...would be nice if you could turn off auto actions all together tbh as I always only want it for a rig
If you open the timeline, in the header you can enable/disable auto keying. In addition, you can enable the button next to it (active keying set only) and also pick "whole character" keying set. Now only the rig bones will get auto keyed and only their channels which haven't been locked.
I'm having a strange problem the I've never had before. I'll try to export my low poly with a cage, but xNormal complains that it doesn't match with low poly.
I've triangulated my low poly model, duplicated it to a cage. Inflated the cage, and exported both to .obj.
Sometimes I see Blender flickering the shading at times when sculpting or editing the model. But there shouldn't be any difference between the low poly and the same duplicated low poly cage.
I can't use Blender for baking normals now, because the high poly from Zbrush is about 1 GB. It definitely wouldn't handle the HP in the viewport.
flickering is often two faces superimposed. Try a W / remove doubles or look if you didn't have n-gons under your faces. It append often to me when I do F instead of J to join edges.
flickering is often two faces superimposed. Try a W / remove doubles or look if you didn't have n-gons under your faces. It append often to me when I do F instead of J to join edges.
The model is originally made with quads and triangles, no n-gons. And remove doubles removes 0 vertices. So, it's a solid piece of mesh.
EDIT: I think I figured it out, probably. I have tick the "Keep Vertex Order" in Blender's OBJ export setting for both low polies, then it seems to work. I'll have to test more...
I can't use Blender for baking normals now, because the high poly from Zbrush is about 1 GB. It definitely wouldn't handle the HP in the viewport.
Set viewport draw type to bounding box, import the highpoly, select it in the outliner and name it appropriately, and then in the Properties panel->Object tab (i.e. the yellow cube)->Display rollout set Maximum Draw Type to Bounding Box for the HP. This will let you see your lowpoly meshes easily once you set the overall draw mode back up to Solid or whatever.
Now the speed of the baking might be a bit of a different story, but if you can get a nice high-res bake running overnight it shouldn't be a problem.
Set viewport draw type to bounding box, import the highpoly, select it in the outliner and name it appropriately, and then in the Properties panel->Object tab (i.e. the yellow cube)->Display rollout set Maximum Draw Type to Bounding Box for the HP. This will let you see your lowpoly meshes easily once you set the overall draw mode back up to Solid or whatever.
Now the speed of the baking might be a bit of a different story, but if you can get a nice high-res bake running overnight it shouldn't be a problem.
I wouldn't consider an normal map bake taking a whole night to be acceptable though.
Two solutions for that :
1 - bake in Xnormal, which is a specialized tool. It bakes in seconds/minutes.
2 - reduce the model to be imported into Blender using decimation master. It is extremely effective !
(I personally rely on both concepts : baking in Xnormal, using a decimated high. That way bakes are really, really fast!)
Yeah when you get down to it a lot of these applications are more similar than they are different.
If you don't have Zbrush for whatever reason and you don't want to use the decimate modifier, you can also do the decimation using Meshlab.
It's hard to say exactly how long a particular bake would take. You don't need eleventy shizillion Cycles samples for a normal map, but you might for a lightmap. On the other hand Blender seems to take quite a while to fill up an 8k image.
The main issue (now that the memory issue is ameliorated) that I have with Blender baking is that there's no real antialiasing, so you basically have to bake things at 4k or 8k and scale down to get a good result. Other than that it's made quite a bit of progress since the old days. I still think it could be improved, though, and xNormal is a testament to that.
BTW, I thought I'd share this - a guide covering the different OBJ export settings as per 2.74 and their interpretation by the Max2010(Guruwave) and Zbrush 4r7 OBJ importers. As far as my own usage is concerned "Custom export settings 3" is the most versatile setup.
I hope this helps - and let me know if you spot any errors !
Hi direwolf, you can use in the outliner
shortcut B (boxselect)
the lights are highlighted, now use'rightclick'
in the menu you can then use 'select'
now those lights are selected in the 3D Viewport.
BTW, I thought I'd share this - a guide covering the different OBJ export settings as per 2.74 and their interpretation by the Max2010(Guruwave) and Zbrush 4r7 OBJ importers. As far as my own usage is concerned "Custom export settings 3" is the most versatile setup.
I hope this helps - and let me know if you spot any errors !
Pior maybe you can help me out on something that I've been trying to figure out and banging my head against the wall about the GoB export settings from Blender to Zbrush. I'm still pretty new to Blender. I'm trying to export a new retopoed mesh back to Zbrush from Blender using the Gob addon but it seems that the mesh itself won't align with the original mesh. I thought that by pressing either the "Insert" or "Append" Subtool that the mesh itself would correct itself by following the same offset measurements as the currently selected subtool but it doesn't seem to do the trick.
I've searched for any solutions on Google and all I can find mostly that have the closest solution to the problem is the GoB thread in Zbrushcentral. I read on that thread that you have to update Goz in order to correct the offset problems in Goz. The thing is that I'm working with R7 which I think Goz is already updated. Am I missing some export settings or maybe my files are put in the wrong places? I am at a loss here.
I've tried everything that I could think to fix the offset problems with GoB export to Zbrush. One solution that I found is to "Unify" all the Subtools but that would mean that the scale of all the Subtools would adhere to Zbrush's scale which in turn make everything tiny when you do export it out to other 3d applications. I'm using the "Julie.ztl" by the way to conduct my tests. The only other solution that I found is to physically move the retopoed mesh in Zbrush using the Transpose tool and align it as best as I can to the original mesh and then Goz it back to Blender and apply the "Shrinkwrap Update" in order for the retopoed mesh to stick onto the hi poly reference mesh.
