Added spec and normal maps. My MAIN focus was the wood which I think came out nicely for my first (ever) spec map. The normals were tweaked from CrazyBump and Photoshop. The stone shouldn't be looked at as I will probably go the high res route and bake in those normals.
There is a noticeable seam on the horizontal wood pieces. Mostly because you have it getting really dirty/dark at one end of the wood piece, but not the other. However, I would advise against doing that all together, because even if you fix the seam itself by doing it on the other side too, it will be quite repetitive.
Yeah, I'd definitely go with a high res model for the stone or something. They are just not popping out well IMO.
There is a weird vertical line on the rocks just to the left of each wooden pillar. What's that about? Did you bake the AO then move the pillars slightly or something?
Tweaked just about everything. Alot of the issues before are fixed, and I UVed some other parts of the puzzle. About to call the "texture set" done. Normal map needs tweaking.
The first thing that stands out to me is how blobular the wood appears to be. It has no hard edges. Even if you were intending to have the wood be rounded off, it doesn't feel so much like it's rounded off, as it's just the way the smoothing groups ended up. Did you do a high poly piece for this? Even if you do just a very basic high poly piece to just get those edges, it could make a huge difference.
The wood texture feels odd to me. It almost feel as if the wood was stained or treated some way. I'm not sure what the theme for the scene is, but if it's supposed to be recreating a fort/castle wall back in the old days, you wouldn't likely see any stained/finished wood on the fort walls. As you see in Chris' paintover, the wood is very dull (not much spec to speak of), a bit rugged, etc. I don't know, just feels to modern.
The rock texture is pretty "smooth" looking. The normal map might be a bit strong for the bumps on the rock, and not very noisy for the little details (that's why I say it looks smooth). I'd say just take the diffuse, and run just the old photoshop normal map filter on it, then overlay it ontop of normal map you have now, and see how that looks. It looks like you may have done that, but I can't see it really. The spec on it seems a bit too high. Mostly the gloss/power is a bit too tight. Maybe tone that down a bit.
Also might want to consider adding a bit more dirt/grime kick up at the bottom. You have it being a bit darker towards the bottom, which is cool, but dirt and grime doesn't go on quite so evenly in reality. I know it can be difficult to pull off when you are tiling so often, just play with it and constantly look at it on the model to see if it gets repetitive or not. As long as your eye doesn't get drawn to it like a pimple on the first date (sorry for the stupid analogy), you should be okay.
Speaking of that, the top wood spade pieces all use the same UVs. So you want to make sure that it's as bland as possible while maintaining the integrity of the material. You have some dark patterns about 3/4 of the way up that makes it extremely noticeable when you have 5 or 6 of them lined up next to eachother. I would try to minimize how much of that is in there, because it will be very obvious that it's tiling.
You could also try adding some decals or something it appear to not tile. Some chipped rocks, rotting wood areas, etc. A little trick I tend to use as well as it to make the diffuse use a composite material. Put the diffuse map into both of the slots, but go into the Output>RGB Levels and lower the value (to your liking) on ONE of the slots. Then for the mask, you could make one custom, or just use a noise material. This will blend the two maps together (in this case it's using the same map, so it's not adding memory) and you should get areas that are darker and areas that are lighter. This trick is subtle, but can really detract from the tiling. And if you are okay with using a 2nd diffuse map, you could could really take it much further.
The stone fits much better I think. I bought an Intous3 9x12 and it gets here tomorrow so I will do a high res stone tomorrow! (Can't wait) and I really think that will make it look better.
I'm very thankful for you guys and the help your giving me! Thanks!
In the mean time, I switched to Maya, and I LOVE it! I also went back and modeled high res stones / wood to give more umph to the normal map, which turned out nicely!
A quick note though, currently each piece has it's own unique texture, which at this point, is very hard to keep everything consistent (I know the spec is too much on the rocks and jazz)and to modify all those textures is too much work so I'm currently going back and just making a tile able stone and wood texture and having all the pieces share from that. It should conserve on VRAM too.
Could you post a copy of the rock you are using for reference, and an example of what you are trying to achieve?
2nd - what are you assuming the grout between the rocks made of, other than "black"? Varying the contrast of colors between the grout and the rock would be a good idea.
