First off, xrg's videos are awesome and a big reason I wanted to pick Blender up and give it a try. kat's tutorials were also helpful for clearing up a few points.
I've been trying out Blender for the past couple of days. I think I'm having similar hangups as SHEP did coming from a background using Max. I was foolish at first trying to model a humanoid character (and failed miserably) and switched to a much more newb-friendly sword. I'm sure it's mostly just being unfamiliar and having to unlearn all the muscle memory, but maybe you guys have some answers or solutions.
While I can totally see the usefulness of the 3D cursor, binding it to one of the mouse buttons seems crazy. That's valuable real estate and you map the 3D cursor to it? Seems like more people would use the Snap menu to move the 3D cursor around? How often do you actually use the mouse button to place the 3D cursor? Wouldn't a contextual drop-down be better for the mouse button?
Modeling just felt really weird, like it focuses too much on edge loops and not enough on a "free-for-all" style of modeling. That might be a result of it feeling like certain commands that I thought should be simple took two or three more steps than I thought they should, slowing things down (and being unfamiliar with everything).
Triangulation also felt iffy. There were a couple of instances where I created an edge bridging a gap without any faces connected to it instead of altering the triangulation of the face. Could have been a terminology error on my end.
Unwrapping was a complete mess for me. Felt very different from how I do things in Max. You have to select all the faces to see them in the UV window causing the model to turn orange. So then when you select a face in the UV window how do you see it on the model when the entire thing is already selected? I think again it comes down to the unwrapping workflow being different from how I normally do it.
3D Painting related:
- Wires in the UV window don't go away unless I go into Object Mode.
- I think my Mirror modifier was screwing up painting on the model so I did most of my painting in the UV window. It wasn't a big deal because I was just painting a sword that's mostly projected from the one side, but something to keep in mind when doing a more complex model.
- At first the color selection window annoyed me, but there's an option to change it in Preferences > System. Good to know!
- Colorpicking with a small brush sometimes caused it to pick the color of the brush circumference (red) instead of the color underneath.
- I was looking for some kind of masking, and I found the "Face Selection Mapping" button. Is there a way to turn off the white dotted line after selecting the faces? Is there a better way of masking areas off?
Sorry if it sounds whiney. I know most of it will be solved with familiarizing myself with the program more, something I'm interested in doing. These are just my first impressions. I learned a lot and it felt good to shake things up a bit. Hopefully I'll make a character soon and see what the animation tools are like.
Always nice when my videos inspire people. I really didn't expect even 10 views when I threw them up.
GSoC Paint branch has fixes for a bunch of your issues. Wire view is toggleable; Mirror painting and color picker is fixed. You can usually find builds over at Graphicall.org (GSoC-Paint).
Unwrapping was a complete mess for me. Felt very different from how I do things in Max. You have to select all the faces to see them in the UV window causing the model to turn orange. So then when you select a face in the UV window how do you see it on the model when the entire thing is already selected? I think again it comes down to the unwrapping workflow being different from how I normally do it.
The active UV face is highlighted with small white dots all over it, both in UV and 3D view.
You can make it so you don't have to select the whole mesh to see it in the UV view, though. Try activating this button:
xrg: Sweet, thanks for the resource! I haven't dove into the user-created stuff yet but it seems promising.
Volantk: Thanks, that will definitely solve a few of my problems. Don't know how I missed that button before.
I guess the problem I was having before was less with the faces (although it was annoying that the model was slathered in orange) and more to do with verts and edges. It was impossible to tell what you had selected on the model until you started moving them around and saw how it affected the texture on the model. This little button seems to have fixed that though.
While I can totally see the usefulness of the 3D cursor, binding it to one of the mouse buttons seems crazy. That's valuable real estate and you map the 3D cursor to it? Seems like more people would use the Snap menu to move the 3D cursor around? How often do you actually use the mouse button to place the 3D cursor? Wouldn't a contextual drop-down be better for the mouse button?
The first thing I did was change the behavior of the right and left mouse button. Then I mapped the 3d Cursor to ctrl+shift+right click. Since I almost never use it this isn't as awkward as it might sound.
Then I used my newly liberated right click to act as a hotkey for opening the Mesh Select Mode menu. I've attached an image of this so it'll be clear what I'm talking about. This helped to preserve the spirit my marking menu centric workflow in Maya.
Lastly I set one of my bonus gamer-cool side mouse buttons to open the merge menu that you otherwise have to press alt+m for. These changes removed 90% of the awkwardness from Blender for me.
The first thing I did was change the behavior of the right and left mouse button. Then I mapped the 3d Cursor to ctrl+shift+right click. Since I almost never use it this isn't as awkward as it might sound.
Then I used my newly liberated right click to act as a hotkey for opening the Mesh Select Mode menu. I've attached an image of this so it'll be clear what I'm talking about. This helped to preserve the spirit my marking menu centric workflow in Maya.
Lastly I set one of my bonus gamer-cool side mouse buttons to open the merge menu that you otherwise have to press alt+m for. These changes removed 90% of the awkwardness from Blender for me.
nice tricks. I am so much faster now.
Dont forget to do the same for the UV Editor.
The first thing I did was change the behavior of the right and left mouse button. Then I mapped the 3d Cursor to ctrl+shift+right click. Since I almost never use it this isn't as awkward as it might sound.
Then I used my newly liberated right click to act as a hotkey for opening the Mesh Select Mode menu. I've attached an image of this so it'll be clear what I'm talking about. This helped to preserve the spirit my marking menu centric workflow in Maya.
Lastly I set one of my bonus gamer-cool side mouse buttons to open the merge menu that you otherwise have to press alt+m for. These changes removed 90% of the awkwardness from Blender for me.
One of blender's core developers has a programmer he's worked with that will implement split vertex normals allowing for hard and soft edges.
The goal is to provide game artists and those using Blender in a larger pipeline, to have the same functionality with smooth and hardened edges found in similar apps such as Maya and 3DS Max.
