Looks Sweet. Your Texture quality has come a really long way.
A few ideas:
- Do you have an Anti-Aliasing filter applied or the Camera
- Might look cool with some SSAO to help ground a few objects to there surfaces
- More Dynamic Lighting perhaps, I notice in areas such as under the stairs are still receiving as much light as under the table, as in the shadows and light intensity are the same
Thanks Carl! To address your bullets:
I feel stupid but after I got all my shots taken and prettied up for my portfolio I noticed my camera didn't have AA activated. I will eventually go back and fix it but I am putting all tweaks on the backburner for now so I can roll right into my next project.
I hadn't even considered SSAO. I have no idea how to do that in Unity, but I guess Google is a thing. I'll add this to my backburner tickets (I use asana for project management).
Yeah, I'm not the best at lighting obviously; there are glaring issues with what I have currently, again all to be fixed later. I think my ambient light is just too bright, and I also didn't tweak the sky color in the lightmapping options. I will tinker with lowering that and throwing in some more subtle omni lights all over.
And on that note, here's a refreshing change of style! I'll be using these for my next project, which I will talk more about at a later time.
Sweet man! I really happy to see you finish this project, great follow through! There are always things that you can improve, but that's not really important at the moment; you got it out the door and learned from it.
I would say that in the future, try out UDK or Cry. The advantage of working within standard engines is the amount of docs. community, tutorials, etc. They're also easier to work in and give you a better understanding of foundation stuffs.
As far as the rocks go, I would try to make them less blob like; and I wouldn't worry about the cracks and little details, (they can be put in in photoshop, crazy bump, etc.) Worry about the over all shape and hardness of the rocks.
Sweet man! I really happy to see you finish this project, great follow through! There are always things that you can improve, but that's not really important at the moment; you got it out the door and learned from it.
I would say that in the future, try out UDK or Cry. The advantage of working within standard engines is the amount of docs. community, tutorials, etc. They're also easier to work in and give you a better understanding of foundation stuffs.
As far as the rocks go, I would try to make them less blob like; and I wouldn't worry about the cracks and little details, (they can be put in in photoshop, crazy bump, etc.) Worry about the over all shape and hardness of the rocks.
Thanks! Yeah, follow-through is definitely important! I have a number of adjustments I will follow up with at a later time, probably near the end of my break if this piece is even still in my portfolio by then. I really appreciate your feedback you gave while I was working on it, it helped out a lot.
I was originally going to have the tavern in UDK, however some problems early on forced me to change my plans. Now with UE4 being out and looking much more user friendly, I will be less hesitant to use it for one of my projects.
Again, thank you for the advice and imagery for improving my stone sculpts! I haven't really done rocks before so I expect my sculpts to start terrible and be much better by the end of the project, just like with my hand painted stuff. I have been checking out the Rawk thread and have found some good resources and tips there which I will be exercising soon.
I'm still trying to keep up on daily hand-painting studies despite my next project being closer to photorealistic. I tried to paint some metal plates today. I think they would have come out better if my features were larger, such as the edge highlights, scrapes, and general panel thickness.
Hopefully I can get a pretty nice process down for stone. I'm using kind of a combination of ZombieWells' tips and some others that I came across in the Rawk thread. In the rock below, I was toying around mostly with TrimSmoothBorder with a square alpha and TrimDynamic with a "rocks" brush alpha that I got from ZBrushCentral. I think it adds a nice texture with no effort and not a terrible amount of repetition.
I would have gotten more done today but I had to stop half way through my day to prepare some asset store stuff.
One of two or three boulders I'll be using to generate most of the walls and ceilings. I got the general detail down to make sure the normal map baked out alright. Once I go back to texture it, I will also add in more cracks and marks. I'm still trying to get away from the blobby feel. I might end up not even using this one.
I wasn't satisfied with my last rock so I decided to make another. The previous boulder had too many sharp-ish edges that felt artificial. This one looked much better so I textured it up. No spec yet but I think it's a pretty good base.
More rock stuff. Here's a first pass on a chunkier texture. I intend on coming back around and adding in moss and other grimy buildup. This was literally a 10 minute job after I got all the maps baked out. I think the longest part was making the rock base seamless. I included a gif showing the texture breakdown.
I am going to hit this real quick, (I just don't have the time currently) your rocks are way too soft, and, what I see is a very common error in making rocks... have now become mashed potatoes. I have also, never seen a rocks look that muddy texture wise. I looks to me like your guessing, (common error) at what a rocks looks like, and not using reference. Your stair case deals are way too nosy and read like a choppy mess. With any art your looking for quite places your eyes can rest.
There are rounded rocks, that have been worn by rivers and such...So I don't know what your going for- post your reference!
Your low poly for that pillar is way too high. you shouldn't be cutting in those impressions, let your normal do its job. More on the way to why these things are a bit strange.
Thanks again for your continuing critiques ZombieWells, they definitely help me out.
The brazier did indeed have some excessive polygons so I went in and cleaned those up.
I decided not to remove the lower indentations simply due to those spots being too flat which the normal map can't make up for. I think I can still cut out some of those small triangles along the indent corners, but I definitely need to keep that little slope.
Alright, on to those mashed potatoes! The textures I'm working on now are meant for underground cave systems, not so much a shattered, jagged pile of rubble like in your photograph. I'm trying to make similar textures present in Skyrim's caves. This particular texture I've made will be used on stony formations that protrude from the walls and ceilings, as seen below in the shot straight out of Skyrim.