3rd - The "spikes" at the top of the wall pieces, how is this attached to the wall? You need to keep construction in mind before you just embed geometry for no obvious reason.
4th - Vary the sizes of the rocks up more. Lots of small ones in between big ones. Right now it looks like they are all within about 75% - 125% the same size of each other.
Another thing with AO maps - If you are rendering them out and using anything but a White object with white light and black shadows, you will be multiplying or overlaying the gray onto your texture and it will desaturate it more than intended. Also, don't forget to overlay your AO onto your spec as well.
[ QUOTE ]
(I know the spec is too much on the rocks and jazz)and to modify all those textures is too much work
[/ QUOTE ]
I'm sure the guys who made E.T. for the Atari probably said that making a good game would have been too much work, but they still did it in less time than it's taken you to do a wall.
Thanks for the crits! I'll post my references sometime here soon, hitting the sack - responding before I go to bed.
The "spike's" true shape would be more of a <font class="small">Code:</font><hr /><pre>/\
||</pre><hr />, meaning, the wood goes into the grout.
I'm currently (almost done) redoing the normal map and I'm fixing that.
The AO map I'm using is from Xnormal, so it came out white, but I used the normal for a cavity map to enhance the diffuse. The colors I am currently working on, as I did notice that issue (not as apparent in the diffuse, but onces in Maya's HQ render, it gets blacker).
And you ending comment, I see no point as to why you had to say that. This last post is a revisit from a few month old project that I slacked on (and was my first project of this size). If you ment to bring me down as an artist, and want me to give up, then pat yourself on the back, great job.
First off, anyone who has been on this forum for more than 20 seconds know that if you are looking for candies and cream and sugar on top of your crits, head on over to 3dbuzz.com, because polycount aint it. Do you think your boss at a game studio is going to sit you in a corner office and ask you what you would like to model and how long you would like to work on it?
Secondly, my statement wasn't to put you down, but to point out that saying that something "would take too much time" - that is as important as your color / spec maps - isn't really an excuse.
From what I gather, this project was for you to learn, and to do something RIGHT, so to state that something would "take too long" when you know it is wrong isn't a work mentality that will get you far.
If you can handle someone telling you your work is complete ass, you have failed in life, your mother is a hooker, and you are staying late tonight until you get it right - you are halfway to being a game artist.
You have to look at your stuff VERY critically and push yourself until it IS right.
As far as creating walls like this, it is pretty close to how this sorta stuff would be done in a current-gen title with the modularity, but I would suggest varying it up with more than just one. Perhaps two or three variations, and just randomize them up as you snap them together. You really start to notice the repetitiveness on the surfaces with the same support design you have currently. You could vary the direction, size, and damage on the support struts and then use those to hide your seams, and just Re-UV along where those are.
I agree with the Ott on this one TSM. Ott has given you some great advice here, and this is not the forum that shoves sunshine up peoples ass's, he was straight and to the point. So try not to get to upset, he has valid comments. Good luck with your models.
Yeah listen to the two ass clowns that posted before me. Ott has always been Greys bear type lover and plus he is the top.
All joking aside yeah man your stuff is starting to come together and by adding small variances like suggested it really helps push it even more.
Polycount is by far one of the open and honest forums out there. If you shit in a bag and toss it up here we will call you on it. We will also give you good feedback and help you become better at what you do. Trust me I have had my ass torn into a million pieces at times on some of my art but that is what pushes us to become better.
Just keep plugging away good sir and post up some work constantly and you will see how you get better and better each time.
Some more stuff, hope I improved some! 1024x1024^3 here. Will post maps if requested. Also, there are black lines on the sword.. I think it'd due to Xnormal's filtering... can anyone confirm?
Metal on the axes looks like granite instead of steel or iron. Normal maps on the handles look off to me too, more often than not the handles on these things were polished smooth enough so they shined. A crazybump filter on the wood texture just doesn't cut it for this aplication, you would be better off without the normal map and using a good spec map to make the grain of the wood really shine through.
As for the line on the sword it looks like you have a seam there maybe try unwrap the blade in 2 pieces (back/front) instead of splitting it in half.
Thanks for the response! The material is a stone for the axe heads. I don't see kingly knights using these weapons (except the sword), more Goth's or Franks. There is a ranking system in this... each one is a bit "high quality" then the next. I figured a barbarian wouldn't want to sand down the wood at all.. would take too much time on his part.