Since Blender is a powerful modeling application, this would be a great boost for artist who need a solid modeling and UV app for game production and freelance.
The lack of vertex normal based smoothing has been preventing many game artist from adopting Blender. This can change. The catch is that this developer still needs to be hired.
The goal is to raise $800 to fund this development. Once theres more detailed information and a link to a kickstarter/fundme/indiegogo page, we can begin gathering donations.
...The goal is to raise $800 to fund this development. Once theres more detailed information and a link to a kickstarter/fundme/indiegogo page, we can begin gathering donations...
IMHO, whilst this is good news, it should be funded through Blender Foundation (or an official channel therein) so as to provide some reasonable expectations by way of guaranty to completion.
IMHO, whilst this is good news, it should be funded through Blender Foundation (or an official channel therein) so as to provide some reasonable expectations by way of guaranty to completion.
I was in the dev irc channel when the discussion took place. Its about as legit as it gets. Cambell set it up, and he's perhaps one of the biggest core contributors to Blenders development.
In the BA forums they asked the same question you asked. The result was a guarantee that it will be supported by the BF after implementation.
I wish it was a BF initialized project too but it seems they have their hands full. They dont seem to stray far from the road map for the next two versions.
Other discussions in the IRC channel involve the future relationship between blender as a asset creation tool and Valve, so ultimately there is interest from the developer side in making Blender game dev friendly.
Alright, change of info. The developer is currently proposing this:
"* Core function (C) takes a mesh, scans it to detect edges marked as sharp, computes a normal for each vertex for each face accordingly, and stores those normals into a Loop CD layer. Note edges considered as sharp do not only include those marked as such, but also all boundary edges, as well as all non-manifold edges (i.e. edges used by more than two faces).
* Then any export (or other) addon needing that data can request it through usual RNA or bmesh APIs. I will add at least support to export those in FBX and OBJ formats."
Would this still be able to address the general needs of those using Blender for game assets? Or is it still lacking the basic functionality for most of you?
From what that sounds like, we're getting smoothing group functionality, but I could be wrong. Never heard of a "Loop CD layer". Doesn't sound like we're getting manual vertex normal editing. In most cases that solves a lot of problems, but in some cases (like foliage), manual editing is necessary. FBX and OBJ exporting would cover all my needs, the only thing lacking is the manual editing.
What he is proposing is, to convert sharp edges to split vertex normals upon export instead of spliting the mesh.
It should work for those who want to export meshes with proper smoothing without broken edges, be it to xnormal or other applications.
This has nothing to do with manual editing.
On an important note, if there are people who need this feature and can chip in
please make yourselves apparent.
I personally think it's a bit of a hacky solution and 800$ seems a bit rich for it as such.
I think Blender Foundation should be most interested in providing this basic compatibility, but it's definitly not the case.
$800 isn't that bad - it's going to take him a few days of work to do, and developers gotta eat. The Blender Community can raise that much pretty rapidly.
Making split vertex normal data available to export (and import?) plugins is a good thing. If that's all I'm going to get, I'll take it, and be mostly happy
What about previewing the hard/soft edges in Blender's viewport? Will we need to physically split the edges in Blender to get an idea of what's going on, then merge them back again and export with hard edges "marked" to get the real split vertex result? I used to have to do that in Modo and it was a PITA.
I've got to say though I think the split vertex normals issue is very related to the way Blender seems to handle calculating and storing vertex normal data in general. For game artists that need to be able to import, export, and edit custom vertex normals (which includes but is not limited to vertex normals that are simply split to enable hard or soft edges) this is an extremely frustrating issue, and IMHO quite frankly a deal breaker in regards to studio adoption.
Import/Export split vertex normals will be enough for many game development use cases, but it certainly won't solve the underlying problem of how Blender handles vertex normals in general. Allowing for fully editable vertex normals would solve this issue for all game development use cases.
See what's being worked on in to address the lack vertex normal editing in Blender these threads (note the frustration and problems caused by how Blender handles vertex normals internally):
IMHO it's better to fully fix this problem then commission a workaround hack even though that workaround would be a pretty good solution for most models created for games.
Look how much went into explaining the simple problem of not being able to preserve smoothing without splitting edges in this thread.
A lot the post are written by Blender users who have no idea about the game art workflow.
Almost no one to test the smoothing groups implementation, it sends a message.
There's really not a lot of people who do serious game related stuff and
not a lot of recognition of the problems even from the users.
Proper support for multiple normals, editing and exporting isn't really going to happen, definitely not from Blender Foundation.
It would require to much effort, and probably too much deep changes for what it's worth to satisfy a small bunch of inactive game artists.
The guy who went pretty far with it scrapped the code, cause it produced too many compatibility issues.
What about previewing the hard/soft edges in Blender's viewport? Will we need to physically split the edges in Blender to get an idea of what's going on, then merge them back again and export with hard edges "marked" to get the real split vertex result?
Are you not familiar with EdgeSplit modifier?
It splits the sharp edges in the viewport, nondestructively when added.
You have to apply it (e.g. on export) to split for good.
_ _
Basically what it all boils down to is
we can have working export of meshes with sharp edges via multiple vertex normals or nothing.
Don't expect anything will be done if you don't chip in.
It splits the sharp edges in the viewport, nondestructively when added.
You have to apply it (e.g. on export) to split for good.
Sorry, I've been trying to get into Blender on and off over the last couple of years so I'm not intimately familiar with how the edge split modifier works... thanks for the clarification I just wanted to be sure what's being proposed will integrate well with existing Blender workflows.
Don't expect anything will be done if you don't chip in.
Yes, I completely understand. I just stated what I did earlier since it was asked and we're on the topic of Blender's vertex normals. Personally, I'm currently learning C programming albeit slowly... maybe one day I will be able to contribute to Blender development in a more meaningful way. Until then all I can do is speak up and make suggestions.
