If the masks aren't having an effect, and you know you are saving/exporting properly, then try increase your in game graphics options. That might be what is holding you textures back.
This was the setting I used and I was saving with different names every time i imported it into the client. Does anyone have trouble using dota 2 test? The workshop interface is broken in mine.
It's not an issue with the Dota 2 settings. Post your masks so we can have a look and see what the issue is. Keep in mind that if your alpha channel of either mask is 100% black or 100% white the masks will not work.
@bronto
I'm fairly certain that the masks are correct.
Here are my masks coupled with the original mask for easy comparison. I resized everything to 256 so that is what the engine reads. I also fooled around with the mask 1 R channel because glowing stuff .
The main difference that I see was the rim light (mask 2 channel g), but I only did that because it looked like the rim light was too strong at my end.
You know how valve created a button to toggle from lod0 to lod1? I think an improvement to that is for people to also toggle between original mesh to new mesh. Switch valves content to new content, therefore its easier to compare.
Did you bake those normals in handplane? Your normals are all borked and look less a tangent space normal map, more an object space normal map.
If you did bake them in handplane, make sure the handplane model has its mirrored UV islands shifted by 1UV space left/right. Also, make sure that that model you're using handplane on is the same or at least has the same 0-1 UV value islands as the one you baked the object space map from.
@snowstorm
I see where I made the mistake. I converted a tangent space map rather than an object space map.
Well I rebaked in xNormal and I still get the problem.
Can someone explain moving the mirrored UV coordiantes? I never read anything on that. The term would be enough for me to do some research on it.
If you look at this image, the high poly bits I want to bake are in black, I've selected the polys I want to bake them onto with orange, and any overlapping UVs are highlighted red. All the non-selected polys aren't highlighted orange. If you look at the UV window the overlapping UVs are all shifted to the right by exactly 1 UV coordinate, meaning that the game still interprets them as being in the same position as those non-overlapping bits. Notice that from UV values 0-1 (inside the main square with highlighted polies) none of the UVs have overlaps.
Generally having overlapping UVs can cause all sorts of baking errors, or in handplane's case shading errors. If I had to guess it's because having overlapping UVs tends to confuse renderers, though there's probably a more accurate technical way of putting it.
If you want to do further reading, this is a good starting point (though it doesn't illustrate the steps specifically):
No they're different. If you're using bent normals you can get object space maps from xNormal by going into the "..." for Normal Maps in your Baking Options and unchecking the check box beside Tangent Space. Then just bake out the Normal Map instead of the Bent Normal.
I don't know if this will help but these are my normal maps that I baked from xNormal. It looks a little weird to me.
Left is tangent and right is Object.
@Teriyaki
Your maps look a little strange, but maybe thats because its only 128x256. You should fill the UV space as much as you can.
My tangent space nomal map. Its baked in xNormal because the conversion from oject to tangent didnt work for me.
Edit: Fixed my problem and figured out why handplane gave em weird results. It gave me weird esults because when I rigged it and gave it an offset, handpplane read the offset. So I just used the model I had that was still at 0,0,0 then rigged it and used that for the handplane inputs.
Hi, nice work there, Just started mine earlier today, will try to break my fingers on Zbrush, never really been using it before, so far i have High poly done in Max
Well, the month is ending soon, and I figured that we should talk a bit about the 'Noob Challenge'. At the start a lot of people showed interest and I thought that we would have something pretty cool, but I think that only a handful of people really tried to do something, and a lot of people either just stopped at the highpoly or lowpoly, with very few getting textures and the skinning done. So, I gotta ask, is there something wrong with the thread model? Or do you guys think that it's fine the way it is? I don't think that the lack of interest was due to the thread not being "marketed" enough, because it was posted on reddit and on some CG groups on other places, and there was definitely some interest there, but I don't know, I would love to see more people joining in and going all the way through, from start to finish. I personally couldn't do it because I'm pretty busy with some other stuff, but I see a lot of people trying to do Dota stuff which should give the noob challenge a try.
Anyway, what about next month? Do you guys still think that this format can keep up? Any suggestions? Something you'd like to be changed? I'd really love to read what you guys have in mind.
In my case it was really even a challenge to start to model something. With that in mind, be sure this thread did at least help 1 person to get into that stuff.. me
Also Im really encouraged by guys like you giving back something to the community and don't hesitate to critique and give feedback to new guys like I am myself. There might be a few reasons why people stopped in between, but at least it was a pleasure to see all the wires and how people approached some of the stuff.. So even when the will to participate sunk at a certain point, Im sure people followed and learned something from the things you guys posted and explained.
