[UE4] Physically Based 3D Art: 2016
interpolator
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Edit: As of June 29th, 2016, this thread will no longer be updated with new art. I will be rehosting its contents outside the site (to be determined at a later date).
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Next month will mark 1 year since I made the first major push into stylized art and despite some pitfalls on the way, I have a better understanding now of what kind of art I want to make.
Right now, I've gone many sleepless nights putting in more effort and detail into some new props I'm working on, so I will be posting them as they come along.
Edit: As of June 29th, 2016, this thread will no longer be updated with new art. I will be rehosting its contents outside the site (to be determined at a later date).
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Next month will mark 1 year since I made the first major push into stylized art and despite some pitfalls on the way, I have a better understanding now of what kind of art I want to make.
Right now, I've gone many sleepless nights putting in more effort and detail into some new props I'm working on, so I will be posting them as they come along.
This discussion has been closed.
Replies
It's like explaining to someone last gen why the Wii couldn't get PS3/360 games. Nobody was saying the art wasn't good enough. Everyone was saying it wasn't powerful enough.
Before I was like.. wth is this guy talking about?!
But now that you explained yourself it all makes sense now!
Now 4 years later and it's "your art is too good. downgrade it for real time".
Yooooooooooooooooooooo. But I like this. It's mean I've finally done something.
I know there are plenty of cool people on Polycount, and those who just seek to post art or help others with art. I applaud you. So you have to understand it makes me feel a bit sad inside when I see a thread like this just get derailed into this boring CG vs game argument. All the time I spent today arguing this could have been used to make 2 or 3 props I'm working on.
So please believe me, I do not like this. I don't like having to explain everything like this. I just want to make art like everyone. Something being CG, real time, cartoon, whatever. It shouldn't matter.
People are trying to help you get a better understanding of the 3D medium, you're at a stage where you need that. As i said before, fully understanding your choices is a vital skill. If the choices seem weird to people who are far more experienced than you, they will try and help you get a better grasp of why that is. Ignore it if you wish, thats up to you. Depends how much you want to learn, really.
We're just trying to correct the misinfo. We're not saying game engines are always better, we're not saying offline renderers are useless. We're just saying, if you think game engines aren't powerful enough for your current needs, you're just wrong.
And you're choosing to comment here instead of working on those props. You could just as easily stop typing, and start modelling, and post it instead of commenting. I understand the compulsion to reply, but no one is forcing you.
Just realize, if this many people are frustrated at your reasoning, it's probably because your reasoning is incorrect. A humble person accepts that, and seeks to learn it for himself.
I feel like I've been intentionally lead to feel like I have no idea about anything. Just because this forum is focused on real time, doesn't mean we can spread misinformation about offline rendering. Just as how if I went to a TV/film forum and they said something wrong about how games work, doesn't mean I just believe it because that's what they say.
Both Danr and Magihat seemed to have missed the point of the chart. That was made by Nvidia. A company who builds GPU's for gaming and is also one of the biggest vendors for VFX as well. It wouldn't make sense for them to lie about their own products. I'm sure they've done their research, and are well aware of the huge differences that comes with rendering time versus the amount power a single GPU can do. 3 years ago is not even a long time ago. That's not enough time to seal the huge computational differences that VFX has invested all its money on trying to bring about the next big thing. Even if visual fidelity has gotten closer, that doesn't change that there are still hours and hours needed to create demanding CG that games aren't going to achieve in an instant today. I've seen it myself, at both studios who still rely on rendering to create images for TV, as well as having just done a render in 3DS Max that takes hours to complete.
There are hundreds of tech papers out there. It's science to say there are plenty of reasons you can't do such and such in real time. I would never believe people wasted research on this, if the answer was "games can already do this".
I never ever once said UE4 was completely incapable. I want to make art with it as it's the only real time render I use. But I'm not going to believe its the only tool out there and it's without any faults. There are clearly things it can't do and I posted since the beginning, I wanted to try offline to not deal with those limits. If you don't believe there was any argument, why were people attempting to refute this fact of trying this?
My one hope is that there is someone on Polycount who is reading this and can understand I'm speaking truth. "Hey, JordanN isn't crazy. He tried to back everything he said with real statements". Wanting to learn or become better at 3D doesn't mean I should deny my own well being.
What do you think my needs are? I've established I wanted to do what a game engine couldn't do. It would sound like a huge contradiction if I said I wanted do something impossible in UE4. Like wanting to drive a car that can't move.
Nobody misunderstood the chart or just doesn't get what you're saying. The argument people are making is that you haven't given any real substance for using offline, and that making that switch won't help your art in the slightest because the work that you've shown would look virtually identical in any renderer. Historically you have not listened to feedback or taken it well at all. You're hellbent on doing things your way without being able to communicate a direction.