I think the Blender developers are doing a simply awesome job with the manpower, time, and other resources they have available. I truly appreciate their efforts.
Hi! I`m the guy that made the thread at blenderartists. Seeing as some of you guys are interested in supporting this proposal, what would be the prefered donation method for you?
Look how much went into explaining the simple problem of not being able to preserve smoothing without splitting edges in this thread.
A lot the post are written by Blender users who have no idea about the game art workflow.
Almost no one to test the smoothing groups implementation, it sends a message.
There's really not a lot of people who do serious game related stuff and
not a lot of recognition of the problems even from the users.
Proper support for multiple normals, editing and exporting isn't really going to happen, definitely not from Blender Foundation.
It would require to much effort, and probably too much deep changes for what it's worth to satisfy a small bunch of inactive game artists.
The guy who went pretty far with it scrapped the code, cause it produced too many compatibility issues.
Being realistic seems to be based, in this case, on the subjective. One persons "realistic" might not be the "realistic" others see.
Whether you think proper support for multiple normals is going to happen or not is kind of moot, since its up to the core developers. Given the latest proposal for 2.7 and onward, as well as the latest valve connection, it seems that such a concept as multiple normals isnt so "crazy" after all.
I would say whats being realistic is that the core developers are realizing it might be better to work within a larger pipeline than try to be the only pipeline. Now whether this gets results now or later is up for grabs, but I dont think they are adverse to the bigger changes. Its more a matter of when not if.
As for the Blender Artist users (specific forum) not having any idea about the game art workflow, that generalization is probably not very accurate. I think it would be safer to say it that many of them are not familiar with the technicalities, as in artist but not necessarily technical artist.
To be honest, it does seem like a confusing subject because it requires a bit of knowledge on the general working of things, where as an artist will just know whats on the surface. In Maya, you select edge and make it hard or soft, then export. On the surface thats whats apparent. What is going on in the background is not, or how it differs between applications.
It is true, donating will make this happen as its been presented and if that is enough for some then it doesnt hurt to push forward with that proposal.
Look how much went into explaining the simple problem of not being able to preserve smoothing without splitting edges in this thread.
A lot the post are written by Blender users who have no idea about the game art workflow.
Almost no one to test the smoothing groups implementation, it sends a message.
There's really not a lot of people who do serious game related stuff and
not a lot of recognition of the problems even from the users.
Proper support for multiple normals, editing and exporting isn't really going to happen, definitely not from Blender Foundation.
It would require to much effort, and probably too much deep changes for what it's worth to satisfy a small bunch of inactive game artists.
The guy who went pretty far with it scrapped the code, cause it produced too many compatibility issues.
This one isn't on the Blender devs or the Blender Community. This one is on people here. The problem is figuring out what the fuck people actually want. People are quick to say what they don't want, but figuring out what they do is amazingly difficult to do.
People here wanted smoothing groups, not BA, and very few heretested it just like nobody here is currently testing the FBX import. These are things people here need, not the typical Blender users at BA. Therefore, people here need to step up.
As far as I can tell, the people that were bitching about smoothing groups this whole time actually meant editable vertex normals? Even the proposed solution for that isn't right? The discussion is too vague for me to follow. I doubt the devs are going to have the patience to keep guessing what you guys want.
1.The goal of that Bloodwork's thread was precise from the start, it was to preserve smoothing without breaking edges.
2.Campbell didn't solve the issue, cause writing smoothing groups in obj isn't enough.
These are the facts.
Some people started talking about every other kind of shit, including importing custom normals, editing normals, confusing things, saying we don't need this, sharp edges are enough etc.
This exact situation happened in other threads about this btw, hence my opinion about the Blender users.
Therefore, people here need to step up.
That's exactly the point, there's too few game artists on BA and to much people posting bullshit.
Just a thought, frmdbl, could you at least try to not comment in such a provocative and belittling manner? I think it would go a long way if you could comment without disdain and belittlement, while having some articulation that at leaves open to the possibility you can be constructive and professional.
The way you write makes it sound like you want to go around picking fights, and if thats not your intention then being aware of it might be a good way to curb that effect.
As far as I can tell, the people that were bitching about smoothing groups this whole time actually meant editable vertex normals? Even the proposed solution for that isn't right? The discussion is too vague for me to follow. I doubt the devs are going to have the patience to keep guessing what you guys want.
Both are features that are used in game development. I'm sure the developers can understand what we're talking about even if you can't.
If you truly believe that frmdbl, keep the sentiment out of it. Why drag your "anger?" over here to polycount? It makes no sense. If you want this, and think its all about raising money, then do something that is actually productive rather than verbally abuse or provoke the user base in both forums.
This is common sense. If you call a spade a spade, then look in the mirror and state your unproductive actions for what they are.
1.The goal of that Bloodwork's thread was precise from the start, it was to preserve smoothing without breaking edges.
2.Campbell didn't solve the issue, cause writing smoothing groups in obj isn't enough.
These are the facts.
Some people started talking about every other kind of shit, including importing custom normals, editing normals, confusing things, saying we don't need this, sharp edges are enough etc.
This exact situation happened in other threads about this btw, hence my opinion about the Blender users.
That's exactly the point, there's too few game artists on BA and to much people posting bullshit.
i find this situation happens since it seems the blender devs and hte BA community inherently don't like outsiders, and take there suggestions and features requests as people trying to change blender away from what it is, and how the old user base works with it.
at the same time the general user in communities like PC know what they need to do there work, but generally cant be bothered and don't want to make friends and participate with the blender community since all they want to do is work on there own projects be that art or there own code.
in short of ba and blender devs want the big picture they need to listen to more than just there own.
Both are features that are used in game development. I'm sure the developers can understand what we're talking about even if you can't.
No, they don't; that's the problem.
Blender developers do not have experience creating art assets for games using commercial software pipelines. So when people approach them with poorly thought out complaints like "Blender needs smoothing groups like Max!" they have no idea what the fuck they're talking about. Whatever problem exists cannot be inferred when people state their problems that way. Blender Developers do not use Max. Blender Developers do not assemble art assets for games. Especially next gen games.