So, that being said, Im really looking forward to the next month and hope for even more kind of criticism and tips from established artists like you. And last but not least, thanks everyone for posting in here and especially to those who wrote just a few kind words which can really make the difference.
@belkun - Not many of people actually modeled something to the exact detail of the asset (which I think was the purpose, right?) people seemed more interested in adding their own flavors which makes it difficult to critique their work, as there is obviously no benchmark to hit.
We also need to keep in mind that most people who frequent this forum are competent (or believe they are competent despite being behind) & thus feel the topic is beneath them.
Past those points, I'm not sure if I can put in any valuable suggestions.
Personally I enjoyed this challenge - it helped me solidify my workflow for creating items. Here's my final product. Reverse shot is to show off the rim lighting. My handle cloth also needs to be darker on the diffuse.
Thanks again guys, this was fun! I'll participate in the next one for sure, regardless of any changes made.
I think we should still continue the challenges because I learned a lot from tthis. However when I reached a wall (masks) I started to procrastinate thinking that the deadline is too far to worry about, so I think 1 month is just too long to make 1 item. Perhaps make this bi-monthly ending at the 15th of the month?
This competition did teach me a lot but there just wasn't enough to go on with a plain sven sword. So what about the next competition, has more choices such as
-Weapon basic
-Armor basic
-Weapon w/ advance rigging (ex: Dazzle or WD staff)
-Armor w/ advance rigging (ex: huskar helm)
-Maybe ward that goes with DEFAULT animations?
(The ward thing might go a little overboard though)
Here are more progress things. I've been getting a jitter on the blade and was wondering what is causing this. Some help would be greatly appreciated.
Well, the month is ending soon, and I figured that we should talk a bit about the 'Noob Challenge'. At the start a lot of people showed interest and I thought that we would have something pretty cool, but I think that only a handful of people really tried to do something, and a lot of people either just stopped at the highpoly or lowpoly, with very few getting textures and the skinning done. So, I gotta ask, is there something wrong with the thread model? Or do you guys think that it's fine the way it is? I don't think that the lack of interest was due to the thread not being "marketed" enough, because it was posted on reddit and on some CG groups on other places, and there was definitely some interest there, but I don't know, I would love to see more people joining in and going all the way through, from start to finish. I personally couldn't do it because I'm pretty busy with some other stuff, but I see a lot of people trying to do Dota stuff which should give the noob challenge a try.
Anyway, what about next month? Do you guys still think that this format can keep up? Any suggestions? Something you'd like to be changed? I'd really love to read what you guys have in mind.
My reason for not continuing it was that I worked until I learned what I wanted to learn (controlling hard surfaces in Zbrush) and the rest I already knew how to do so I just didn't see the need to continue it.(Of course I got busy too)
I think if you continue to make the challenge over something that people have problem with, people will be more participative.
My suggestions for next challenge:
1. Hard surface sculpting (learn how to make stuff in Zbrush that don't look like clay) - maybe harder than this month's
2. retopo + cage + Bake: Let's pick something that is hard to retopo it. The challenge will be to manage to get retopo a highpoly object and define a cage and get straight clean bakes without dirty tricks (such as cleaning the bakes in Photoshop). I think if one of the seniors that have items in game is keen enough to share one of their sculpts with us, it will be really nice. I personally would go all the way to the end.
3. a lot of people(including myself) wanna learn animating as well. But surprisingly we are all looking for tutorials not because they are not animating tutorials out there but because we don't know the workflow of dota2 animating. let's pick up a ward or a summon (or maybe a courier) and learn the animating workflow in Dota 2.
I stoped as well because my holiday was over and because of that I am accesing to computer 3 times a week or maybe non at all. And thw montly contest and my lion head took more time than they should. And another reason was like Reza said, I was trying to improve myself at making lod1 and lod0 so I stopped after I made them.
I think that I can see now some of the main problems with the challenge, the first being that the month may have been a bit too busy for lots of people, and the second that maybe completely finishing the challenge isn't a goal for many, even though I think that it should be.
