Graphics/Animations question for the programmers out there.

Gyva02

I really don't know anything about the programming end of things when it comes to these games. So my question is. 

Have we come far enough now that it is possible to change out and replace graphics when the time comes to do so? I remember asking in EQ threads why don't we just keep everything the same and just make the game look more current. I figured all you would have to do is change out the skins and textures on everything.

The replies I got back though, were people saying the code would not allow this to happen so we were stuck with the older textures and skins. I'm curious if Pantheon's code will be able to be updated in the future to keep up with the times if we find ourselves still plaything this a decade or even longer from now.

Again graphics are in no way a game breaker for me Pantheon looks great with what I have seen so far, just wondering if this would be possible in the future. 

Rattenmann

Sadly, no.

Better optics are usually gained by increasing textures / making better. They still need to fit the 3d model tho,... so that is a hard limit. Redoing the 3d model means EVERY animation has to be redone as well. And then we need the doublechecking if every armor / weapon / whatever plays along nicely.

Also the 3d work is usually the most expensive work and you can barly, if ever, sneak in "some" improvments on the fly if you have a day spare time. You could do that with many other game aspects / bugs.

delete5230

I don't care about graphics at all, just give me good game play......Just saying

Rattenmann

delete5230 said:

I don't care about graphics at all, just give me good game play......Just saying

I mostly agree. 
Yet it is becoming harder and harder to log into EQ1 for me. Most of that is due to it having changed too much for my taste. But graphics is what keeps my GF from trying it out, which in return would make me keep playing and not just checking it out every few months. 

That being said: Pantheon graphics will be fine for my taste for at least 20 years, so ill be set.

Deivos

Rattenmann said:

Sadly, no.

Actually It's been done in the past with games like Dark Age of Camelot. Mythic had launched that game on the NetImmerse engine. They did an updrage to that engine that brought with it rendering enhancements. They then went on to do a complete engine replacement into the Gamebryo engine that also allowed them to introduce replacements of a lot of the game's assets.

Another example would be the upcoming engine upgrade for APB:R as they are moving to a later version of the Unreal 3 engine. The shift is coming with more rendering and performance upgrades rather than asset upgrades though.

The Forgelite engine that SOE/Daybreak developed is actually forward compatible. The engine has a bunch of shaders and graphics settings that were developed and then basically cut off alongside more dynamic memory and CPU usage. It's an example of a system that's intended to open up features as hardware progresses to more seamlessly allow improvement in a game's visual quality over time.

The primary problem is that most game engines are not scalable. Generally there is very little practical purpose in designing an engine that relies on hardware that doesn't exist, as it introduces a lot of problems with compatibility unless you focus on the few consistent parts of a system.

On a technical and somewhat semantic level, the graphics are upgradable in most any game. The hitch is the amount of effort it takes and just how much of the game/engine you need to rebuild in order to support the change. Eventually it crosses the point where the engine limitations are too much and you are basically just making a new game.

The amount of effort as such, is generally not deemed a good investment.

Ridelynn

Depends entirely on the engine being used on the back end. Most MMO engines are optimized for a high variance in models and textures - the need to be able to call up any of hundres, maybe thousands of different models and textures on demand. Think of walking around a hub city, and every time a new player with different armor walks by, that's a (possibly) new model + texture.

That's a lot different than a FPS or on-rails game, where you can control exactly what models are shown when, and know when to cache them (and when not to). So you can make them a lot more detailed.

Wizardry

Well you do not need to change models at all,we cover them up with gear anyhow.
A quite acceptable change would be to use better textures and just make surfaces look a lot better.

Unreal T back in 99 released with an added hi res texture pack and it was quite noticeable the improvement.I tried before but couldn't find any videos that would do it justice,you need to see yourself in game and that was back in 99 when texture sizes were likely max of 640x640.

Truth is developers don't see any profits in it,you update graphics where do you get the money for that?a Smart dev would think ok we are giving back to the players who supported us but nope they are greedy as heck,each CEO and big suit sees $$$ and quarterly bonuses.

A perfect example,Square said they couldn't do it ,too costly ,when FFXI fans were asking for it,yet here we are they are rebooting every single past FF game with updated graphics.So they lied they will do it if they see the profits from doing it.people don't even realize what FFXIV really is,it is a just a scramble of FFXI with better graphics  to make it look like a new game to warrant a fresh new price.

Quizzical

It's always possible to improve graphics.  It's only a question of how hard it is.  And the answer to that on an already released game is usually, hard enough that the return doesn't justify the cost.

Making a lot of good quality artwork is expensive, and for an already released game, you've already paid that cost.  Indeed, it was a considerable portion of the cost of developing the game in the first place.  Throwing out what you've already paid for to replace it by something better is expensive, and generally won't gain you much.  Replacing or fixing a few egregiously bad models is very doable, but a wholesale replacement of a broad swath of the game just isn't viable.

Deivos

Oh, forgot to mention that Asheron's Call even did a graphic upgrade at one point.

They added necks to the player models. :awesome: 

Ridelynn

Rattenmann said:

delete5230 said:

I don't care about graphics at all, just give me good game play......Just saying

I mostly agree. 
Yet it is becoming harder and harder to log into EQ1 for me. Most of that is due to it having changed too much for my taste. But graphics is what keeps my GF from trying it out, which in return would make me keep playing and not just checking it out every few months. 

