How Close Can Indies Get To AAA Games?

Ultimately though, this entire thread is really about costs.

"How close can indie get to AAA" is literally the question of "How cost efficient can you possibly be?"

That's the entire ball game, and even if some of us may not be comfortable with royalty free assets - not discussing them is not attempting to answer the question seriously.

Royalty free assets are the most cost efficient solution there is, bar none. The question of how we can best use them is an important one if we really want to be as efficient as possible with time and money.

 

A good asset flip is almost impossible to pull off. Not only is it very hard to get anything to fit together by sheer coincidence or any rational amount of tweaking, but you can count on two hands the number of large packs of models that are both good enough quality and comprehensive enough to satisfy a significant amount of the art needs (i.e. MK4 and a few others). Problem is, players can probably spot them from a mile away.

IMO the main issue is the textures. If models were commonly designed for tiling texture arrays and separated AO/wear maps, then they could be useful but very few are. Substance materials are another way to give range to the character of a model, but again there aren't many and flexibility is usually limited.

Unfortunately, I think the best thing is to grind at modelling/texturing until you're good, or hire someone.

 

There is a lot of middle ground there - vertex painting and the like are also tools for this. A lot can be accomplished with those approaches, and with various shaders.

For example, adding dirt, decals, moss, other details to a given instance can make something feel fresh and different while also fitting the aesthetic of the scene. The blacksmith demo has good examples here, with various procedural moss shaders being added to viking village buildings.

They took those available assets and fit them to the darker environment and aesthetic they were building for the blacksmith reel.

 

It's an interesting topic to discuss as I think the market for selling assets to devs has way more potential than it seems. If it was possible to modify any given asset through a 'slider' type interface, then there would be a ton of opportunity.

This interface would not adjust properties specific to any asset but rather modify more general properties:

• Automated material separation and generating a 'splatmap', with the possibility to assign tiling materials (or generate a texture by tri-planar projecting the new textures according to the splatmap);
  
• Adding wear and tear based on curvature etc;
• Removing vertices and closing off the cut;
• etc;

It would depend on how the object was made and to what extent the software could read the underlying structure of the texture information, but I think it would be really useful if it worked out.

 

This is the one that caused all the kaffluffle (it is a word) https://forum.unity3d.com/threads/fallen-spirit-rpg.225407/

I wish I still had the project so I could show the actual sheer size of the game instead of select screenshots, but it was massive.. I'm not talking Nier Automata style big, I'm talking twice the size of Skyrim big. From the "city view" you could see the coastline for miles down..

There's no way I could do that without the asset store, none and I did try to start swapping meshes out whilst combining textures etc. to add in my own artwork.. There's mesh improvements / re-textures on page two and the house to the right of page 1 (picture 4) was one of twenty kitbash houses I made and used textures from kits to keep them inline with each other.

On the caves picture second one down, I re-did all the shaders / particles shaders and tried to give a completely new "feel" to the aesthetics.. Ultimatley obviously it wasn't good enough. The prototype essentially failed (due to technical reasons as well) and whilst it was "efficient" assets were one of the reasons why, I can't ignore feedback completely..

For the most part I didn't get "jumped on" for it, a lot of the criticism was clear / insightful and accurate so I thank people for giving the time to do so.. Although I've had to reduce the scope of the game nearly 95% (roughly) to now custom build it from the ground up.

Sure, I'll use megascans / speedtree and since it's used in near enough EVERY major AAA game of this type then I'd love to hear the logic behind that being an issue. After all ShadowK IS birch yo.!

After that mess over two years, I have to admit my motivation has somewhat plumeted but I'm sure I'll get there again..

 

A lot of this stuff exists to some extent, but the most useful stuff for really making a game probably won't sell as well as stuff that looks like it can be really useful for making a game.

One of the most constant challenges I need to deal with is normalizing stuff. Different artists selling different assets tend to use different ranges of value for everything - from specularity to heighmap height to albedo. Having tools to 'color balance' the various maps that games need - that alone would be a hugely useful tool.

Having really good automated tooling for normalizing albedo against a certain tone and color palette would also go a long way.

I look at those screenshots and I see good work. I am very familiar with some of the assets you used, but the composition is solid and the scene is whole. I can't imagine a player calling that an asset flip.

In all honesty, if you pulled the most iconic pieces out - like that arch - I wouldn't have known for sure if it was an asset or not and I know my bloody assets

Games are all smoke and mirrors anyway. If you can fool someone like me into not knowing its a stock asset, you're good to go.