So assuming the developer doesn't brush them off immediately, they have to play 20 questions to try to work past their shitty communication skills in order to figure out what the problem actually is. At the same time there are other people popping in generating useless noise. "Both features are used in game development!" and "WHY AM I THE ONLY ONE TESTING!?" are examples that do nothing but muddy the waters for a developer trying to understand an issue.
If people didn't propose things with the communication capabilities of Rainman, ya'll would probably be a lot more successful. Ideasman42, the main Blender developer discusses how frustrating it is communicating with people in this Blender Podcast.
i find this situation happens since it seems the blender devs and hte BA community inherently don't like outsiders, and take there suggestions and features requests as people trying to change blender away from what it is, and how the old user base works with it.
at the same time the general user in communities like PC know what they need to do there work, but generally cant be bothered and don't want to make friends and participate with the blender community since all they want to do is work on there own projects be that art or there own code.
in short of ba and blender devs want the big picture they need to listen to more than just there own.
I don't know that they don't like outsiders. It's just literally a Community-run project. They generate revenue off selling t-shirts and DVDs, not licenses. These are affordable to virtually anyone, so it doesn't make much of a difference if a user is a hobbyist or professional from their point of view.
If an influx of users, demonstrating that they do wish to use Blender for Next-Gen games or whatever showed up, more dev would be focused in that area I'm sure. Staying outside the Community, the demand just seems really low I think. Especially compared to users interested in Cycles and things. You get out of it what you put in really.
xrg: then why the hell make the Blender Game Engine? Nobody is asking for Max like smoothing groups - Max users call them smoothing groups, Maya users call them hard and soft edges.
Read Dataday's post, it sounds like one of Blender's devs knows this:
The concerns being voiced aren't new, they as old as Blender itself. The Foundation knows exactly what needs to be done because some of usHAVE been telling them what they need to do, FOR YEARS in fact.
The problem you're missing or perhaps don't have the history to understand is that BF ONLY implements things that are conducive to the direction they are taking Blender, film (Ton himself has said this over the years several times and meant it), everything else takes a back seat and or gets the "Blender is open source so you do it" throwaway response - I can't enumerate the number of times, over the years, some of us have tried to get BF to support some of the things game developers need only for those very requests to be shut down, sabotaged through over complication or "do it our way or the high way". It happens. Too often.
I'm not totally familiar with open source development but why hasn't there been a game branch? There's a flavor of Linux for everyone out there.
It's a shame they seem hell bent on film, a large portion of games use 3D, from the indie up to the big AAA games. For movies that's mostly concentrated to the big budget movies, AAA movies if you will.
There has been talk about it over the years and one or two attempts to at least start a 'game' branch, but without support (or at least acknowledgement) from BF it's all died a death - open source is all very well but without recognition from Blender HQ a lot of those efforts just languish in obscurity, which is one of the reasons why people give up.
Blender developers do not have experience creating art assets for games using commercial software pipelines. So when people approach them with poorly thought out complaints like "Blender needs smoothing groups like Max!" they have no idea what the fuck they're talking about. Whatever problem exists cannot be inferred when people state their problems that way. Blender Developers do not use Max. Blender Developers do not assemble art assets for games. Especially next gen games.
So assuming the developer doesn't brush them off immediately, they have to play 20 questions to try to work past their shitty communication skills in order to figure out what the problem actually is. At the same time there are other people popping in generating useless noise. "Both features are used in game development!" and "WHY AM I THE ONLY ONE TESTING!?" are examples that do nothing but muddy the waters for a developer trying to understand an issue.
If people didn't propose things with the communication capabilities of Rainman, ya'll would probably be a lot more successful. Ideasman42, the main Blender developer discusses how frustrating it is communicating with people in this Blender Podcast.
Are they incapable of using Google? Not to get pissy but I imagine that "3dsmax smoothing groups" would generate a lot of useful information for them complete with screen shots, videos, examples, etc.
The concerns being voiced aren't new, they as old as Blender itself. The Foundation knows exactly what needs to be done because some of usHAVE been telling them what they need to do, FOR YEARS in fact.
The problem you're missing or perhaps don't have the history to understand is that BF ONLY implements things that are conducive to the direction they are taking Blender, film (Ton himself has said this over the years several times and meant it), everything else takes a back seat and or gets the "Blender is open source so you do it" throwaway response - I can't enumerate the number of times, over the years, some of us have tried to get BF to support some of the things game developers need only for those very requests to be shut down, sabotaged through over complication or "do it our way or the high way". It happens. Too often.
Touching up on your post...
There have been a couple of catalyst which indicate a change in direction for Blender, and that's kind of what I have been betting on.
Signs include the proposal by Ton regarding the future of Blender and the ongoing dialog between Valve and Blender (hinting at some sort of steam/steamworkshop integration).
With work being done on threaded drawing and updates, viewport (compositing) effects, unified physics, node based animation, and everything thats currently real-time in Blender already, I also propose to refocus the current game engine to re-use much more of this work.
Or more radically worded: I propose to make the GE to become a real part of Blender code to make it not separated anymore. This would make it more supported, more stable and (Im sure) much more fun to work on as well.
Instead of calling it the GE we would just put Blender in Interaction mode. Topics to think of:
Integrate the concept of Logic in the animation system itself. Rule or behavior based animation is a great step forward for animation as well (like massive anims, or for extras).
Support of all Blender physics.
Optimizing speed for interactive playback will then also benefit regular 3d editing (and vice versa)
Singular Python API for logic scripting
Ensure good I/O integration with external game engines, similar to render engines.
What should then be dropped is the idea to make Blender have an embedded true game engine. We should acknowledge that we never managed to make something with the portability and quality of Unreal or Crysis or even Unity3D. And Blenders GPL license is not helping here much either.