Despite Skyrim having a realistic appearance, there does seem to be a subtle amount of stylization going on so I have to improvise a bit. I'm using what I see in Skyrim as a reference, and realistic boudlers as a starting point. I've heard I shouldn't using other art as a reference, but I don't know a better way to match a game's art style.
The "Flat" one is 722 and the "Nice" one is 1278. There will be exactly 10 in the scene which means the nice ones cost ~5000 more tris than the flat ones.
okay, so now I want you to think about this. From the perspective your showing me that nice one is only "nice" from the left side, looking at the front end... its doesn't make any difference. Now look at that brazier if were a player that is standing up. How much of a difference is there between the two really? almost double the tri count difference? Now take all that geo you just saved and put it into those rocks, the bowl, or something a bit more important than cuts in the bottom of a brazier. Unless the brazier is HUGE and the player is really seeing it from that perspective, or if this is a show piece... I am assuming that its not, then its not worth it... Also when your making these things, it always pays to plan a head and think about the low, how it's going to cook down from the hi, and the impact on the player... order of importance really. I am sure, that the hi could have been made with just as much artistic impact, with less need for geo in the low. Meaning that it could have been one big dip instead of five.
Skyrim- I love Skyrim, one of the greats, played the hell out of it. Still, you must remeber, that although those guys did an incredible job on all the terrain, etc. The world was Gigantic! So they didn't have a lot of textures or geo room to move around in. I think you have a bit more wiggle room to make more defined and unique meshes... so look at real world caves and other ref. and textures etc.
Yeah, that's a good point. Those indentations don't help the silhouette any either. I guess I'll leave them out then! On that note...
I still have a hard time drawing the line between how low to really make in-game assets. Obviously it depends on the platform and other variables, but a few of my buddies received critiques at GDC about how some of their models were too low poly for current/next gen, and then there are games like Star Citizen where I'm hearing the in-game ship models clock in around or above 200K triangles.
It makes sense to cut as much out as possible for mobile platforms because of limited rendering capabilities; but with PC, what's really the deciding factor on polygon budgeting? I've been reading that most performance issues nowadays are related to hitting a memory ceiling/draw calls rather than a polygon cap. If occlusion culling and all that cuts out a lot of rendering (especially in enclosed areas like caves/tunnels), when does it start to become a problem?
I think it came out noisier than expected. It kind of reminds me of sun-dried mud. The larger rocks probably won't tile so well across large expanses either so I'll probably re-bake without them. Any other crits are welcome, as usual.
models were too low poly for current/next gen- I am guessing that they butchered their silhouettes and used low texture sizes. If you have any examples I could tell you for sure. IF your making a show piece, then throwing a bit of extra geo, or upping the texture size is fine. Just don't chuck polys at something and call it hi res... what happened here is that you chucked a bunch of polys at something very simple, and at some details that didn't add too much to the over all piece. If you modeled some sick eagles coming out of the corners or something like that, fine bump them out dude! that's worth more geo.
polygon budgeting?- Your going to have limits on budgets depending on what your working on that's a given. The thing to keep in mind is no matter what your making, try to be as efficient as possible. Silhouette, player view, size, use, impact. Make it look kickass, but do it in a smart way!
(Those bold lines in my shot are UV seams, no wacky double edges going on or anything)
I definitely understand the importance of keeping stuff as low as possible, however my decisions aren't haphazard. I am adding a more than usual amount of topology in my low poly models for a few reasons. This project is another piece for my portfolio and, although the brazier is a pretty unimportant asset, its place is in the main hall of a room which will be the primary focal point of the project. If I decide to take a beauty shot, lighting will play off those indents far better than if they were cut down as flat as possible, relying only on normal maps (I would add more extravagant features, however I'm attempting to stay true to the original idea). I'm also pretty confident these extra verts (across all my assets) will hardly impact performance, if at all; but if it comes to that, it's much more convenient for me to cut polygons out of my LP than it would be for me to reopen my HP source file to rebuild and readjust UVs to add in more detail.
My shots of the brazier probably don't do me justice. I did say in the first one that I had an excessive poly count that I needed to clean up. I hope I don't sound like I'm discounting your advice and knowledge. On the contrary, I very much appreciate and look forward to your responses.
So here's what I've been putting together today. A few modular platforms and a pillar-type of object. Absolutely not final versions. I know they are pretty bland right now, especially with that unfinished ground texture. I like the way the one with the brick flooring is coming along. I still need to fix up the textures as well as add in foliage such as plants and hanging moss and such to add some life. There will also be some alchemical reagents added here and there (gratuitous glowing mushrooms), but since those assets can't really be changed without confusing the player, I won't be modeling them.
Sorry for the slow updates lately. Here's what I've been doing lately
Some seamless tiles and assets made from them
Here are most of the major platform and barrier terrain modules (sorry the shot is so far out, I'm lazy)
And here's a shot of those modules in action.
As before, textures are still WIP; I'm trying to get all the models out of the way before going back to tweak textures. I also still need to model small rubble and other debris assets to add more variation to the terrain, as well as vegetation such as hanging roots, vines, moss, ferns, and perhaps some plantlife native to the Shivering Isles.
All lighting, fog & other effects, as well as alchemy nodes, will be added in the Creation Kit.
Heyo its time for a foliage pass! Here's some small leafy twigs that I will be using for small shrubs or debris. I still need to do fern leaves, some hanging moss, and dangling roots.