So, do you guys think that we should have something a bit more free form? Like having different concepts, or having an artist from the community "donate" a concept for us? Maybe just "enforce" people to complete at least one "stage" of the item creation and everything else being an extra? Maybe something more complex? I tried going for a simple sword just to start things up, but something a bit harder would definitely be better and more challenging. Perhaps instead of doing a 'challenge' we could do something different, like a 'sketch' thread where people try to create items (either just a concept, a model, textures or all of them) just for practice? Just having a brainstorm and throwing some ideas, would love if you guys gave some thoughts.
I'm really really happy that this thread could help some of you guys, and if possible I'd love to keep this up for the next month and the month after that, until we have no more noobs here, haha! However, I need you guys to help me out with that in order to make the next thread even better!
I have some ideas:
1. We can do it as parts, I mean If one of the pros post their work files part by part (lod0, lod1, high poly, unwrapping, texturing, rigging, masking) people can comtinue from the part they like, I mean if they dont think they need to.model it but they think that they need to practice in texturing, masking and rigging. With this they can download the unwrapped and baked model to continue. I know it can be hard to do but I will drop this here.
2. There can be different concepts as a diffuculty. Like very easy model like this for starters, a little harder one for better, and I hard one like finishing project. If someone finish the hard finishing project he wouldnt be noob anymore. Maybe??
A model that isn't in game would probably keep more interest. It was cool to see everyone working on the same item but could also be cool to watch several concepts be worked on. Or give a basic silhoutte, then everyone could use that to develop their own style weap.
I think that there's no 'challenge' on doing just parts of stuff. The point is to learn new things, and if you just do one or two parts, there's not a lot of new things to learn. Also, how we would choose the parts? Maybe get the full files from someone who got an item ingame, like basemesh, sculpt, lowpoly and textures? I think that it would be a bit hard to get something like that
I like the silhouette idea! Maybe something like Concept Cookie's Shape Challenge? That would help people make a concept too! I definitely like where this is heading!
One more thing.
One month was too much for this sword. (since it was an easy one too)
At the beginning I joined this somehow cause I wanted to support the idea, but then since I had a lot of time to postpone, I lost interest
The silhouette would be a good idea. So will the next competition be silhouette to item? Should there be requirements like if someone chose to make a WD staff from the silhouette would they be expected to use the jiggle bones or a particle effect on anti-mage?
There should also be a random silhouette for each weapon type or shield, or at least enough of them so people's items don't look so similar (maybe 3?
Wouldn't this change allow people to submit their item to the workshop, since it's not copying something else and its just puerly from the one person themselves?
The silhouette would be a good idea. So will the next competition be silhouette to item? Should there be requirements like if someone chose to make a WD staff from the silhouette would they be expected to use the jiggle bones or a particle effect on anti-mage?
If we go with the silhouette idea, the challenge would be much more open, meaning that everyone would do something different and there would be no restrictions. I don't really get that last question, but still, anyone would be able to do anything for any hero, it would just need to match the provided silhouette.
Wouldn't this change allow people to submit their item to the workshop, since it's not copying something else and its just puerly from the one person themselves?
Yes. Since everyone would need to create an original concept just based on a silhouette, you would be able to submit it to the workshop with no issues.
I loved the silhouette idea, but I think that will turn "noob challenge" into a another version of "montly contest". It will be cool to have a real theme.
This is my first post and my first model. I only just found out about this activity so I unfortunately will not be able to finish in time. However I look forward to the next one! Sorry I do not have time to leave additional comments.
Hey Geos, welcome to the gang! The main thing I can recommend you is to be very careful with those triangles, specially if you're going to subdivide it. But it seems that you got a nice shape!
Anyway, this is the last day on this month, and if possible I'd love to get March's Noob Challenge by monday, so what do you guys think we should do? The silhouette idea, another item which is already in-game or try to get an artist to donate a concept?
@hmShark, ask belkun he's the one in charge of the noob challenges here.
also, how should I do presentation images? not promo art but just to display the screenshots.
If possible, a minimum of two screenshots, one from the loadout view and one from the in-game view. If you want to do a promo, that would be really cool too, but it's not really needed.
This probably isnt the place to talk about such things but, I really wanna do this stuff myself however I get completely lost and confused in it all, but I have nice concepts. If the owner of this here forum is happy to let me I can provide concepts for you, whoever submits it to Valve if it gets approved earns whatever there is to earn. I just want to see my ideas in there XD
Replies
I'm fairly certain that the masks are correct.