On the positive side I think that the main cool feature of our GE is that it was integrated with a 3D tool, to allow people to make 3D interaction for walkthroughs, for scientific sims, or game prototypes. If we bring back this (original) design focus for a GE, I think we still get something unique and cool, with seamless integration of realtime and offline 3D.
I work for Valve (http://www.valvesoftware.com/company/). We would like to make our digital distribution platform Steam (www.steampowered.com<http://www.steampowered.com>) one of the places where you can download Blender. The long-term goal would be to make it easier for people to build their own mods for PC games with Blender and share these mods with other gamers. So I was wondering if there are any Blender users on this list who are interested in PC games and could see themselves working on an integration between Blender and PC games that offer official modding support such as DOTA 2.
Long story: Valve is a company that is built on modding. The original Half-Life was built on a modified version of the Quake engine. All our major games since then started out as mods which we found cool, hired the people who built them and released them as major game titles. This is true for Counter-Strike, the original Team Fortress, Day of Defeat and DOTA 2 (Portal was not technically a mod but a student project - but you see the pattern). Similarly, one of the most successful features of our Steam platform is the Steam Workshop (http://steamcommunity.com/workshop/), which is an interface for users to share, discover and install mods for their games.
Essentially, you can publish your mod there and other gamers can bring your mod into their games with a single mouse click. This is something that we think would be a cool feature for Blender to tap into. Like modeling a sword in Blender, pushing a button and having it available to all users of Skyrim. But we bet there are more creative ideas out there than this one. What we are currently looking at is offering a completely vanilla version of Blender as a free download on Steam that is completely the same as that offered on other websites. We'd hope that this will get enough of our users exposed to and interested in Blender so they will be inclined to work on Blender plugins that would talk to Steam's backend services such as Workshop.
If you think you might be interested in being part of that, we'd be happy to hear from you!
Best, Jan-Peter
...
Hi all, There have been some really good points brought up here.
The suggestion to ship at least some existing plugins that lend themselves very well to modding sounds like a very sensible one. SMD/DMX plugin seems like an obvious plugin to bundle with a version of Blender. If anybody else wants to suggest other plugins that are useful for modding and should be shipped in this bundle, I'm all ears. The topic Ton brought up - how to do a license-compatible way to send models generated by Blender to another program - is something that I will need guidance on from the experts here. I have looked at the license that ships with Blender, which is GPL V2.
What we want is essentially provide the possibility to send API calls to our online service that make it possible to send data out of Blender (models, textures, animations,...?) there. As I said, the main call I have in mind is about Steam Workshop. My understanding of GPL V2 is that this would in any case not be a problem if we did just a WebAPI call where you send the data via a clearly documented http requests.
At the moment, what the games on Steam that integrate with Steam Workshop do is they send the models to a compiled C program and that sends the data to our server. My understanding of the GPL V2 is that this is still ok since we are not making function calls that will hide functionality - we are only sending off data, the same way that we would if we exported a file locally and then uploaded the data manually via the Steam Workshop website, except we are doing this in one mouse click.
Is this a fair assumption? If not, what would be the license-compatible way to export models from Blender to a different program?
Best, Jan
The following back in forth between Ton and Valve's Jan P. involves getting Blender out there so game artist and modders know what it does and can do.
Bringing all this up again, I hope to highlight that Blender might be on the verge of really focusing on the game development/artist needs. It is also the time where the rest of the game artist out there need to get more vocal and help influence the changes needed in order to turn Blender into a powerful tool for game art, and the larger pipeline. Doing so is good even if its just to have a fallback application with no cost to the user.
Additionally Blender's success drives other major applications to be more competitive and accessible.
Previously such requests were hard to sell, and you are right, Ton and co seemed focused on film. Now there's a shift in that mindset, all it takes is a firm push to keep it going in a more pro-game route. To make it happen (faster?) this requires more feedback to come in from users outside of the Blender Artist forums, or at least fill the forums with more Polycount users as well.
Wishful thinking perhaps, but with Steam Workshop integration on the horizon (it could just be a mirage...) maybe Valve will fork Blender and add the features we need for game art.
With respect to Value (and broader issues regarding game development uptake) I can only temper this conversation with what I've experienced concerning the MD3 & MD5 formats.
They are both very popular community driven tools. Often used by educational establishments teaching 3D game programming, but also by a surprising number of indie game studios. That is until they realise the tools break each time there's a Blender update (every two or so months at present).
They may stay-the-course once or twice through these breakages but eventually have to go elsewhere when they realise just how much time and resources have been lost to production checks that shouldn't really be necessary. In that kind of environment are studios and professionals (which are the crux of this discussion) *really* going to invest in Blender? Or are they going to go elsewhere, i.e. 3DS Max and/or other industry standard applications?.
This is one of the main reasons why companies like Valve throw support for products like Blender at the community; it means they don't waste resources having to deal with the fast update cycle and the consequences of that. It makes application adoption fundamentally unreliable so they just won't do it themselves (id software did this with both the MD3 & MD5 formats, Bethesda with support for Morrowind/Skyrm, Bioware(?) for NeverWinter Nights tools... the list is endless).
Indecently, it's not without a certain sense of irony that the fastest adoption rates for commercialised use of Blender occur in advertising (particularly in so called 'developing' nations - Brazil, India et-al) - why? Because the short term, quick turnaround projects they usually engage in fit into Blenders equally frequent update/release cycle.
Here's the website where you can donate to the aforementioned project, that is aimed at bringing the export of sharp edges without splitting geometry to Blender.
Wishful thinking perhaps, but with Steam Workshop integration on the horizon (it could just be a mirage...) maybe Valve will fork Blender and add the features we need for game art.
Unlikely I think - more likely they would just donate back to the project and keep it in one branch. Blender Foundation is open to other contributions, they just don't want people putting code in and running away leaving them to maintain it.