These will all be baked onto a texture sheet for use on planes
All right, what are you trying to do here? Are you remaking a skyrim dungeon? are you trying to improve on a skyrim dungeon? Are you making your own? Even still, these assets need to go in an engine, or in the creation kit that comes with pc skyrim for modding, (I think your planning on do this). It is hugely important. You need to start see how your textures are coming together; where your going to add geo, composing the scene, etc. Right now I can't tell heads or tails from what I am seeing. It looks like brown blobs and contrasted mashed potatoes. Also, the geo you saved on that brazier could be used on those terrain assets, (as they look super low... we are talking lower that Dragon Age Origins, (2009)). You seem to be falling into that trap of order of importance, and not making your hi polys work well with low poly creation. Harping: the brazier was just an example of how to think about asset creation. I don't care if its 5k for your portfolio, just make it worth the 5k. Those patterns at the bottom are could be overlayed, and the supports between the gaps are really thin and look super weak. It looks like your concentrating on, what is a fundamentally the bottom of a box, on a bad decision to inset those patterns, and low balling the terrain, which is what your going to see more of.
Now getting to those rocks.
So how does look like are rock to you? Textures are ranging from purple, sick yellow, green... it looks like Stallone beat up a two pieces of living dough stuck together.
I just put in Skyrim, and I am now running around in a dungeon. Looks like they have square sharp rocks in there to me, and the flat round ones are tiled textures with bumped out geo. Look Skyrim is huge, just remember what I said about the budget... last comment I think.
Rounded rocks are what you find in river beds, here are some cave rocks... from caves.
then you have stuff like this
Now on to foliage.
I don't mind the veins,(you should put them in) in the leafs, but taking this into zbrush is a waste of time. It makes your mesh really heavy and it's hard to arrange things for the bake, (if your doing a large shrub or something similar). Once again, focus is on minor stuff. Most of that detail can be done with overlays. The cracks in the leaves, the blocky bark. Also you have to ask... what are you spending your time on? Yeah the way that branch gets really thin in the middle... ummm
Make a leaf, make a stem, group, make a bigger stem, use the smaller stem leaf groups to make the bigger one. Make alpha cards, add geo, bend, pull, push. Make a few variants, etc.
example, (this is for a tree but still) its light, easy, flexible, and does everything you need it too. and if you want to break up the silhouette, delete some groups.
All right, what are you trying to do here? Are you remaking a skyrim dungeon? are you trying to improve on a skyrim dungeon? Are you making your own? Even still, these assets need to go in an engine, or in the creation kit that comes with pc skyrim for modding, (I think your planning on do this). It is hugely important. You need to start see how your textures are coming together; where your going to add geo, composing the scene, etc. Right now I can't tell heads or tails from what I am seeing. It looks like brown blobs and contrasted mashed potatoes.
My goal is to develop a small cave system that would fit right in to Skyrim. Many of the shots I have been posting are WIPs of things I am both learning to do as I make them as well as using them in the dungeon as final assets. I was really hoping for numerous people to be jumping on me and offering critiques and such, but it's really only you (thank you) so by the time I receive any feedback, I've already moved on to another asset.
I've got the whole level already blocked in. Every asset is instanced so tweaking textures, UVs or topology doesn't cost me any time in regards to replacing modules. I can get around to that at any time, and frankly I'm leaving much of that to the polish stage. All the terrain elements use seamless textures so there is no high poly to rework. That being said, I've been taking notes on what you've been saying and I will be posting more relevant updates as I execute the changes.
Also, the geo you saved on that brazier could be used on those terrain assets, (as they look super low... we are talking lower that Dragon Age Origins, (2009)).
Yeah, I know my stuff is kind of simplistic right now. Once I get my texture scaling down pat I will be going back through during the polish phase and adding more geometry and pulling out certain areas to match the textures. This will enhance the craggy feel. (Again, all my assets are instanced so I don't lose any time doing this.)
Make a leaf, make a stem, group, make a bigger stem, use the smaller stem leaf groups to make the bigger one. Make alpha cards, add geo, bend, pull, push. Make a few variants, etc.
Yep, this is exactly what I'll be doing. The stems I posted before were WIP as usual (I fixed the thin part of the branch), but I'll be baking them down to a texture holding all my transparent assets. Nice example for the tree branches, I had planned to do something similar with the branches I have, but for small shrubs and such. What I posted before was one leaf an one stem for each of the two different leafy clusters.
At the end of the day I am a single person on a very strict and limited time frame. I don't have company-level resources or the convenience of getting handed assets to work with, or handing off my assets for them to be blocked out by someone else. Skyrim had entire teams dedicated to various portions of this pipeline. I'm making 100% of the HPs, LPs, textures, UV mapping, level design/blockout, and polish. Granted, Skyrim is much bigger than my single dungeon but I hope you see my point. My primary goal is to get all the assets developed and in place before I worry about perfect textures or perfect topology. I don't mean to come off as defensive, I'm just laying out the facts so you know where I'm coming from
Today I'm blocking in all the stray vines/exposed roots for the level.
All right, I looked back over this thread and its seems to me that you have eaten through about half your time already... was afraid of that. You have a goal; do you still want to work at one of those 3-4 studios? I have some more question actually; since the pressure is starting to become apparent, and I want to stop assuming that you know certain things.
what are you modeling in now? maya and zbrush right? and your using Unity for your engine correct?
are you using maya straight out of the box or are you using scripts and plug-in? are you still modeling your hard surface stuff in Z?
Do you know how to hi-poly model in Maya or Max?