 

Tbh, I don't think it would be useful enough until you could replace an entire material. IMO materials are about 75% of the character of an asset - just look at textureless objects, they could potentially look like anything.

 

I always forget about that thread of yours @ShadowK

At least from the comments on the thread other than one or possibly two haters (jealous much?) the feedback was overwhelmingly positive regarding the visuals. The main theme that I see in that thread is "looks fantastic when will you move on from making graphics to the other aspects?" at least that is what I got out of reading through it.

I don't think this can be used as an example of asset store use being a bad thing. If anything I think this project would have served as a shining example of how one should best use such assets.

 

Just want to point out, this isn't a commercial project per see, I mean as a end goal.

 

Which project? Is there a project defined now?

EDIT: @neoshaman to clarify that ^^^ I mean do you have the project defined now to know clearly what you are building? It seemed like you were planning on doing that all up front instead of iteratively.

 

I meant my "exercices" I do not close the door, but it's not meant to be sold at the end "as a goal". Right Now I'm trying to fetch a story that makes sense for the structure and define the biomes and theme of each region. This will led to characters, enemy types, environments and props for each region.

 

Okay so right now the project is still in the phase of defining what the project is. Sounds good! Obviously need that.

 

I won't expect to spend one or two years like AAA defining the identity though lol, I'm trying to see how fast I can go through this too, it's kind of fun. Obviously it will slow down when execution will be necessary.

 

Hopefully not a year or two... if so you should bookmark this thread now because even it will likely be long buried by that time and a search will reveal thousands of similar threads. lol

I am doing nothing on any game dev. I almost did. I'll probably get back to it this weekend or so. And when I do I think I will build out the level like before (only tiny) and use primitives for everything. Mainly because I have mentioned so many times about making an rpg completely out of cubes. I was thinking cube environment as well but I think this will be fine. Might be more visually interesting than yet another rpgish looking game with the game environments we've seen countless times and the characters and barrels and wagons / carts and so forth.

So the idea is less work for me to dig and find the assets and just maybe have something a bit fresh as a result. Well actually I might do some ultra low poly objects for collectibles and props but definitely all characters will be primitives. Different size cubes. Different size pyramids. Etc. I think a big part of it is I just find it more interesting to work within these constraints and try to pull it off.

 

My main problem now before jumping to next step is just finding the main opposition and what the ramping is according to open world structure, finding the opposition basically mean finding the motivation and the stakes, what slow me down is that I arbitrary decided early on that it needed 7 plot coupon to collect in any order because open world (though I think I will have a prescribe order through soft gating with level and other character progression trick).

The nature of the plot coupon can be very flexible, it's either actual objects (like seven gems thingy, but then I must justify a bit more), requirement (for example trial or simply skill and power, or a mixed of the types) or "revelation" (basically you access hints at something bigger, and when all the hints are put together the final quest can happen, for example having all kingdom having seemingly unrelated problem, but they seems to have all the same sources, there is an invasion coming, and all incident is to prepare that). I still need a strong sense of direction from the beginning (to avoid aimless progression until 2 third in the game), something to make the player care.

 

I'm not sure if it's a smaller problem than I think it is or you're too kind, anyway that thread was far from the only comments I got about it. As @Billy4184 said, after a certain amount of time it becomes more of an issue to correct / change all this artwork than making it from scratch yourself..

 

I'd much rather someone come to me asking for a model from scratch rather than tweak something someone else has done. I've done it once or twice and it's long, frustrating manual work on textures/UVs, I probably wouldn't do it again.

I still think that a well-made software could potentially do a good job on a model with fairly clean and well-defined materials. But how difficult it would be to make something that could get a good result is a good question.

 

@ShadowK @Billy4184 the really interesting thing about this is I remember Unity used to create threads asking for people to share their game development successes with the asset store. And many folks posted about using assets from building to kits to various scripting systems to art.

So they of course recognize this... it is good for business... and it is a way to get some free marketing when they feature your game in such a write-up.

 

Well I do try to be a decent human being instead of an ass unless someone else is clearly being an ass. lol But in this case I just went by the thread. It is loaded with praise for the visuals and excitement over the game in general.

 

Well can't say it ultimatley matters anyway, my vision is somewhere between the Adam demo, Crysis 3 and what I've seen of Star Citizen for the main bit and then split into several biomes.

It's somewhat like "Noah's ark", originally it was meant to be set on a "planet ship" with levels and levels for different races and some of them spanned massively. The plan is the same really but now it's been reduced to the size of a terraforming outpost where it splits off into different segments..