We have already reached 50+% of the required funding! Even a single day hasn`t passed sisnce the begining of the fundraiser. http://mont29.wordpress.com/
Replies
I've been trying out Blender for the past couple of days. I think I'm having similar hangups as SHEP did coming from a background using Max. I was foolish at first trying to model a humanoid character (and failed miserably) and switched to a much more newb-friendly sword. I'm sure it's mostly just being unfamiliar and having to unlearn all the muscle memory, but maybe you guys have some answers or solutions.
While I can totally see the usefulness of the 3D cursor, binding it to one of the mouse buttons seems crazy. That's valuable real estate and you map the 3D cursor to it? Seems like more people would use the Snap menu to move the 3D cursor around? How often do you actually use the mouse button to place the 3D cursor? Wouldn't a contextual drop-down be better for the mouse button?
Modeling just felt really weird, like it focuses too much on edge loops and not enough on a "free-for-all" style of modeling. That might be a result of it feeling like certain commands that I thought should be simple took two or three more steps than I thought they should, slowing things down (and being unfamiliar with everything).
Triangulation also felt iffy. There were a couple of instances where I created an edge bridging a gap without any faces connected to it instead of altering the triangulation of the face. Could have been a terminology error on my end.
Unwrapping was a complete mess for me. Felt very different from how I do things in Max. You have to select all the faces to see them in the UV window causing the model to turn orange. So then when you select a face in the UV window how do you see it on the model when the entire thing is already selected? I think again it comes down to the unwrapping workflow being different from how I normally do it.
3D Painting related:
- Wires in the UV window don't go away unless I go into Object Mode.
- I think my Mirror modifier was screwing up painting on the model so I did most of my painting in the UV window. It wasn't a big deal because I was just painting a sword that's mostly projected from the one side, but something to keep in mind when doing a more complex model.
- At first the color selection window annoyed me, but there's an option to change it in Preferences > System. Good to know!
- Colorpicking with a small brush sometimes caused it to pick the color of the brush circumference (red) instead of the color underneath.
- I was looking for some kind of masking, and I found the "Face Selection Mapping" button. Is there a way to turn off the white dotted line after selecting the faces? Is there a better way of masking areas off?
Sorry if it sounds whiney. I know most of it will be solved with familiarizing myself with the program more, something I'm interested in doing. These are just my first impressions. I learned a lot and it felt good to shake things up a bit. Hopefully I'll make a character soon and see what the animation tools are like.
GSoC Paint branch has fixes for a bunch of your issues. Wire view is toggleable; Mirror painting and color picker is fixed. You can usually find builds over at Graphicall.org (GSoC-Paint).
These are videos he made for his midterm evaluation. Not sure what else he has planned.
[ame="
[ame="
You can make it so you don't have to select the whole mesh to see it in the UV view, though. Try activating this button:
Volantk: Thanks, that will definitely solve a few of my problems. Don't know how I missed that button before.
I guess the problem I was having before was less with the faces (although it was annoying that the model was slathered in orange) and more to do with verts and edges. It was impossible to tell what you had selected on the model until you started moving them around and saw how it affected the texture on the model. This little button seems to have fixed that though.
The first thing I did was change the behavior of the right and left mouse button. Then I mapped the 3d Cursor to ctrl+shift+right click. Since I almost never use it this isn't as awkward as it might sound.
Then I used my newly liberated right click to act as a hotkey for opening the Mesh Select Mode menu. I've attached an image of this so it'll be clear what I'm talking about. This helped to preserve the spirit my marking menu centric workflow in Maya.
Lastly I set one of my bonus gamer-cool side mouse buttons to open the merge menu that you otherwise have to press alt+m for. These changes removed 90% of the awkwardness from Blender for me.
nice tricks. I am so much faster now.
Dont forget to do the same for the UV Editor.
This pie menu addon looks alot like maya:
http://blenderartists.org/forum/showthread.php?267414-Pie-Menus-%282-66%29-Update-01-28-13
One of blender's core developers has a programmer he's worked with that will implement split vertex normals allowing for hard and soft edges.
The goal is to provide game artists and those using Blender in a larger pipeline, to have the same functionality with smooth and hardened edges found in similar apps such as Maya and 3DS Max.
Since Blender is a powerful modeling application, this would be a great boost for artist who need a solid modeling and UV app for game production and freelance.
The lack of vertex normal based smoothing has been preventing many game artist from adopting Blender. This can change. The catch is that this developer still needs to be hired.
The goal is to raise $800 to fund this development. Once theres more detailed information and a link to a kickstarter/fundme/indiegogo page, we can begin gathering donations.
The developers profile is here: https://www.ohloh.net/p/blender/contributors/274877939555
Stay tuned for more info.
Thanks for posting this here Dataday, I might have missed it otherwise
I was in the dev irc channel when the discussion took place. Its about as legit as it gets. Cambell set it up, and he's perhaps one of the biggest core contributors to Blenders development.
In the BA forums they asked the same question you asked. The result was a guarantee that it will be supported by the BF after implementation.
I wish it was a BF initialized project too but it seems they have their hands full. They dont seem to stray far from the road map for the next two versions.
Other discussions in the IRC channel involve the future relationship between blender as a asset creation tool and Valve, so ultimately there is interest from the developer side in making Blender game dev friendly.
NP, getting the word out is how we can make sure this kind of development gets funded faster. It takes a community to push such requests.
"* Core function (C) takes a mesh, scans it to detect edges marked as sharp, computes a normal for each vertex for each face accordingly, and stores those normals into a Loop CD layer. Note edges considered as sharp do not only include those marked as such, but also all boundary edges, as well as all non-manifold edges (i.e. edges used by more than two faces).
* Then any export (or other) addon needing that data can request it through usual RNA or bmesh APIs. I will add at least support to export those in FBX and OBJ formats."
Would this still be able to address the general needs of those using Blender for game assets? Or is it still lacking the basic functionality for most of you?
It should work for those who want to export meshes with proper smoothing without broken edges, be it to xnormal or other applications.
This has nothing to do with manual editing.