Other than that, here's what you would want to do. Look at skyrim or whatever and make a list of assets for the area you want to make. then GRAY BOX BLOCK IT OUT and PUT IT IN AN ENGINE. Run around in it, and make sure you have your basic lighting set up and your focal point, composition, (fist pass). Once your happy with the way it looks, start picking out assets, break them down on a sheet of paper and plan out how your going to make them. What is a unique assets, whats tiled, etc. (this takes two seconds) its portfolio stuff so it can be over budget: geo, textures, make it look awesome, just do it in a smart way, and do go stupid nuts. Think about whats important, and this also translates directly to TIME, (which your are running out of). Are you going to spend a few hours screwing around with miniscule leaf details that no one will ever see in Z, or do it in PS in two sec. If you have a kickass statue that you want to spend the time really making it sweet for a show piece, spend your hours on that, I can see that happening.
Do you know how to overlay normal maps in PS? Crazy bump? Knald?
Start making your assets, you already have "hand painted" stuff on your port, so I would stick with the style your trying to get a job in. This time go borderlands or something, or not, as long as its in the same hand painted stylized vein. Also post some reference and say, "THIS IS THE STYLE I AM GOING FOR!" it helps people get exsited about what your doing because they can visulize it in their heads. Also gives you a clear focus of what your doing. Like this.
I am making this! this style, but in a Skyrim dungeon, (sounds badass to me).
REPLACE GRAY BOX WITH PROPS. tweak textures, lighting pass, (second pass) do this as necessary throughout process.
Do you know UDK?
Do you know how to make materials in UDK or Unity?
(UDK is awesome because there is a ton of information out there already. Pretty handy. I would recommend it for straight up learning)
Finnish building scene, final textures, finial lighting, (finial pass). FX pass, (not a must but nice to do).
Scene DONE.
Beauty shots. Showcase/Hero prop shots. Scene break down.
Your on a good path, keep going! Your a hard worker, your putting in the time. Its not an easy journey, or a short one; but you can make it, I wouldn't be helping out other wise.
If you have questions on how to go about making certain things... like that rock. Make a new thread, or post it on that rocks thread. Ask for help. Don't wait for help to come to you, go get it! Also, you would be surprised at who's keeping an eyes on this thread but don't comment...
what are you modeling in now? maya and zbrush right? and your using Unity for your engine correct?
are you using maya straight out of the box or are you using scripts and plug-in? are you still modeling your hard surface stuff in Z?
Do you know how to hi-poly model in Maya or Max?
Do you know how to overlay normal maps in PS? Crazy bump? Knald?
Do you know UDK?
Do you know how to make materials in UDK or Unity?
I use Maya and ZBrush primarily. I "grew up" on Unity so I haven't touched UDK.
I have buckets and buckets of scripts to supplement my modeling/unwrapping speed.
No, I'm not still modeling hard surface stuff in ZBrush
I can high poly model in Maya, yes. I'll be posting a bunch of those Weekly Hard Surface Challenge starting next week.
I know how to overlay detail normal maps on top of other normal maps.
My knowledge of node-based shaders is limited since I never used UDK. However, I did pick up Shader Forge at the beginning of my last project but I haven't had the opportunity to dive in just yet. One of the projects on my list is a Borderlands style, so that will give me the opportunity to mess around with Shader Forge.
Also post some reference and say, "THIS IS THE STYLE I AM GOING FOR!" it helps people get exsited about what your doing because they can visulize it in their heads. Also gives you a clear focus of what your doing. Like this.
Ha! I guess my delivery has been piss-poor huh? I originally started this thread to post daily updates regardless of how complete they were, sort of like a 3D sketchbook. I think I might start limiting my posts to more complete assets. For the next project I will try to make it a bit more organized as far as presentation goes.
All right, I looked back over this thread and its seems to me that you have eaten through about half your time already... was afraid of that.
Believe me, I am always conscious of the time that has elapsed. I usually take a day every so often (usually after major milestones of each project) and reflect on some stuff. Here are some things I have concluded:
So far, my projects have been far too over-scoped. I actually tried to dial back on the scope of the Skyrim mod, but it still ended up taking more time than expected (what project truly gets finished on or before its projected timeframe?)
My projects are taking longer than they should because I'm learning certain things while I'm working on the project. For instance, those mashed potato rocks. Those were some of the first assets I started making for the project as well as some of the first rock assets I've had to sculpt; and now, a month later, they are unacceptable to me so I will have to go back and replace it.
However, the things I learned taking that extra time is certainly not a waste. I've picked up new skills and techniques, learned how to do things... And how not to do things.
No more massive levels. I can present just as much style and atmosphere in a diorama than I can using an entire fleshed out level. This, along with the experience from my previous projects, will speed up my project completion rate considerably.
I don't have any real game dev experience. By that I mean, I haven't worked at a company that has a solid pipeline from start to finish. My last job had us scrambling from project to project haphazardly throwing shit together in hopes that we would make our deadlines. Certainly not something that awarded me any sort of dev experience. I knew it was a terrible method.
I'm actually not incredibly behind schedule for my sabbatical. Maybe two or three weeks. The biggest cause for this was a couple months back I made the mistake of jumping into one of the monthly environment challenges without thinking and, after a couple weeks, decided I really didn't want to do the project in the first place so I scrapped it.
Your on a good path, keep going! Your a hard worker, your putting in the time. Its not an easy journey, or a short one; but you can make it, I wouldn't be helping out other wise.
Ah, thanks man. Again, your support is much appreciated. It's definitely been slow-rolling so far but I see things speeding up soon.