Here are my masks coupled with the original mask for easy comparison. I resized everything to 256 so that is what the engine reads. I also fooled around with the mask 1 R channel because glowing stuff
The main difference that I see was the rim light (mask 2 channel g), but I only did that because it looked like the rim light was too strong at my end.
You know how valve created a button to toggle from lod0 to lod1? I think an improvement to that is for people to also toggle between original mesh to new mesh. Switch valves content to new content, therefore its easier to compare.
Thanks for the reply
If you did bake them in handplane, make sure the handplane model has its mirrored UV islands shifted by 1UV space left/right. Also, make sure that that model you're using handplane on is the same or at least has the same 0-1 UV value islands as the one you baked the object space map from.
The normal map should look like this: http://www.hardworkerstudio.com/HWR2013/wp-content/uploads/2012/08/normalmap.jpg
Normal maps generally don't have rich reds or yellows.
I see where I made the mistake. I converted a tangent space map rather than an object space map.
Well I rebaked in xNormal and I still get the problem.
Can someone explain moving the mirrored UV coordiantes? I never read anything on that. The term would be enough for me to do some research on it.
If you look at this image, the high poly bits I want to bake are in black, I've selected the polys I want to bake them onto with orange, and any overlapping UVs are highlighted red. All the non-selected polys aren't highlighted orange. If you look at the UV window the overlapping UVs are all shifted to the right by exactly 1 UV coordinate, meaning that the game still interprets them as being in the same position as those non-overlapping bits. Notice that from UV values 0-1 (inside the main square with highlighted polies) none of the UVs have overlaps.
Generally having overlapping UVs can cause all sorts of baking errors, or in handplane's case shading errors. If I had to guess it's because having overlapping UVs tends to confuse renderers, though there's probably a more accurate technical way of putting it.
If you want to do further reading, this is a good starting point (though it doesn't illustrate the steps specifically):
http://wiki.polycount.com/NormalMap/#Mirroring
It didn't really work because my bakes are still weird (same as above)
Are bent normal maps the same as object space maps?
Left is tangent and right is Object.
Your maps look a little strange, but maybe thats because its only 128x256. You should fill the UV space as much as you can.
My tangent space nomal map. Its baked in xNormal because the conversion from oject to tangent didnt work for me.
Edit: Fixed my problem and figured out why handplane gave em weird results. It gave me weird esults because when I rigged it and gave it an offset, handpplane read the offset. So I just used the model I had that was still at 0,0,0 then rigged it and used that for the handplane inputs.
Anyway, what about next month? Do you guys still think that this format can keep up? Any suggestions? Something you'd like to be changed? I'd really love to read what you guys have in mind.
Also Im really encouraged by guys like you giving back something to the community and don't hesitate to critique and give feedback to new guys like I am myself. There might be a few reasons why people stopped in between, but at least it was a pleasure to see all the wires and how people approached some of the stuff.. So even when the will to participate sunk at a certain point, Im sure people followed and learned something from the things you guys posted and explained.
So, that being said, Im really looking forward to the next month and hope for even more kind of criticism and tips from established artists like you. And last but not least, thanks everyone for posting in here and especially to those who wrote just a few kind words which can really make the difference.
And big thanks to you belkun for taking the time
We also need to keep in mind that most people who frequent this forum are competent (or believe they are competent despite being behind) & thus feel the topic is beneath them.
Past those points, I'm not sure if I can put in any valuable suggestions.
Personally I enjoyed this challenge - it helped me solidify my workflow for creating items. Here's my final product. Reverse shot is to show off the rim lighting. My handle cloth also needs to be darker on the diffuse.
Thanks again guys, this was fun! I'll participate in the next one for sure, regardless of any changes made.
This competition did teach me a lot but there just wasn't enough to go on with a plain sven sword. So what about the next competition, has more choices such as
-Weapon basic
-Armor basic
-Weapon w/ advance rigging (ex: Dazzle or WD staff)
-Armor w/ advance rigging (ex: huskar helm)
-Maybe ward that goes with DEFAULT animations?
(The ward thing might go a little overboard though)
Here are more progress things. I've been getting a jitter on the blade and was wondering what is causing this. Some help would be greatly appreciated.
My reason for not continuing it was that I worked until I learned what I wanted to learn (controlling hard surfaces in Zbrush) and the rest I already knew how to do so I just didn't see the need to continue it.(Of course I got busy too)
I think if you continue to make the challenge over something that people have problem with, people will be more participative.