Anyway the point is the asset store can't cover it in the slightest, even if most of the packs were dedicated there wouldn't be enough to cover what I needed. The story has to be done via voice actors specific to the project, once I've started I might as well finish.

If nothing else after all this time, I have a clear picture of what I'm trying to achieve and what I'm going to use to achieve it.

 

Okay, I spent some time on end of my lunch break working on this... and I think what I'll do is still keep everything very simple but anything that is super easy & quick to add I will do that.

So basically instead of going for something like this...


I am now going for something more along the lines of this...


I think is doable. It will be reasonably fast to knock out and I personally think it is much more visually interesting already.

EDIT: Well actually I can't say that because I think the original way I was testing also looks very visually interesting as well or at least it looks like pretty high quality. I think what it basically comes down too is I am not too keen on games looking sort of like real-life. But I tried to make the mountains look a bit otherworldly for that reason and I think they do add some visual interest. I do like the water a hell of a lot though and the whole scene was so easy & fast to make.

The great thing is I now have found 2 game development tools I can use that are super intuitive for me and that is very empowering and motivating. Actually this reminds me I need to bug @ShadowK about something.

 

In all honesty - I just wish there were more posts from more people who were actually doing something involving material progress in a game of some sort.

I'm a simple dude, just show me screenshots of any kind of progress toward a real playable anything.

I don't like blocks of text, I don't like abstract thoughts on what might be or plans for hypothetical futures. I want to see material progress.

 

I agree but wrong forum for that stuff. Go to WIP.

I definitely am the same way though. I do want it to look interesting (not saying great or even good just interesting) but mainly I just want to get to the parts of actually implementing stuff. Going this ultra low poly way I can knock out the stuff fast.

So hopefully can knock out enough content tonight to start programming sometime this weekend and getting stuff happening.

 

Problem with WIP is it's too focused on a specific project, I want a sampler

 

Yeah that's why spectacle is always more entertaining, but most stuff have boring part to slog through before going to the fun part. Nobody like chores (well most body lol) but they are part of the job!

There is a thing as starting too early without a plan, it makes it more difficult after when you have to patch yur things up, and since the goal is AAA quality I would rather no redo the mistake I made on my own key project already.

But I do this in part BECAUSE I want to show the "chores" too, the proces from simple planning to execution. Anyway I'm just a decision away from a first prototype where I need to validate what basic unit of scope will be and then start again making list. SO stay tune.

The entire IRONY is that I want to ensure my game has minimum spectacle in the progression (ie I need a convincing story progression, even if generic). I gave myself one week to decide before defaulting to a lesser solution.

I do this because I'm less effective under stuff I can see the plothole too easily (I wont avoid them though if necessary, great story have them in spades), so instead of having zillion of small decisions later, to try to ignore a problem I made for myself earlier, I want to have one big decision that makes all others logical from it later on. So I can focus solely on quality rather than small logical leap and retcon.

 

Does it have to be a game?



Currently in the process of moving my CA program (what I showed in the AI thread) from Unity over to a Winform application. I worked in Unity because I was more familiar with it, but as I finish up the main features of the program ("first order" or unmoving, yet state-changing, and "second order" or actual moving agents on a grid, stochastic (unlike most which are deterministic) CA) I feel like a more typical program interface would be way more intuitive and just less cumbersome than all that Unity business.

The UI has honestly been super easy to set up, but I've had a bit of trouble with the programming - first bigger one was with VS / Microsoft auto-interpolating low-res images, so I had to get a workaround for that (though as you can see in the second image, the grid image isn't centered correctly, so I need to fix that somehow). Next one I'm having trouble with now is that there's not really an "update" method where I can basically make an infinite loop scot-free - I'm going to have to add a second (maybe background worker) thread. But I have absolutely no experience with that and what little I've tried thus far hasn't worked out really. Other than that though I'm having lots of fun and it's going okay.

I've also been working on a visual novel, I suppose I could post excerpts for you I'm not really working on a "game" at the moment, because the two of these have been taking up my time. Hoping to start to wrap up the CA "port" near the end of next week at latest (I just started it this week), so maybe I can focus fully on the VN and the preceding RPG (they're going to be part of a trilogy following one overarching plotline) demo scene (which is a prologue to the game, so it sort of sets the stage).

 

So... I did absolutely nothing tonight. lol Well I had a beer and watched a few episodes of Ancient Aliens while checking out game dev forums.

@EternalAmbiguity that's an interesting project. I watched the video from the AI thread and first thing that popped into my head is this is a dual-purpose application... first your original purpose of experimenting with cellular automation and second just preset the colors and names for terrain / map generation.