Haven't posted in a bit! I have decided to start posting a little less frequently and only show stuff at major milestones. So! I decided to take ZombieWells' advice and set up a better presentation for my projects. Perhaps that will evoke more responses
Coincidentally, my next project is in fact Borderlands themed! I will be making a small diorama containing some supporting elements to help frame a centerpiece. The centerpiece is a "weapon crate" similar to all the varying kinds that present sweet loot when you open them, like this:
I noticed only Hyperion and DAHL (and Bandits I suppose) are really the only in-game companies which manufacture loot crates. I always wondered what the other companies' crates would look like, such as Vladof, Tediore, Maliwan and Jakobs.
My play style in Borderlands tends to gravitate toward the heavy-hitting weapons that Jakobs is mostly known for. As such, Jakobs is going to be my inspiration behind the weapons container I'll be making.
There isn't a whole lot of content to go off, mostly just the feel of the brand logo and the special skin that Claptrap has. However, with those two things I think it will be fun to blend an old West/Texan feel to what is foundationally a sci-fi style. I foresee lots of wooden panels and swirly leafy designs.
Here's a number of reference pictures I grabbed, mostly for mechanical functionality:
I've gotten started with a blockout for all the elements of the container. I don't want to jump right into making the high poly until I'm satisfied that all the components read well and appear functional/real.
The red color indicates compartment doors.
The green color indicates locking mechanisms.
I toyed around with the idea of having one of those big spinning handles to unlock all the compartments, but I think that is a bit too far from the general style and probably wouldn't lend well to a timely animation.
Cool man, yeah terrain stuff is an art form in it's self. Hard to do when it looks sooo damn easy... Anyhow, happy to see that your showing examples of what your going for. This should help out quite a bit. Also I might put projects like this on it's own thread with borderlands fan stuffs heading or something. It will let people know what your doing right away and not have to dig though almost a years worth of work to find something relevant. Also I looked back at your portfolio and from what I remember of what it was before, is much, much better. My only comment is that some of those edges are really tight in the HS section, for quick fix you can try to relax some of it, but I would go back an scale my edges, (something you can't do quickly in zbrush Yeah cool block out, but too early to crit!
Hey guys, sorry to just butt in but I am thankful for the thread. I am a 3d generalist making the transition to game art and am about to start a similar process.
Bartalon, seeing your process and what worked/didn't work over the past months really helps me plan out my next 6 months or so. Going through this thread shows your improvement, keep up the hard work and dedication.
ZombieWells, thanks for all your feedback. I read most of your longer post.. esp your 5/25 post about start to finish macro workflow.
Not trying to hijack at all, just encouraging you both.
ZombieWells Kind of a late response but I took your advice and made a dedicated thread for just the Borderlands project. I think from now on I'm going to post less frequently on here, but with more relevant and complete content; and post my WIP stuff in specific threads.
oxygencube Thank you! I'll keep an eye out for your stuff popping up in the forums as well
My Borderlands project is finally wrapped up, here are the shots for that:
I posted this a while back already but I forgot to post it in this progress thread. Small project that took a few days. It's the F-Ray from Futurama from S2E04.
Man, everything in that shadow box is dead on. The attention to detail is incredible, although I am surprised you opted for 3D lettering instead of a name tag. Outstanding work!
Lt_Commander: Sorry for the suuuuper late response. Thanks for the compliment! In retrospect, yeah, a name tag probably would have been more appropriate! Darn...
---
Six weeks later, I post the finished version of my latest project. You can see the development of this project in the original project thread here:
Here are some final renders of my scene, rendered in Unreal Editor 4, textured entirely with Substance Painter. You can also check out the full detailed breakdown on my portfolio site, here.
The lighting is great! I like how you divided the room into areas by lights' colours.
Objects' normals and textures are very good. Variance of materials on normal maps is well readable. I'm not sure about final materials though, because they look better in the portfolio IMO. Why are the surfaces so diffuse compared to portfolio breakdowns? Do you have enough reflection capture spheres?
Out of curiosity, because I don't know that: What is the purpose of the Albedo Color Overlay in the master material?
Avvi: Thanks! So yeah there is a definite difference between the individual shots on my folio and the final result here in the scene. I'm pretty sure I had enough reflection spheres. I had one for the whole room, then smaller ones around each light, covering those local areas.
I believe the difference here is due to the Substance Painter shots having full cube map illumination (to show the details), while in the actual scene the ambient cube map settings in the scene are very low to accommodate the dark, underground location. Although it could very well simply be my inexperience with UE4
The Albedo Color Overlay (and also the Desaturation) shifts the albedo color of some assets. In relation to my scene, I could have changed the saturation or made slight hue shifts to the bins, crates, concrete slabs, plaster walls, etc.
Replies
I feel stupid but after I got all my shots taken and prettied up for my portfolio I noticed my camera didn't have AA activated. I will eventually go back and fix it but I am putting all tweaks on the backburner for now so I can roll right into my next project.
I hadn't even considered SSAO. I have no idea how to do that in Unity, but I guess Google is a thing. I'll add this to my backburner tickets (I use asana for project management).
Yeah, I'm not the best at lighting obviously; there are glaring issues with what I have currently, again all to be fixed later. I think my ambient light is just too bright, and I also didn't tweak the sky color in the lightmapping options. I will tinker with lowering that and throwing in some more subtle omni lights all over.
And on that note, here's a refreshing change of style! I'll be using these for my next project, which I will talk more about at a later time.
I would say that in the future, try out UDK or Cry. The advantage of working within standard engines is the amount of docs. community, tutorials, etc. They're also easier to work in and give you a better understanding of foundation stuffs.
