You do? Ok, please explain to the folks at Gamedev.net how long does it typically take to develop a current gen 3D game (like Doom 4), how many hours a day do the developers work, what part of the development process is creating the environments, models, textures and other in-game assets, are there any side projects, creating a new engine, tweaking an existing engine. The time taken in polishing and testing. The time spent outside asset creation and coding.
Here is my reasoning on that post: Because with Unity and Unreal engine 4, the user has most of the tools needed to create a similar title given enough time. (Minus the voice acting and using other modes of animating other than the motion capture rig). Many necessary assets can even be got from the internet free. Reducing the number of devs to one means no salary involved. Putting existing assets together, well it has got to be a piece of cake. But the game does not have to be so long, only the length that reasonable time will allow.
This is what I was trying to tell those at Gamedev, It seems they misunderstood my definition of AAA, but in that post my definition is as follows:
I mentioned an AAA type game, meaning any 3D game that is by my definition a game with the quality similar to a title typically released by professional game dev studios of any length and size, with any generation, including 4th to 5th generation technology (n64 to ps2 era).
So from the above definition, I was trying to explain that a skilful sole dev can develop a Quality 3D game of any length or size (including current gen) that time will allow.
I'll tackle the visual/environment parts, since this is my area of expertise. There's so much work involved into creating a believable world than just making an asset flip (Steam already has tons of those, and they are of generally very poor quality). Buying/finding assets doesn't automatically make the world you're building. Finding assets certainly doesn't make a coherent or interesting world unless you put the work into it.
I'm a level artist. When I started working in the industry we did most of the visual part of a level ourselves. A single level artist (with the help of a level designer, who does the gameplay part) would model a low resolution version of the level respecting the designer's rules. They would also create a couple placeholder textures or use some already available. The level is kept low-ish resolution because the artist and designer need to iterate several times on the level, make changes to the layout, try new creative ideas, cut sections because of time/budget, etc. Then the artist would refine the modeling and textures to a shippable level of quality. in some cases a dedicated modeler/texture artist would help out for certain assets, like statues or a texture of a unique painting, and tweak textures for color consistency, but back then it was pretty common for the level artist do have been responsible for 90% of the graphical assets a level (characters and effects excluded).
An important detail is
lot of work goes into scene composition and flow.
This is aside from the actual graphic assets themselves. These is the decisions to where things are and how they're integrated, and this is an
enormous amount of work that can't simply be downloaded, or automated by a plugin in Unity. it's not enough to simply model an environment and call it done. You need to study the style the art director wants so that it's consistent with the rest of the game. Also extremely important are reveals and first impressions when entering any area. What does the player see first? is there anything in the way? Is the point the player needs to be attentive of framed by the rest of the environment? Can I have the sunlight hit the point of attention to highlight it? Does that angle of the sun make sense for the rest of the level? Where do I put the most detail in an area? Where do I specifically
avoid putting detail to not attract attention to stuff that would confuse the player? Does this thing look like a ledge that I can climb? Should the player be able to climb it? If it looks like a ledge but I don't want the player climbing up on that balcony, how do I block it in a way that makes sense, visually? Is this a grocery market stall and how do I make it look convincing? At what angle do I enter the grocery market? I see it's from this angle, but then all the roofs of the stalls are in a boring even line and equidistant from each-other. How do I fix that to make it more pleasing to the eye? What if I raised some of the stalls on a platform? How does that impact the AI who doesn't like the small step? x1000 other questions a level artist has to consider when building the environment. Clarity, first impressions, flow, environmental storytelling (what happened here), does it look lived-in, etc. All very important. And then there is all the re-doing of all of the above, if needed.
Composition and visual storytelling is
VERY important. And it's a
LOT of work if you're trying to raise the bar high. You just can't download that part of creating an environment like you can an asset. It's a big reason why 99% of asset-flips on steam look terrible, even though the quality of individual assets might be okay. Because little thought is put into composition and visual storytelling. Where you end up with something that has no consistency, clashes with itself, is visually too noisy (random detail in the wrong places), looks empty yet runs poorly, etc.
Level artists usually have help from concept artists who do paintovers or photobashes (collages from photos) for what the layout of key areas should look like. The background of an important cinematic for example. But there isn't enough time or budget for concept art to be done of every angle of an environment, so the level artist usually has to do most of the visual storytelling/composition work themselves.
Oh, also, there are the tech limitations you have to consider at all times. is there enough occlusion in the terrain or architecture? What if an area up ahead lags a bit, should I bottleneck the gates leading up to it? How do I do that, by changing the model of the gate a bit? Or stacking crates at eye-level on the side of the gate? What if an area needs to not have the framerate drop below 40 because it's a boss fight? New considerations have to be met. Then there's memory. The biome assets I use in one area use 8 2k ground textures but the next area is running out of memory. Can I sacrifice one of my 8 ground textures? This gravel looks too similar to this small pebble texture, I think I can substitute one for the other. Better start repainting parts of my terrain. You're also probably in charge of collision meshes and making all that respect the game physics.
MOST importantly, most of this has a subconscious effect on the player. They're enticed and guided to go in certain places without realizing it. They enter an area and are in awe with how beautiful it looks. Half of the reason is the fidelity of the assets, yes, but the other half is the
visual storytelling, the
framing, the good use of
negative space, etc.
The AAA graphics pipeline has changed dramatically since 10-15 years ago. Now, level artists do far less modeling and texture art, and focus more on composition, layout and visual storytelling. The reason is simply because the fidelity of individual assets now is way too high to be done by one person, so you have teams of modelers doing their magic in zbrush, maya, max and substance painter. The assets are then given to the level artists (this is an oversimplification, there are many more steps before this)... who push the visual storytelling I described earlier to the highest level they can muster, all while respecting technical limitations and gameplay vision.
This is where using existing assets can be a benefit. If one understands that composition and visual storytelling (without needing to actually create the assets) is just as much work in modern games, as modeling/texturing the level as well as doing the composition stuff was, back then. The difference comes with how much higher the bar is now, as well as the object density. How much more
involved the player is with the environment the player is in new games, vs older games where the artist did everything.
The positive reaction a player gets when playing a game comes from a whole lot more than just the fidelity of assets, or the stability/featureset of an engine. It takes an enormous amount of work, depending on how you define your "high end" bar. Which is why I asked what your benchmark was. It can be done by one person, if you properly define "the scope of your game, and what "high end" means. You're never going to make a GTA V-like game by yourself, even if you can find every single asset the game needs, freely available, as well as the perfect engine for it. You might be able to make "a map the size of GTA", with "GTA-like assets in it", but it won't feel very good to the player. It won't feel "high end".
Because you won't have spent the thousands of manhours into considering every aspect, every reveal, every angle the player can see a landmark, how the landmarks are framed, what stories the environment tells (even from existing assets), the proper balance of negative space and highlighting important gameplay elements, while respecting technical constraints, all resulting in the Lump Sum of the player uttering "wow, this game feels high end!" when entering the area. And that's just for one department (environment art), out of many that game dev involves.
But if you're just making a single-room party-game with the same visual fidelity of AAA, then yes. A single individual could do that. It's in large part about scope, which needs to be defined.
I'm only scratching the surface.