 

Yeah I've thought about something like it for terrains. And one could do white and black cubes so you could have a heightmap from it too. You could use interpolation to smooth the blocks out.

Maybe that should be what i do for the PCG Game Jam Does that count as PCG?

 

I think absolutely no doubt it is. And would be a useful tool for folks to use as well. Win-win and maybe you also win. lol

 

Cool, nice to have a little nudge. I've already been messing around for the last hour or so and I've got terrains being created at runtime with the press of a button, so I can incorporate that and then make it a heightmap. Should be interesting.

 

AS much as I like to push for knowledge structured pcg instead of random stumbling, it does

I made a bit more planning and try to track dependency of overall task, I have major decision to take in how I want to build the world in term of theme, conflict, antagonist, etc ... in order to decide how to use the assets and their number, to give the player solid direction, ground some logic (ie structure) to the world and be engaging in at least a pulpy sense.

BUT that's not necessary to start prototyping the cost of "space", ie the basic gameplay, the basic navigation, density of elements and the lay out logic. So I will start experimenting and releasing demo made with blitz3d. Why blitz3D, well I have a faster workflow to test visual and basic procedural idea that ask for no complex data (which also encourage to find Occam razor implementation of idea). Also I can use any image to send to the heightmap terrain, it pick the R channel to build the heightmap, and I can use the the other channel to put object ID with code, so I can draw literally the map on any paint soft that support layer.

1 region being 750² x 2m, that's the space I need to test.

 

CA is definitely "knowledge based," though it of course has a pretty strong random factor to it.

I look forward to your prototype.

 

I meant knowledge in level design, ie have implicit or explicit reasoning of the flow of progression. But really whatever people find good lol, outside of pcg thread, I have no stigma about what people call pcg lol. I would just like people to think about what pcg mean applied to level design that isn't just "randomness" but applied intentional design.

 

@neoshaman you seem to lean toward these other tools similar to the way I do. I agree if you're going to make something like an rpg or pretty decked out fps B3D and a few others would be better IF you are experienced with them.

Although I completely understand your choice of Blitz3D for this keep in mind it will be limited graphically because it is locked to DX7 IIRC but at least I think you won't wrestle so much on the implementation side (assuming you have the programming experience).

I am using AGK2 for my current project. Gonna build a quick n easy world editor for it this weekend maybe tonight. I'll send a pm to you about something that might be useful.

 

I don't post very much, but here is a screenshot from my game.

It is a strategy game inspired by old games like Age of Empires. It will focus more on controlling a smaller number of units and using the landscape to your advantage. Instead of building things all over, there will be one basecamp that you can move around the map and must defend.

Well, I'm not ready to reveal much about the game, but here is a recent screenshot. I've been working hard on making the landscape look nice.



Here was the last video I made talking about it:

 

I don't know how to make everything in blitz3D, it's mostly to visualize space and having custom terrain code where I hijack the drawing function. Since the terrain is basically just an image, I can draw directly on it using the blitz3D 2d painting tools.

On unity I may or may not have to write code to draw primitive on the terrain, because there is no such tools on unity. SO making something like a line or a triangle is a diversion when I just want to carve a path.

On blitz I would just blend one image buffer onto another, done! However the challenge remain about using unity, so I may be able to generate data from blitz3D and feed unity with it later!

My hypothesis is that I need to define a procedural workflow and this will be the key to realize the game, aka automatic level building, since it is one of the most consuming part and where I see a lot of people giving up on project with simpler ambition.

I want to use pcg as a the equivalent of a "HTML/CSS" (not sure it's the right metaphor lol) implementation for level design, define the descriptions and let the algorithm do the lay out. Which I hope will translate and scale to smaller asset too. That's why upfront planning is more critical to me. In order to achieve that I need clear intent and structure.

 

I fully intend to have logic applied to the level design. But I'll leave it at that, I don't want to give too much away

 

Now we're talking!!

How far along are you?

The ground/grass textures look great, as do your props!

Your video reminds me of some of my earlier videos - you need to learn to cut them down to like a few minutes so people watch

 

Thanks!

I should probably do videos more often so I don't have as much to talk about. I do tend to get carried away.

Anyway, I'm hoping to get a playable demo out by the end of the year. After that, I can start wrapping things up and enter the final stretch of polishing, building out the rest of the levels, and getting ready for release.

I make my props/characters in Modo, and texture with Substance Painter. I'm mostly an artist, but I am learning programming as best I can to get this done solo.

 

If you want a hand with some of the code stuff, I'll trade you assistance for some character models

 

I'll keep that in mind.