I think the source of these complaints is that you come across as treating offline like it's a crutch to improve your work. You did the same with PBR when you fought with people in your last thread. This will ultimately fail, just like how many people jumped to modern PBR workflows and still made sub-par art. A pretty renderer is either a bandage or a booster and for you it is unfortunately the former.
And not that I really care to get into an argument over the minutiae of realtime rendering, but from raytracing to millions of polygons, every example of the undoable you've given have actually come to exist in just the past 3 years. Unless you can actually back those claims up, I wouldn't make sweeping statements you seem unqualified to make. Especially since NVIDIA, who's chart you trust so much, have been at the forefront of pushing the things you claim can't be done... 2 years ago:
https://blogs.nvidia.com/blog/2014/09/18/debunked/
So, what features of an offline renderer do you really need? What cannot be done? How will this decision make your work better? I could think of many concrete examples just off the nose, but over the last page you've given us a 3 year old chart and some nebulous stuff about tflops.
Lastly, you began to actually post some decent content with the interior scene but now it's just back to the same old stuff that we've seen in every one of your threads for years. Clashing styles and all over the place quality, gross misunderstandings and misuse of terminology, arguing with more good feedback than you honestly deserve, and a stubbornness to do it your own while damning anything else.
I stated CG was more powerful.
Edit: I think I'm done arguing for the day. I've defended myself many times while being open minded to everyone's opinions that have come by. I think the time should now be used for making props now.
I saw your original 2D art thread a while ago and though it looked fine. People were critiquing a lot of stupid things in my opinion. I though it looked interesting. I even skimmed this seeing some negative stuff here and there and I thought the artwork looks cool, so I went back and re-read. I once was making a cartoon style pixar game. I've also had many projects over the last 10 years losing some focus and what not.
I say this as a graphics programmer who has worked for EA. I've read a lot of papers. You titled your thread in gigantic letters as if this is a final project/ thesis. There is no conclusion here. Everyone is creating/rendering physically based artwork, so why is your's in bold gigantic letters as if what you are doing is unique? It is called PBR, everyone does it. All you have done is build some chairs and props that seem to fit a cartoon world.
The most confusing thing here is that you completely changed your requirements, and very "stupidly" if I can say so. "My cars are too high poly and I want real-time reflections and super-high res light maps, so therefore I'm just going to make an animation instead." Ok......do you want to make a game or an animated movie...., very odd choice to change. You can still take all that and look super good in Unreal 4. You might lose a couple pixels.....
[quote](i.e I wanted to have the cars completely reflect each other as they pass by)[/quote]
Who cares...? You can still do this in real-time. You could do it for the closest car instead of all, or just using static cube-maps works pretty well.
[quote]
I want to make game art, but I don't believe that should mean I ignore evidence of offline rendering being more powerful.
[/quote]
Actually I disagree 100%. You should ignore all mediums. I do. Everyone here does. We make games. That is our requirement. Also, offline rendering isn't that much more powerful. What techniques do you really need? What are you doing that doesn't allow real-time or to down-tune a setting so it doesn't look 100% raytraced. If you are going to complain about light map resolution then just buy a GTX 1080 and bump your Horizon based AO to ultra.
You have a perfectly legit project, but if you are going to just build a non-real-time environment then what is the end goal. Either you build a game, or a pre-rendered animation. If you are passionate about making games, then make a game. To give up making a game because of light maps is ......crazy.
In the last 2 or 3 months it sounds like you are switching between projects so often that I don't even see your dedication in completing something.
Also, why is this labeled as hand-drawn? Just confused.
First of, Im starting to see a more mature Jordan here, sure, there is still work to do, but he clearly tries to stay away from pointless arguing. we should encourage this more guys, would hate to see him turn into another Valuemeal.
also, with that said, I am aware that Im doing exactly the opposite of that just now by continuing the argument xD
@JordanN - I agree with most people here, there is sooo much you can do with realtime rendering nowadays and of course we as a game-art forum is passionate about exactly that. the years we spent learning skills to make a few polygons look like thousands, with the help of normal mapping and masterfull topology at the exact right place is a big part of what the polycount community is. so yeah, we can get very passionate about it.
Saying more polygons/render passes = better art is probably more or less blasphemy around here, and I hope you can see how that makes people upset (including me for that matter).