On an important note, if there are people who need this feature and can chip in
please make yourselves apparent.
I personally think it's a bit of a hacky solution and 800$ seems a bit rich for it as such.
I think Blender Foundation should be most interested in providing this basic compatibility, but it's definitly not the case.
http://lists.blender.org/pipermail/soc-2013-dev/2013-August/000272.html
Making split vertex normal data available to export (and import?) plugins is a good thing. If that's all I'm going to get, I'll take it, and be mostly happy
What about previewing the hard/soft edges in Blender's viewport? Will we need to physically split the edges in Blender to get an idea of what's going on, then merge them back again and export with hard edges "marked" to get the real split vertex result? I used to have to do that in Modo and it was a PITA.
I've got to say though I think the split vertex normals issue is very related to the way Blender seems to handle calculating and storing vertex normal data in general. For game artists that need to be able to import, export, and edit custom vertex normals (which includes but is not limited to vertex normals that are simply split to enable hard or soft edges) this is an extremely frustrating issue, and IMHO quite frankly a deal breaker in regards to studio adoption.
Import/Export split vertex normals will be enough for many game development use cases, but it certainly won't solve the underlying problem of how Blender handles vertex normals in general. Allowing for fully editable vertex normals would solve this issue for all game development use cases.
See what's being worked on in to address the lack vertex normal editing in Blender these threads (note the frustration and problems caused by how Blender handles vertex normals internally):
http://blenderartists.org/forum/showthread.php?251402-Addon-Recalc-Vertex-Normals
http://blenderartists.org/forum/showthread.php?259554-Addon-EditNormals-Transfer-Vertex-Normals
http://blenderartists.org/forum/showthread.php?284736-Normals-Editing-in-Blender
As an example of what I'm talking about. We need to be able import, export, and edit vertex normals as seen in this foliage example: http://wiki.polycount.com/VertexNormal#Foliage_Shading
IMHO it's better to fully fix this problem then commission a workaround hack even though that workaround would be a pretty good solution for most models created for games.
Look how much went into explaining the simple problem of not being able to preserve smoothing without splitting edges in this thread.
A lot the post are written by Blender users who have no idea about the game art workflow.
Almost no one to test the smoothing groups implementation, it sends a message.
There's really not a lot of people who do serious game related stuff and
not a lot of recognition of the problems even from the users.
Proper support for multiple normals, editing and exporting isn't really going to happen, definitely not from Blender Foundation.
It would require to much effort, and probably too much deep changes for what it's worth to satisfy a small bunch of inactive game artists.
The guy who went pretty far with it scrapped the code, cause it produced too many compatibility issues.
@Ben Apuna
Are you not familiar with EdgeSplit modifier?
It splits the sharp edges in the viewport, nondestructively when added.
You have to apply it (e.g. on export) to split for good.
_ _
Basically what it all boils down to is
we can have working export of meshes with sharp edges via multiple vertex normals or nothing.
Don't expect anything will be done if you don't chip in.
Sorry, I've been trying to get into Blender on and off over the last couple of years so I'm not intimately familiar with how the edge split modifier works... thanks for the clarification
If we can get this, then it'll be very useful for many game artists, and much appreciated
Yes, I completely understand. I just stated what I did earlier since it was asked and we're on the topic of Blender's vertex normals. Personally, I'm currently learning C programming albeit slowly... maybe one day I will be able to contribute to Blender development in a more meaningful way. Until then all I can do is speak up and make suggestions.
I think the Blender developers are doing a simply awesome job with the manpower, time, and other resources they have available. I truly appreciate their efforts.
Being realistic seems to be based, in this case, on the subjective. One persons "realistic" might not be the "realistic" others see.
Whether you think proper support for multiple normals is going to happen or not is kind of moot, since its up to the core developers. Given the latest proposal for 2.7 and onward, as well as the latest valve connection, it seems that such a concept as multiple normals isnt so "crazy" after all.
I would say whats being realistic is that the core developers are realizing it might be better to work within a larger pipeline than try to be the only pipeline. Now whether this gets results now or later is up for grabs, but I dont think they are adverse to the bigger changes. Its more a matter of when not if.
As for the Blender Artist users (specific forum) not having any idea about the game art workflow, that generalization is probably not very accurate. I think it would be safer to say it that many of them are not familiar with the technicalities, as in artist but not necessarily technical artist.
To be honest, it does seem like a confusing subject because it requires a bit of knowledge on the general working of things, where as an artist will just know whats on the surface. In Maya, you select edge and make it hard or soft, then export. On the surface thats whats apparent. What is going on in the background is not, or how it differs between applications.
It is true, donating will make this happen as its been presented and if that is enough for some then it doesnt hurt to push forward with that proposal.
This one isn't on the Blender devs or the Blender Community. This one is on people here. The problem is figuring out what the fuck people actually want. People are quick to say what they don't want, but figuring out what they do is amazingly difficult to do.
People here wanted smoothing groups, not BA, and very few here tested it just like nobody here is currently testing the FBX import. These are things people here need, not the typical Blender users at BA. Therefore, people here need to step up.
As far as I can tell, the people that were bitching about smoothing groups this whole time actually meant editable vertex normals? Even the proposed solution for that isn't right? The discussion is too vague for me to follow. I doubt the devs are going to have the patience to keep guessing what you guys want.
1.The goal of that Bloodwork's thread was precise from the start, it was to preserve smoothing without breaking edges.
2.Campbell didn't solve the issue, cause writing smoothing groups in obj isn't enough.
These are the facts.
Some people started talking about every other kind of shit, including importing custom normals, editing normals, confusing things, saying we don't need this, sharp edges are enough etc.
This exact situation happened in other threads about this btw, hence my opinion about the Blender users.
That's exactly the point, there's too few game artists on BA and to much people posting bullshit.
The way you write makes it sound like you want to go around picking fights, and if thats not your intention then being aware of it might be a good way to curb that effect.