 

That to me looks like a professionally made indie game, don't take that as anything but a compliment because that's rare.. If you'd of said it was made by the same team as pillars of eternity, I probably wouldn't of questioned it..

This is what I generally mean by "good graphics", it needs to look professional within it's respective field / genre and you've hit the mark on that. Also it's rare I give compliments, looking forward to seeing the rest of it..

 

Hi,

Im one who is trying to make an AAA/BBB looking RTS game ( http://www.metadesc.com/studio/ ). Working on the project since 3 years and I own the half asset strore... The asset in the Unity asset store are awesome, without, indie developers wouldn't have any chance to develop a commercial game. I also bought a commercial license for Unity 5, but this license ended now and I still need at least a year for my game. Unity is a good platform for every kind of games and with the assets from the asset store it is very powerful. In 10 years I could make an AAA game, but I would recommend you to make a good game and save your time in making not so important things. You should not try to make a game studio trailer, animations, cut scenes and so on, concentrate on the most important in your game, then you get something like ABC, but make the most important A. By the way, 99% of gamers don't even watch the cut scenes. Bigger game studios continuously failing in "making good games", use that advantage.

But!

- Unity lacks on stable releases.
- You must know the bugs.
- You have to buy assets in the asset store.

 

Artist driven games in general tend to be more successful than programmer driven games.

I've seen a number of pretty solid looking artist driven solo efforts.

It says a lot about what's most important.

 

Well AAA have been riding off the back of graphics for years now, can't say I've seen any "game" changing mechanics especially in terms of RPG's / FPS's.. So it obviously works .

I don't think graphics are the most important thing once you're hooked on the game, issue is you need to draw the crowd in first. I did offer to do a graphics pass / donate some artwork to your project, but you never seemed that interested ..

 

If you want to donate a couple simple characters, I am super interested. I desperately need some good historically based 11th century peasants / vikings / persians / italians.

I've just never had anyone who said they'd deliver actually deliver any, so I tend to not get my hopes up

Only reason I mentioned it to @DrewMelton is that it seems like he's already producing the kinds of guys I might need - low poly and historically rooted.

 

That looks great dude, got any video?

This is so exciting, people showing their projects is awesome!

Seems like there's some really good stuff out there.

 

Do you all ever check out other sites such as GameJolt, YouTube or even Steam?

The reason I ask that is because there is no shortage of great looking games being made. And maybe this is another reason why I am not as into those as some of you. I see them all the time while checking out various sites or just browsing YouTube game WIPS.

I think all of these look very well done. They are at varying stages of completion.

The thing is there are so many people making games (including many teams of varying sizes) there are plenty of examples of just about any kind of presentation or genre you are interested in. More or less. Oddly I think a good 3D platformerish N64 style action or action rpg is perhaps most lacking out there. lol

 

Well there is a balance, a great artistic asset is never better than the behaviour you slap on it, and vice versa the mechanics is never better than the asset you slap it on

Edit: as a VR dev I'm shocked by of many assets that's are off scale for example (very import in VR)

 

A lot of the AAA-like work from Indies you'll find more on YT. I am pretty sure the bulk of the AAA-like Indie work is being done for mobile only for whatever reason.

For example...

And they are also in languages I cannot understand... but it does seem like a lot of devs in these areas of the world are really pushing and reaching AAA-like experiences.

Anyway there are a lot of em out there...

Or heck just watch this for some of the AAA-like rpgs done in Unity & Unreal

Alright I won't post any more. I just am saying this is why I don't really understand why people get so excited over such stuff when these kind of AAA-like games are pretty much all over. Although strangely more on mobile it seems. That is kind of odd to me because such games would likely do very well on desktop on Steam.

I completely get if you just go by all of the games posted in the Unity forums you may think these are very rare. That's why you need to get out there and look around. It's a big world. lol

DISCLAIMER: And maybe some of these are made by what some would consider AAA companies. I don't know about that part. Maybe some are AAA companies even. All I know is there is a lot of stuff like this out there either done, just getting started or anywhere in between.

Also I am not even sure what the purpose of this thread is any more. Big part of that is because the term Indie has two meanings. One side sees it as being very small teams kind of like homebrew development and the other sees it liie the music industry and in that case it has basically no meaning for discussions like this because that definition of Indie can mean a team of 150 ex AAA developers who each have half a million in their bank account. Obviously that is entirely different from @frosted working alone.

I think when people make these threads they should clarify up front do they mean one person? Or what exactly do they mean? And I also think people posting such threads are coming from the definition of Indie being homebrew development by 1 or a very few people. Otherwise it doesn't make any sense.