a couple of textures and swapped the water for a paid asset I feel like I want to be able to do everything straight away. It's great how much support people are willing to give though, and no doubt we'll be creating AAA games in just a matter of days (or more likely in 30 years!) For now I'm going to enjoy the ride, keep asking the odd question and see where it takes me.

 

Made this in a literally few mins with the new noise stamp - its just plain old billow noise. Stamped, scaled and rotated a couple of times. Textured, used Gena to add the farmstead and the trees. Water is Hydroform, light is enviro, and a light smattering of post fx.

Need to remind myself to stop playing and get back to work!

 

Good lord 153 Forum pages! Very interesting lighting you have going on above with Hydroform.
Speaking of water I wanted to know which water system you used in this picture from Gaia because judging by the images the shore waves look really realistic.



Also I had another question about Gaia, judging by the information on the asset it doesn't seem to allow you to really fine tune the terrain like editing it to include things like valleys and generally do a sort of terrain painting to deform the terrain like Unity can. Is this true or not/is it coming etc?

 

I'm going to throw a guess out there on the water.. I think that may be Suimono.. But I could be wrong. Maybe Adam can remember.

As far as the fine tuning, terrains created with Gaia are Unity terrains. So once you are done stamping, you can click on the terrain and use any tools you would normally edit any normal Unity terrain with to fine tune it. You can also use multiple stamps to continue adding features both in add and subtract mode (add to raise, subtract to lower)..

 

I did not know that the terrains with Gaia were actual Unity terrains that changes the game a bit.
I don't think that water is Suimono but I can't be sure also.