That being said: Pantheon graphics will be fine for my taste for at least 20 years, so ill be set.

I would almost still be playing EQ if they would fix their graphics engine. Having my computer come to a complete stand still slide show every time I panned across more than 30 people at once was a severe pet peeve of mine. 

I could really care less if they looked like EQ1, or like stick figures, I just hated the hitches.

Amathe

I remember when EQ did a graphics update. I read that WoW was going to do one. 

Rattenmann

Don'T get me wrong. Updates CAN be done (obviously everything can be done in coding), but the price is considered too high to be worth it. Yes it has been done, no it did not pay off for them.

Usually something like this is only needed after a few years have passed anyways and no game is getting big enough income that far into its timeline. WOW being the exeption.

Unless Pantheon is going to be a surprize hit and smashes all the charts... it is not going to happen. 

Gyva02

Thanks everyone for your replies,

I was hoping with today's engines you could just copy and paste new code to fill spots around the world. Like if we had slots for every object. For example if we had tree001 and all objects had a specific pixel size box that could be filled (Length Width Height) you could just replace it with an updated tree of that same size (to not interfere with things near it in the world).

Once you made the new tree code and placed it into the tre001 slot, all the trees in the world that had a label of tree001 would change to the updated version. As you did this to all the objects in the world your game would slowly be getting updated... Just need to make everything have a slot that you can quickly delete and replace the code for, but I'm assuming a game engine would have to be designed just for a game like this. But if it was possible, your game would always stay current and possibly even done freely if the players wanted to take it on as a labor of love. I know if this was possible in EQ the players would have already done so. Just a thought I guess but then again, I don't really know how codes and engines all jive together  

xm522

From my limited experience with 3D development and graphics engine I would say that it is possible to transition to a more powerful engine that provides some improved visuals. On the other hand most improvements would be related to lighting and effects, if you want improved environment and improved characters you will need to create new assets for all those objects. 
Again 3D modelling and animation is not my field, so this is almost speculation. 

Dullahan

Gyva02 said:

Thanks everyone for your replies,

I was hoping with today's engines you could just copy and paste new code to fill spots around the world. Like if we had slots for every object. For example if we had tree001 and all objects had a specific pixel size box that could be filled (Length Width Height) you could just replace it with an updated tree of that same size (to not interfere with things near it in the world).

Once you made the new tree code and placed it into the tre001 slot, all the trees in the world that had a label of tree001 would change to the updated version. As you did this to all the objects in the world your game would slowly be getting updated... Just need to make everything have a slot that you can quickly delete and replace the code for, but I'm assuming a game engine would have to be designed just for a game like this. But if it was possible, your game would always stay current and possibly even done freely if the players wanted to take it on as a labor of love. I know if this was possible in EQ the players would have already done so. Just a thought I guess but then again, I don't really know how codes and engines all jive together  

This is very possible with new engines. Maybe more problematic in the past. Now you could change the texture or the entire mesh of an object.

I don't know all the limitations, but I could see someone using Unity and having the ability to overhaul the graphics much easier 5 years after launch. I mean, just since they started Pantheon, Unity launched a new version which added some significant graphical improvements. Permitted the engine is still supported, they will have the benefit of any changes made to it. Even if it wasn't supported, the engine is so modular that you can tweak everything yourself.

Montreseur

Art production for large high fidelity games is already man hour intensive.

Just wanted to drop some information here on the process to give you guys a perspective on what it would take to "switch textures or meshes".


 In most cases every single game object has a High Polygon model, which is usually at least a few hundred thousand polys, up to a few million for each mesh. The high poly model never  makes it into the game, but its detail is projected into various texture "maps" (Normal Map, Ambient Occlusion, curvature, thickness, color id, etc). From here we retopologize (create a very, very low polygon version) this is usually just a silhouette of the high poly mesh; this is the model that makes it into the game. Another mesh is generated in some cases (a cage) to help the detail project properly.



From here we can start texturing. Long ago many techniques were invented for faking detail. Every object you see in a game has several texture maps laying on top of one another that the engine reads differently and adds together for the final render. The typical maps can be seen in the image below-



These are your typical Physically Based Rendering maps, which is the new standard model most next gen games use. So on top of your projected maps from the high poly(which are used in the texturing process), you generate the above with your final texture, and these are all wrapped onto the meshes. There are other maps that we generate in special occasions (emissive, opacity, translucency, height, displacement.) 

As an extra layer on top of this, objects need to be placed and optimized. This usually comes down to extra planning for modularity, combining large sections of smaller meshes, and created simpler and simpler meshes for LOD fading (progressively lower texture resolution, and lower polygon counts, all the way down to a "billboard" at the farthest distance).

Lokero

Modelling has been covered pretty well, but I thought I would point out that even Unity allows animation copying, so long as the skeletons/avatars are the same basic structure(humanoid, for example).  If you are redesigning the model, the skeleton is likely to remain the same, and so the animations should(in theory) be transferable.

So, I don't think animations would be too huge of a problem as say adding more geometry to a character model.