As far as the rocks go, I would try to make them less blob like; and I wouldn't worry about the cracks and little details, (they can be put in in photoshop, crazy bump, etc.) Worry about the over all shape and hardness of the rocks.
[IMG][/img]
Thanks! Yeah, follow-through is definitely important! I have a number of adjustments I will follow up with at a later time, probably near the end of my break if this piece is even still in my portfolio by then. I really appreciate your feedback you gave while I was working on it, it helped out a lot.
I was originally going to have the tavern in UDK, however some problems early on forced me to change my plans. Now with UE4 being out and looking much more user friendly, I will be less hesitant to use it for one of my projects.
Again, thank you for the advice and imagery for improving my stone sculpts! I haven't really done rocks before so I expect my sculpts to start terrible and be much better by the end of the project, just like with my hand painted stuff. I have been checking out the Rawk thread and have found some good resources and tips there which I will be exercising soon.
More updates to come
Hopefully I can get a pretty nice process down for stone. I'm using kind of a combination of ZombieWells' tips and some others that I came across in the Rawk thread. In the rock below, I was toying around mostly with TrimSmoothBorder with a square alpha and TrimDynamic with a "rocks" brush alpha that I got from ZBrushCentral. I think it adds a nice texture with no effort and not a terrible amount of repetition.
I would have gotten more done today but I had to stop half way through my day to prepare some asset store stuff.
And a carpet. Can anyone guess what it's from?
Crits are always welcome.
Here's the main post if you want more info:
http://www.polycount.com/forum/showthread.php?t=134266
I could probably clean up the wireframe considerably... Maybe I should look into using Hand Plane so I can cut out some of the supporting geometry too
There are rounded rocks, that have been worn by rivers and such...So I don't know what your going for- post your reference!
Your low poly for that pillar is way too high. you shouldn't be cutting in those impressions, let your normal do its job. More on the way to why these things are a bit strange.
The brazier did indeed have some excessive polygons so I went in and cleaned those up.
I decided not to remove the lower indentations simply due to those spots being too flat which the normal map can't make up for. I think I can still cut out some of those small triangles along the indent corners, but I definitely need to keep that little slope.
Alright, on to those mashed potatoes! The textures I'm working on now are meant for underground cave systems, not so much a shattered, jagged pile of rubble like in your photograph. I'm trying to make similar textures present in Skyrim's caves. This particular texture I've made will be used on stony formations that protrude from the walls and ceilings, as seen below in the shot straight out of Skyrim.
Despite Skyrim having a realistic appearance, there does seem to be a subtle amount of stylization going on so I have to improvise a bit. I'm using what I see in Skyrim as a reference, and realistic boudlers as a starting point. I've heard I shouldn't using other art as a reference, but I don't know a better way to match a game's art style.
Skyrim- I love Skyrim, one of the greats, played the hell out of it. Still, you must remeber, that although those guys did an incredible job on all the terrain, etc. The world was Gigantic! So they didn't have a lot of textures or geo room to move around in. I think you have a bit more wiggle room to make more defined and unique meshes... so look at real world caves and other ref. and textures etc.
I still have a hard time drawing the line between how low to really make in-game assets. Obviously it depends on the platform and other variables, but a few of my buddies received critiques at GDC about how some of their models were too low poly for current/next gen, and then there are games like Star Citizen where I'm hearing the in-game ship models clock in around or above 200K triangles.
It makes sense to cut as much out as possible for mobile platforms because of limited rendering capabilities; but with PC, what's really the deciding factor on polygon budgeting? I've been reading that most performance issues nowadays are related to hitting a memory ceiling/draw calls rather than a polygon cap. If occlusion culling and all that cuts out a lot of rendering (especially in enclosed areas like caves/tunnels), when does it start to become a problem?
I think it came out noisier than expected. It kind of reminds me of sun-dried mud. The larger rocks probably won't tile so well across large expanses either so I'll probably re-bake without them. Any other crits are welcome, as usual.
polygon budgeting?- Your going to have limits on budgets depending on what your working on that's a given. The thing to keep in mind is no matter what your making, try to be as efficient as possible. Silhouette, player view, size, use, impact. Make it look kickass, but do it in a smart way!
I definitely understand the importance of keeping stuff as low as possible, however my decisions aren't haphazard. I am adding a more than usual amount of topology in my low poly models for a few reasons. This project is another piece for my portfolio and, although the brazier is a pretty unimportant asset, its place is in the main hall of a room which will be the primary focal point of the project. If I decide to take a beauty shot, lighting will play off those indents far better than if they were cut down as flat as possible, relying only on normal maps (I would add more extravagant features, however I'm attempting to stay true to the original idea). I'm also pretty confident these extra verts (across all my assets) will hardly impact performance, if at all; but if it comes to that, it's much more convenient for me to cut polygons out of my LP than it would be for me to reopen my HP source file to rebuild and readjust UVs to add in more detail.
My shots of the brazier probably don't do me justice. I did say in the first one that I had an excessive poly count that I needed to clean up. I hope I don't sound like I'm discounting your advice and knowledge. On the contrary, I very much appreciate and look forward to your responses.
So here's what I've been putting together today. A few modular platforms and a pillar-type of object. Absolutely not final versions. I know they are pretty bland right now, especially with that unfinished ground texture. I like the way the one with the brick flooring is coming along. I still need to fix up the textures as well as add in foliage such as plants and hanging moss and such to add some life. There will also be some alchemical reagents added here and there (gratuitous glowing mushrooms), but since those assets can't really be changed without confusing the player, I won't be modeling them.