I truly believe a good artist can make amazing art with whatever restrictions handed to them (and in this case not that many, UE4 is stupidly powerful). and I think what people are trying to say here is your still learning, and saying you'l make better art with an offline renderer shows lack of understanding on a fundamental level, this is whats "triggering" people, not really ofline vs realtime. and of course non-realtime rendering is more powerfull, I dont think anyone is arguing against that.
anyway, thats my rant, in the end I think your doing good work here JordanN. just keep going and please dont switch from UE4 just yet. I can assure you if baking times are to long, or you dont feel the quality is good enough, people here will gladly help you optimize or give you advice on what settings and the likes you can use to make beautifull yet more effichent art.
You are right in saying that you are never going to achieve all the fidelity as an offline renderer. However you can get pretty damned close though with all the shortcuts but that takes knowledge and time. That fidelity you get with offline is not going to make your work look better. It might make the path easier in some ways but it won't improve the art.
It's tougher to work with realtime restrictions but that's hell of a lot of the job working in the games industry. Your art is useless in the industry unless it's performant.
Where do you want to work?
If you want to work in the games industry then you need to make that leap, invest in the knowledge and the methods which take extra time.
If you want to work the the CG industry then you need to make a different leap - upping your fidelity within your assets which takes time.
When you have your answer to that question - either way you have gone you need to continue to chase that improvement of your art at its most fundamental. That's why people are suggesting to put aside fancy rendering and rushing to the final product. Take some time to make a single asset or small environment. Take care over it, get feedback, improve it. It's a slow process but your artistic judgement, your fundamental skills will improve.
Flash forward to Polycount now, I started posting new props but it was becoming obvious I wanted to focus on drawing more stuff instead of showing more in-engine stuff. I had begin making other new projects for UE4 instead but I didn't want the cartoons to be tied to this one.
TLDR: Original thread wasn't suppose to be game art. UE4 was used to preview what my drawings could look like in 3D but it was not going to be permanent. The next scene I began moving the cartoons to CGI, and wanted to replace it with real game props for UE4.
I realize though this means I wont be able to post CG stuff on here, but I will make another thread that contains actual game props.
But as I said, I had always been making other props that I do want to show in games.
This is my own point of view about this whole thing.
I'm just going assume you don't have a games job or any other job for that matter and I'm also going to assume you want one (be it in games or other 3d area).
Now, if that is the case, then you need to start working towards a portfolio that can show your skills in a why that attracts employers.
Because, this -> (me pointing at the whole tread) this is nothing. It shows nothing. I wouldn't hire you.
And you might get all defensive and say you're doing what you like and enjoy, and that's all fine and dandy if you say to me, "Oh, I'm just doing this for fun, I don't want to be a professional in this area."
If that is the case, then fine. Just get a job elsewhere. Doing whatever. And keep posting your cartoons. Because then we won't fall on you as we do now because we know you have no desire to make a career out of this.
Otherwise, man up. Realize you are no where near the quality required by an employer and that you don't know as much as you might think you do and start grinding.
Good luck.
The other thing you really need to think about in using an offline render, while yes they can provide better raytracing and all. But all the the high poly objects and raytracing in the world won't make up for other inadequacies. None of what I'm seeing would make an appreciable difference using an offline renderer. Sure, your shadows will be a little more crisp, but the overall composition won't change much, better materials would definitely make a much larger impact than an offline renderer ever could. Take this:
That is the first thing I ever modeled (done about 2 years ago), rendered in Vray really quick. Lighting quality is great, but with a subpar model it really doesn't make it look much more palatable.
Next thing I noticed, was for a UE4 scene your objects were exceptionally high poly! Like WAYYYYY high poly... I understand it's meant for still shots, but unless that object is meant to be right next to the camera (only reason you'd want something to be reasonably high poly in a real-time shot) there should only be enough edges to carry the silhouette with the frame of mind of how close you want that object to be to the camera. Example:
Both the same object just taking up different amounts of screen real-estate. Obviously, the jagged edges are significantly less noticeable on the smaller one, if the camera (or player if it will be in realtime) isn't going to have their nose pressed against it, it really shouldn't have that many poly's, so that is a design decision you should actively be making.
Now you can take all of this with a grain of salt, I do work in the industry but not in an art-related position (QA) but I've been trying to work towards that for about 2 years now getting my own portfolio ready and I feel like I've learned a thing or two but I hope this all helps.
On a side note, your drawings look good, as do the cars you modeled for that scene. Just keep up with those and push the materials as much as possible and it'll turn out well.
~~~~~~~~~
Other than that, I still believe that working on these cartoons served a purpose. I wanted to have a project that more or less gave me creative control over everything. That meant the style, modeling, textures, characters, plots, scenario etc. In today's world where risk and originality often go hand in hand, I wanted to do something that was free from being completely mainstream or pressure.
I understand this move doesn't play too well with games, so this I why I'm excited now to have moved this outside. I still get to keep the plants I created or the colorful cars. All the ideas stay with me, while I do create other things that are more game art friendly.