My suggestions for next challenge:
1. Hard surface sculpting (learn how to make stuff in Zbrush that don't look like clay) - maybe harder than this month's
2. retopo + cage + Bake: Let's pick something that is hard to retopo it. The challenge will be to manage to get retopo a highpoly object and define a cage and get straight clean bakes without dirty tricks (such as cleaning the bakes in Photoshop). I think if one of the seniors that have items in game is keen enough to share one of their sculpts with us, it will be really nice. I personally would go all the way to the end.
3. a lot of people(including myself) wanna learn animating as well. But surprisingly we are all looking for tutorials not because they are not animating tutorials out there but because we don't know the workflow of dota2 animating. let's pick up a ward or a summon (or maybe a courier) and learn the animating workflow in Dota 2.
I think that I can see now some of the main problems with the challenge, the first being that the month may have been a bit too busy for lots of people, and the second that maybe completely finishing the challenge isn't a goal for many, even though I think that it should be.
So, do you guys think that we should have something a bit more free form? Like having different concepts, or having an artist from the community "donate" a concept for us? Maybe just "enforce" people to complete at least one "stage" of the item creation and everything else being an extra? Maybe something more complex? I tried going for a simple sword just to start things up, but something a bit harder would definitely be better and more challenging. Perhaps instead of doing a 'challenge' we could do something different, like a 'sketch' thread where people try to create items (either just a concept, a model, textures or all of them) just for practice? Just having a brainstorm and throwing some ideas, would love if you guys gave some thoughts.
I'm really really happy that this thread could help some of you guys, and if possible I'd love to keep this up for the next month and the month after that, until we have no more noobs here, haha! However, I need you guys to help me out with that in order to make the next thread even better!
Again, thank you all for the feedback
1. We can do it as parts, I mean If one of the pros post their work files part by part (lod0, lod1, high poly, unwrapping, texturing, rigging, masking) people can comtinue from the part they like, I mean if they dont think they need to.model it but they think that they need to practice in texturing, masking and rigging. With this they can download the unwrapped and baked model to continue. I know it can be hard to do but I will drop this here.
2. There can be different concepts as a diffuculty. Like very easy model like this for starters, a little harder one for better, and I hard one like finishing project. If someone finish the hard finishing project he wouldnt be noob anymore. Maybe??
A model that isn't in game would probably keep more interest. It was cool to see everyone working on the same item but could also be cool to watch several concepts be worked on. Or give a basic silhoutte, then everyone could use that to develop their own style weap.
I think that there's no 'challenge' on doing just parts of stuff. The point is to learn new things, and if you just do one or two parts, there's not a lot of new things to learn. Also, how we would choose the parts? Maybe get the full files from someone who got an item ingame, like basemesh, sculpt, lowpoly and textures? I think that it would be a bit hard to get something like that
@Baddcog
I like the silhouette idea! Maybe something like Concept Cookie's Shape Challenge? That would help people make a concept too! I definitely like where this is heading!
One month was too much for this sword. (since it was an easy one too)
At the beginning I joined this somehow cause I wanted to support the idea, but then since I had a lot of time to postpone, I lost interest
There should also be a random silhouette for each weapon type or shield, or at least enough of them so people's items don't look so similar (maybe 3?
Wouldn't this change allow people to submit their item to the workshop, since it's not copying something else and its just puerly from the one person themselves?
If we go with the silhouette idea, the challenge would be much more open, meaning that everyone would do something different and there would be no restrictions. I don't really get that last question, but still, anyone would be able to do anything for any hero, it would just need to match the provided silhouette.
Yeah, having more than one silhouette would be a good idea.
Yes. Since everyone would need to create an original concept just based on a silhouette, you would be able to submit it to the workshop with no issues.
This is my first post and my first model. I only just found out about this activity so I unfortunately will not be able to finish in time. However I look forward to the next one! Sorry I do not have time to leave additional comments.
Anyway, this is the last day on this month, and if possible I'd love to get March's Noob Challenge by monday, so what do you guys think we should do? The silhouette idea, another item which is already in-game or try to get an artist to donate a concept?
also, how should I do presentation images? not promo art but just to display the screenshots.
Yup! If I manage to work out how we're going to do it, I'll post it tomorrow.
@GhostDetector
If possible, a minimum of two screenshots, one from the loadout view and one from the in-game view. If you want to do a promo, that would be really cool too, but it's not really needed.
The new thread is up!
Click here to go directly to it!