Some seamless tiles and assets made from them
Here are most of the major platform and barrier terrain modules (sorry the shot is so far out, I'm lazy)
And here's a shot of those modules in action.
As before, textures are still WIP; I'm trying to get all the models out of the way before going back to tweak textures. I also still need to model small rubble and other debris assets to add more variation to the terrain, as well as vegetation such as hanging roots, vines, moss, ferns, and perhaps some plantlife native to the Shivering Isles.
All lighting, fog & other effects, as well as alchemy nodes, will be added in the Creation Kit.
These will all be baked onto a texture sheet for use on planes
Now getting to those rocks.
So how does look like are rock to you? Textures are ranging from purple, sick yellow, green... it looks like Stallone beat up a two pieces of living dough stuck together.
I just put in Skyrim, and I am now running around in a dungeon. Looks like they have square sharp rocks in there to me, and the flat round ones are tiled textures with bumped out geo. Look Skyrim is huge, just remember what I said about the budget... last comment I think.
Rounded rocks are what you find in river beds, here are some cave rocks... from caves.
then you have stuff like this
Now on to foliage.
I don't mind the veins,(you should put them in) in the leafs, but taking this into zbrush is a waste of time. It makes your mesh really heavy and it's hard to arrange things for the bake, (if your doing a large shrub or something similar). Once again, focus is on minor stuff. Most of that detail can be done with overlays. The cracks in the leaves, the blocky bark. Also you have to ask... what are you spending your time on? Yeah the way that branch gets really thin in the middle... ummm
Make a leaf, make a stem, group, make a bigger stem, use the smaller stem leaf groups to make the bigger one. Make alpha cards, add geo, bend, pull, push. Make a few variants, etc.
example, (this is for a tree but still) its light, easy, flexible, and does everything you need it too. and if you want to break up the silhouette, delete some groups.
I've got the whole level already blocked in. Every asset is instanced so tweaking textures, UVs or topology doesn't cost me any time in regards to replacing modules. I can get around to that at any time, and frankly I'm leaving much of that to the polish stage. All the terrain elements use seamless textures so there is no high poly to rework. That being said, I've been taking notes on what you've been saying and I will be posting more relevant updates as I execute the changes.
Yeah, I know my stuff is kind of simplistic right now. Once I get my texture scaling down pat I will be going back through during the polish phase and adding more geometry and pulling out certain areas to match the textures. This will enhance the craggy feel. (Again, all my assets are instanced so I don't lose any time doing this.)
Yep, this is exactly what I'll be doing. The stems I posted before were WIP as usual (I fixed the thin part of the branch), but I'll be baking them down to a texture holding all my transparent assets. Nice example for the tree branches, I had planned to do something similar with the branches I have, but for small shrubs and such. What I posted before was one leaf an one stem for each of the two different leafy clusters.
At the end of the day I am a single person on a very strict and limited time frame. I don't have company-level resources or the convenience of getting handed assets to work with, or handing off my assets for them to be blocked out by someone else. Skyrim had entire teams dedicated to various portions of this pipeline. I'm making 100% of the HPs, LPs, textures, UV mapping, level design/blockout, and polish. Granted, Skyrim is much bigger than my single dungeon but I hope you see my point. My primary goal is to get all the assets developed and in place before I worry about perfect textures or perfect topology. I don't mean to come off as defensive, I'm just laying out the facts so you know where I'm coming from
Today I'm blocking in all the stray vines/exposed roots for the level.
what are you modeling in now? maya and zbrush right? and your using Unity for your engine correct?
are you using maya straight out of the box or are you using scripts and plug-in? are you still modeling your hard surface stuff in Z?
Do you know how to hi-poly model in Maya or Max?
Other than that, here's what you would want to do. Look at skyrim or whatever and make a list of assets for the area you want to make. then GRAY BOX BLOCK IT OUT and PUT IT IN AN ENGINE. Run around in it, and make sure you have your basic lighting set up and your focal point, composition, (fist pass). Once your happy with the way it looks, start picking out assets, break them down on a sheet of paper and plan out how your going to make them. What is a unique assets, whats tiled, etc. (this takes two seconds) its portfolio stuff so it can be over budget: geo, textures, make it look awesome, just do it in a smart way, and do go stupid nuts. Think about whats important, and this also translates directly to TIME, (which your are running out of). Are you going to spend a few hours screwing around with miniscule leaf details that no one will ever see in Z, or do it in PS in two sec. If you have a kickass statue that you want to spend the time really making it sweet for a show piece, spend your hours on that, I can see that happening.
Do you know how to overlay normal maps in PS? Crazy bump? Knald?
Start making your assets, you already have "hand painted" stuff on your port, so I would stick with the style your trying to get a job in. This time go borderlands or something, or not, as long as its in the same hand painted stylized vein. Also post some reference and say, "THIS IS THE STYLE I AM GOING FOR!" it helps people get exsited about what your doing because they can visulize it in their heads. Also gives you a clear focus of what your doing. Like this.
I am making this! this style, but in a Skyrim dungeon, (sounds badass to me).
REPLACE GRAY BOX WITH PROPS. tweak textures, lighting pass, (second pass) do this as necessary throughout process.
Do you know UDK?
Do you know how to make materials in UDK or Unity?
(UDK is awesome because there is a ton of information out there already. Pretty handy. I would recommend it for straight up learning)
Finnish building scene, final textures, finial lighting, (finial pass). FX pass, (not a must but nice to do).
Scene DONE.
Beauty shots. Showcase/Hero prop shots. Scene break down.