I did a basic high res geo for each piece. About 30-50k tris on average I'd say(For the big details, I let crazybump handle the small ones). It's hard to see the details from the small pictures. But I'll definitely work on it more and post some updates.
If they are meant to be stone then the head shapes are all wrong It would take a sculpter a lot of work to make the axe heads shaped that well and the weapon would be all but useless in a fight. Take a look at some stone age weapons on the web and get some ideas there. Otherwise I would turn these into iron axe heads and throw on some rust etc to make them look "lower quality". Making stone weapons look authentic may be above your skill level at the moment so I would go for Iron as the material. The handles were always smoothed on most weapons, the reason being that you do not want to get splinters while using it. In some cases the handles were left rough and wrapped in leather cord to provide grip. Research is your friend.
Replies
http://img223.imageshack.us/my.php?image=texturevx1.jpg
http://img89.imageshack.us/my.php?image=singletestgv1.jpg
Yeah, I'd definitely go with a high res model for the stone or something. They are just not popping out well IMO.
There is a weird vertical line on the rocks just to the left of each wooden pillar. What's that about? Did you bake the AO then move the pillars slightly or something?
Looking good though, keep it up.
It is rather repetitive, though. Try flipping the UVs horizontally or vertically on a few.
http://img67.imageshack.us/my.php?image=renderwj2.jpg
The wood texture feels odd to me. It almost feel as if the wood was stained or treated some way. I'm not sure what the theme for the scene is, but if it's supposed to be recreating a fort/castle wall back in the old days, you wouldn't likely see any stained/finished wood on the fort walls. As you see in Chris' paintover, the wood is very dull (not much spec to speak of), a bit rugged, etc. I don't know, just feels to modern.
The rock texture is pretty "smooth" looking. The normal map might be a bit strong for the bumps on the rock, and not very noisy for the little details (that's why I say it looks smooth). I'd say just take the diffuse, and run just the old photoshop normal map filter on it, then overlay it ontop of normal map you have now, and see how that looks. It looks like you may have done that, but I can't see it really. The spec on it seems a bit too high. Mostly the gloss/power is a bit too tight. Maybe tone that down a bit.
Also might want to consider adding a bit more dirt/grime kick up at the bottom. You have it being a bit darker towards the bottom, which is cool, but dirt and grime doesn't go on quite so evenly in reality. I know it can be difficult to pull off when you are tiling so often, just play with it and constantly look at it on the model to see if it gets repetitive or not. As long as your eye doesn't get drawn to it like a pimple on the first date (sorry for the stupid analogy), you should be okay.
Speaking of that, the top wood spade pieces all use the same UVs. So you want to make sure that it's as bland as possible while maintaining the integrity of the material. You have some dark patterns about 3/4 of the way up that makes it extremely noticeable when you have 5 or 6 of them lined up next to eachother. I would try to minimize how much of that is in there, because it will be very obvious that it's tiling.
You could also try adding some decals or something it appear to not tile. Some chipped rocks, rotting wood areas, etc. A little trick I tend to use as well as it to make the diffuse use a composite material. Put the diffuse map into both of the slots, but go into the Output>RGB Levels and lower the value (to your liking) on ONE of the slots. Then for the mask, you could make one custom, or just use a noise material. This will blend the two maps together (in this case it's using the same map, so it's not adding memory) and you should get areas that are darker and areas that are lighter. This trick is subtle, but can really detract from the tiling. And if you are okay with using a 2nd diffuse map, you could could really take it much further.
Just some suggestions.
Keep it up
J
I changed the diffuse and the spec with this update, I dont know if I like it more though, it just seems.. "different".
Was it a step in the right direction?
http://img137.imageshack.us/img137/3191/renderjr5.jpg
The biggest change was in my spec map.
I reworked the stone and I like it alot more than my previous update.
http://img89.imageshack.us/img89/1514/rendergroupmm3.jpg
The stone fits much better I think. I bought an Intous3 9x12 and it gets here tomorrow so I will do a high res stone tomorrow! (Can't wait) and I really think that will make it look better.
I'm very thankful for you guys and the help your giving me! Thanks!