Both are features that are used in game development. I'm sure the developers can understand what we're talking about even if you can't.
I call a spade a spade, I've put a lot of effort into that thread and it was flooded
with off topic posts.
If you don't agree please reply with arguments not pointless banter or better yet let's focus on what's important which is to raise the money.
This is common sense. If you call a spade a spade, then look in the mirror and state your unproductive actions for what they are.
i find this situation happens since it seems the blender devs and hte BA community inherently don't like outsiders, and take there suggestions and features requests as people trying to change blender away from what it is, and how the old user base works with it.
at the same time the general user in communities like PC know what they need to do there work, but generally cant be bothered and don't want to make friends and participate with the blender community since all they want to do is work on there own projects be that art or there own code.
in short of ba and blender devs want the big picture they need to listen to more than just there own.
No, they don't; that's the problem.
Blender developers do not have experience creating art assets for games using commercial software pipelines. So when people approach them with poorly thought out complaints like "Blender needs smoothing groups like Max!" they have no idea what the fuck they're talking about. Whatever problem exists cannot be inferred when people state their problems that way. Blender Developers do not use Max. Blender Developers do not assemble art assets for games. Especially next gen games.
So assuming the developer doesn't brush them off immediately, they have to play 20 questions to try to work past their shitty communication skills in order to figure out what the problem actually is. At the same time there are other people popping in generating useless noise. "Both features are used in game development!" and "WHY AM I THE ONLY ONE TESTING!?" are examples that do nothing but muddy the waters for a developer trying to understand an issue.
If people didn't propose things with the communication capabilities of Rainman, ya'll would probably be a lot more successful. Ideasman42, the main Blender developer discusses how frustrating it is communicating with people in this Blender Podcast.
I don't know that they don't like outsiders. It's just literally a Community-run project. They generate revenue off selling t-shirts and DVDs, not licenses. These are affordable to virtually anyone, so it doesn't make much of a difference if a user is a hobbyist or professional from their point of view.
If an influx of users, demonstrating that they do wish to use Blender for Next-Gen games or whatever showed up, more dev would be focused in that area I'm sure. Staying outside the Community, the demand just seems really low I think. Especially compared to users interested in Cycles and things. You get out of it what you put in really.
xrg: then why the hell make the Blender Game Engine? Nobody is asking for Max like smoothing groups - Max users call them smoothing groups, Maya users call them hard and soft edges.
Read Dataday's post, it sounds like one of Blender's devs knows this:
That's a feature that's existed in most other 3d apps for about 15 years I believe. I'm sure they saw a white-paper or two about it.
I don't know about the devs but there's enough people like you that give off the vibe that it's an insular community that doesn't like outsiders.
The problem you're missing or perhaps don't have the history to understand is that BF ONLY implements things that are conducive to the direction they are taking Blender, film (Ton himself has said this over the years several times and meant it), everything else takes a back seat and or gets the "Blender is open source so you do it" throwaway response - I can't enumerate the number of times, over the years, some of us have tried to get BF to support some of the things game developers need only for those very requests to be shut down, sabotaged through over complication or "do it our way or the high way". It happens. Too often.
It's a shame they seem hell bent on film, a large portion of games use 3D, from the indie up to the big AAA games. For movies that's mostly concentrated to the big budget movies, AAA movies if you will.
Touching up on your post...
There have been a couple of catalyst which indicate a change in direction for Blender, and that's kind of what I have been betting on.
Signs include the proposal by Ton regarding the future of Blender and the ongoing dialog between Valve and Blender (hinting at some sort of steam/steamworkshop integration).
The 2.7 and 2.8 road map proposals Ton released not that long ago.
The highlight from that proposal is this:
Valve Connection:
http://lists.blender.org/pipermail/bf-committers/2013-August/041483.html
The following back in forth between Ton and Valve's Jan P. involves getting Blender out there so game artist and modders know what it does and can do.
Bringing all this up again, I hope to highlight that Blender might be on the verge of really focusing on the game development/artist needs. It is also the time where the rest of the game artist out there need to get more vocal and help influence the changes needed in order to turn Blender into a powerful tool for game art, and the larger pipeline. Doing so is good even if its just to have a fallback application with no cost to the user.
Additionally Blender's success drives other major applications to be more competitive and accessible.
Previously such requests were hard to sell, and you are right, Ton and co seemed focused on film. Now there's a shift in that mindset, all it takes is a firm push to keep it going in a more pro-game route. To make it happen (faster?) this requires more feedback to come in from users outside of the Blender Artist forums, or at least fill the forums with more Polycount users as well.
They are both very popular community driven tools. Often used by educational establishments teaching 3D game programming, but also by a surprising number of indie game studios. That is until they realise the tools break each time there's a Blender update (every two or so months at present).
They may stay-the-course once or twice through these breakages but eventually have to go elsewhere when they realise just how much time and resources have been lost to production checks that shouldn't really be necessary. In that kind of environment are studios and professionals (which are the crux of this discussion) *really* going to invest in Blender? Or are they going to go elsewhere, i.e. 3DS Max and/or other industry standard applications?.
This is one of the main reasons why companies like Valve throw support for products like Blender at the community; it means they don't waste resources having to deal with the fast update cycle and the consequences of that. It makes application adoption fundamentally unreliable so they just won't do it themselves (id software did this with both the MD3 & MD5 formats, Bethesda with support for Morrowind/Skyrm, Bioware(?) for NeverWinter Nights tools... the list is endless).
Indecently, it's not without a certain sense of irony that the fastest adoption rates for commercialised use of Blender occur in advertising (particularly in so called 'developing' nations - Brazil, India et-al) - why? Because the short term, quick turnaround projects they usually engage in fit into Blenders equally frequent update/release cycle.
http://mont29.wordpress.com
Also, there's a very detailed explanation.
You can read up on Valve discussion here on the mailing list:
http://lists.blender.org/pipermail/bf-committers/2013-August/thread.html
(just do a search for "Steam")