Your on a good path, keep going! Your a hard worker, your putting in the time. Its not an easy journey, or a short one; but you can make it, I wouldn't be helping out other wise.
If you have questions on how to go about making certain things... like that rock. Make a new thread, or post it on that rocks thread. Ask for help. Don't wait for help to come to you, go get it! Also, you would be surprised at who's keeping an eyes on this thread but don't comment...
I have buckets and buckets of scripts to supplement my modeling/unwrapping speed.
No, I'm not still modeling hard surface stuff in ZBrush
I can high poly model in Maya, yes. I'll be posting a bunch of those Weekly Hard Surface Challenge starting next week.
I know how to overlay detail normal maps on top of other normal maps.
My knowledge of node-based shaders is limited since I never used UDK. However, I did pick up Shader Forge at the beginning of my last project but I haven't had the opportunity to dive in just yet. One of the projects on my list is a Borderlands style, so that will give me the opportunity to mess around with Shader Forge.
Ha! I guess my delivery has been piss-poor huh? I originally started this thread to post daily updates regardless of how complete they were, sort of like a 3D sketchbook. I think I might start limiting my posts to more complete assets. For the next project I will try to make it a bit more organized as far as presentation goes.
Believe me, I am always conscious of the time that has elapsed. I usually take a day every so often (usually after major milestones of each project) and reflect on some stuff. Here are some things I have concluded:
However, the things I learned taking that extra time is certainly not a waste. I've picked up new skills and techniques, learned how to do things... And how not to do things.
- I'm actually not incredibly behind schedule for my sabbatical. Maybe two or three weeks. The biggest cause for this was a couple months back I made the mistake of jumping into one of the monthly environment challenges without thinking and, after a couple weeks, decided I really didn't want to do the project in the first place so I scrapped it.
Ah, thanks man. Again, your support is much appreciated. It's definitely been slow-rolling so far but I see things speeding up soon.Cheers
After a week of hard surface challenges, I think I'm primed for my next project!
Coincidentally, my next project is in fact Borderlands themed! I will be making a small diorama containing some supporting elements to help frame a centerpiece. The centerpiece is a "weapon crate" similar to all the varying kinds that present sweet loot when you open them, like this:
I noticed only Hyperion and DAHL (and Bandits I suppose) are really the only in-game companies which manufacture loot crates. I always wondered what the other companies' crates would look like, such as Vladof, Tediore, Maliwan and Jakobs.
My play style in Borderlands tends to gravitate toward the heavy-hitting weapons that Jakobs is mostly known for. As such, Jakobs is going to be my inspiration behind the weapons container I'll be making.
There isn't a whole lot of content to go off, mostly just the feel of the brand logo and the special skin that Claptrap has. However, with those two things I think it will be fun to blend an old West/Texan feel to what is foundationally a sci-fi style. I foresee lots of wooden panels and swirly leafy designs.
Here's a number of reference pictures I grabbed, mostly for mechanical functionality:
I've gotten started with a blockout for all the elements of the container. I don't want to jump right into making the high poly until I'm satisfied that all the components read well and appear functional/real.
The red color indicates compartment doors.
The green color indicates locking mechanisms.
I toyed around with the idea of having one of those big spinning handles to unlock all the compartments, but I think that is a bit too far from the general style and probably wouldn't lend well to a timely animation.
Bartalon, seeing your process and what worked/didn't work over the past months really helps me plan out my next 6 months or so. Going through this thread shows your improvement, keep up the hard work and dedication.
ZombieWells, thanks for all your feedback. I read most of your longer post.. esp your 5/25 post about start to finish macro workflow.
Not trying to hijack at all, just encouraging you both.
ZombieWells Kind of a late response but I took your advice and made a dedicated thread for just the Borderlands project. I think from now on I'm going to post less frequently on here, but with more relevant and complete content; and post my WIP stuff in specific threads.
oxygencube Thank you! I'll keep an eye out for your stuff popping up in the forums as well
My Borderlands project is finally wrapped up, here are the shots for that:
Opening animation:
Interactive turnaround:
[sketchfab]d90a1b9a9f194d0f87c52f49f750e0ab[/sketchfab]
My polycount thread for the whole project can be found here.
A more detailed process can be found here.
Hey, thanks! Sorry for the super late response haha. I've been doing stuff and things.
Speaking of which, I'm still active! I just haven't been posting much on my last project, which is done. It's a Unity Asset Store pack!
More shots can be found on the Asset Store Page. Not trying to solicit my content, it's just more convenient to look at the images from there.
---
Six weeks later, I post the finished version of my latest project. You can see the development of this project in the original project thread here:
[PBR] Environment Challenge
Here are some final renders of my scene, rendered in Unreal Editor 4, textured entirely with Substance Painter. You can also check out the full detailed breakdown on my portfolio site, here.
Objects' normals and textures are very good. Variance of materials on normal maps is well readable. I'm not sure about final materials though, because they look better in the portfolio IMO. Why are the surfaces so diffuse compared to portfolio breakdowns? Do you have enough reflection capture spheres?
Out of curiosity, because I don't know that: What is the purpose of the Albedo Color Overlay in the master material?
I believe the difference here is due to the Substance Painter shots having full cube map illumination (to show the details), while in the actual scene the ambient cube map settings in the scene are very low to accommodate the dark, underground location. Although it could very well simply be my inexperience with UE4
The Albedo Color Overlay (and also the Desaturation) shifts the albedo color of some assets. In relation to my scene, I could have changed the saturation or made slight hue shifts to the bins, crates, concrete slabs, plaster walls, etc.