In the mean time, I switched to Maya, and I LOVE it! I also went back and modeled high res stones / wood to give more umph to the normal map, which turned out nicely!
http://img511.imageshack.us/img511/190/fortba9.jpg
A quick note though, currently each piece has it's own unique texture, which at this point, is very hard to keep everything consistent (I know the spec is too much on the rocks and jazz)and to modify all those textures is too much work so I'm currently going back and just making a tile able stone and wood texture and having all the pieces share from that. It should conserve on VRAM too.
2nd - what are you assuming the grout between the rocks made of, other than "black"? Varying the contrast of colors between the grout and the rock would be a good idea.
3rd - The "spikes" at the top of the wall pieces, how is this attached to the wall? You need to keep construction in mind before you just embed geometry for no obvious reason.
4th - Vary the sizes of the rocks up more. Lots of small ones in between big ones. Right now it looks like they are all within about 75% - 125% the same size of each other.
Another thing with AO maps - If you are rendering them out and using anything but a White object with white light and black shadows, you will be multiplying or overlaying the gray onto your texture and it will desaturate it more than intended. Also, don't forget to overlay your AO onto your spec as well.
[ QUOTE ]
(I know the spec is too much on the rocks and jazz)and to modify all those textures is too much work
[/ QUOTE ]
I'm sure the guys who made E.T. for the Atari probably said that making a good game would have been too much work, but they still did it in less time than it's taken you to do a wall.
Thanks for the crits! I'll post my references sometime here soon, hitting the sack - responding before I go to bed.
The "spike's" true shape would be more of a <font class="small">Code:</font><hr /><pre>/\
||</pre><hr />, meaning, the wood goes into the grout.
I'm currently (almost done) redoing the normal map and I'm fixing that.
The AO map I'm using is from Xnormal, so it came out white, but I used the normal for a cavity map to enhance the diffuse. The colors I am currently working on, as I did notice that issue (not as apparent in the diffuse, but onces in Maya's HQ render, it gets blacker).
And you ending comment, I see no point as to why you had to say that. This last post is a revisit from a few month old project that I slacked on (and was my first project of this size). If you ment to bring me down as an artist, and want me to give up, then pat yourself on the back, great job.
Secondly, my statement wasn't to put you down, but to point out that saying that something "would take too much time" - that is as important as your color / spec maps - isn't really an excuse.
From what I gather, this project was for you to learn, and to do something RIGHT, so to state that something would "take too long" when you know it is wrong isn't a work mentality that will get you far.
If you can handle someone telling you your work is complete ass, you have failed in life, your mother is a hooker, and you are staying late tonight until you get it right - you are halfway to being a game artist.
You have to look at your stuff VERY critically and push yourself until it IS right.
As far as creating walls like this, it is pretty close to how this sorta stuff would be done in a current-gen title with the modularity, but I would suggest varying it up with more than just one. Perhaps two or three variations, and just randomize them up as you snap them together. You really start to notice the repetitiveness on the surfaces with the same support design you have currently. You could vary the direction, size, and damage on the support struts and then use those to hide your seams, and just Re-UV along where those are.
All joking aside yeah man your stuff is starting to come together and by adding small variances like suggested it really helps push it even more.
Polycount is by far one of the open and honest forums out there. If you shit in a bag and toss it up here we will call you on it. We will also give you good feedback and help you become better at what you do. Trust me I have had my ass torn into a million pieces at times on some of my art but that is what pushes us to become better.
Just keep plugging away good sir and post up some work constantly and you will see how you get better and better each time.
468 tris
227 tris
http://img148.imageshack.us/my.php?image=swordlevel2ti7.jpg
484 tris
As for the line on the sword it looks like you have a seam there maybe try unwrap the blade in 2 pieces (back/front) instead of splitting it in half.
Thanks for the response! The material is a stone for the axe heads. I don't see kingly knights using these weapons (except the sword), more Goth's or Franks. There is a ranking system in this... each one is a bit "high quality" then the next. I figured a barbarian wouldn't want to sand down the wood at all.. would take too much time on his part.
I did a basic high res geo for each piece. About 30-50k tris on average I'd say(For the big details, I let crazybump handle the small ones). It's hard to see the details from the small pictures. But I'll definitely work on it more and post some updates.
Thanks!
Good luck.