@myclay thanks! However, the outliner can only display around 20 lines of objects on my screen. It looks like I have to go
B > hi light 20 lights, scroll down > B > hi light more lights > scroll down > over and over a few times to get all the lights.
Is there a faster way than this?
Not without a hack.
- Parent them to an empty, curve or whatever. You can then select Children/ChildrenRecorsive (whole children tree)
- Use the default Layers system (There is an Addon that is shipped with Blender that allows Layer Naming and such)
- Use the Super Grouper Addon
I was going to post it yesterday (I've read an article on Blendernation), but I found it to be really buggy.
If you select a face and set it to Face, it points the normal of the nearby face to that direction (If I've understood correctly the function, it should do the same as selecting the face edges and setting them to hard), then if you want to correct the mess you have done, the Smooth and Flat will produce crazy results as this:
It is not the same fuction of hardedges, well, it is not ONLY that.
Here looks like you have marked all the edges as hard-edges, than you assign the "face" option to all the faces, but this is meanless, once you setted the hard edges the vertex normals works correctly by default. the face option is intend to be used to help to fake bevels as I did in my example before.
If you want to set hard edges with this addon you should play with the "split normals" feature.
My suggestion is to use only this addon or only the Blender tools to edit the vertex normals, not both together in the same time.
Actually I was reversing the operations done on the Normal Editor, by trying to harden the edges back, with that result, Smooth does something weird too.
If I deactive the Normal Editor, the editing is lost.
On a cube, activate the Normal Editor, select a face, and click on one of the 3 options (it doesn't matter), you don't like the result, and click on Restore. Now your face is Smooth, even though your cube was on Flat.
If you then click on Flat, to put it back as its previous state, nothing happens, same with Smooth, but if you select all and click on one of these options, the weird stuff I showed in the previous post happens.
The only way to reverse this mess is to select all the faces, click on Tree (you may need to do this twice, disabling and enabling the Normal Editor), then click on Smooth, and finally on Flat. In this order.
wow, well I think I'm going to use this addon just for few selected tasks, like the one I posted before, for the rest I think I stay on the default blender tools. eheh
Yeah you basically set up your rigid bodies how you want them, then add them to the seamless area where each rigid body gets surrounded by passive copies of itself. The passive copies are offset the dimensions of the seamless area and mirror the active with constraints.
Basically if an active rigid body goes too far away from the others, it interacts with passive copies instead of other actives, while the other actives interact with its passive copies.
The size of the seamless area can be changed at any time, even after its full of objects (although you will have to rerun the simulation). The tools also allow an orthographic camera to be automatically created and scaled with the seamless area as well. Also this is not solely for rigid bodies, it can be used on almost any object. It also comes with node groups for cycles to ease material creation for seamless objects.
There is documentation included with the addon that I hope explains it well enough.
When I have perfectly overlap UV island, UV Editor auto-weld everything. How to stop this? This case a lot of problems with normals etc. @Edit
One of the biggest problem is that if I overlaping two the same UV-islands (mirrored), I have big problem with normal map in game engine. Even If I import this to 3ds max it seems that Blender merged that. I need to move mirroed part to second UV block and then it works good.
Maybe you saw Ton's statements in the other topic of the Blender developement ; the main goal for the next release (in a year) is to upgrade the viewport to handle millions of polys.
Currently the polycount limitation is quite low. There's some workaround by enabling VBO and disabling double sided but to sump it up, no blender can't handle a lot more.
Wow, that almost means that we can throw the exported OBJ models with the highest sub-d level from ZBrush to Blender's viewport. Nowadays I'll need to either set one or two steps back with subdivisions, or decimate beforehand though.
can you actually edit such a high density mesh though without having the app drop to a snail's pace? or is it just some caching to not have the viewport lag at every step? the latter can be found in at least a certain autodesk 3d app and is only partially useful. <bicker bicker bicker>
can you actually edit such a high density mesh though without having the app drop to a snail's pace? or is it just some caching to not have the viewport lag at every step? the latter can be found in at least a certain autodesk 3d app and is only partially useful. <bicker bicker bicker>
Was my thought as well. "Yes, but can you edit them in real time?"
Are you seriously want EDIT that objects from video about 129m?
no. just be able to edit complex geometry and have dev resources focused on speeding up the actual editing tools. you never know when you need the extra performance. in comparison being able to rotate stupid amounts of polygons in the viewport with simple materials thrown on is less useful in general.
there seems that underlying and necessary code re factoring has begun to
be able to edit such massive amounts of data...(129M triangles) albeit not showing yet in pictures due to its early state.
in comparison being able to rotate stupid amounts of polygons in the viewport with simple materials thrown on is less useful in general.
I guess it depends on the workflow - I personally rely quite a bit on having dense (not edit-mode friendly, but at least sculpt-friendly) models in my Blender scenes, so any boost in that regard is welcome ! That being said, 2.74 already behaves quite well when dealing with decimated models from Zbrush, so it is quite workable even now.
Can't wait to see where all this is going. Exciting times !
no. just be able to edit complex geometry and have dev resources focused on speeding up the actual editing tools. you never know when you need the extra performance. in comparison being able to rotate stupid amounts of polygons in the viewport with simple materials thrown on is less useful in general.
From a rendering stand point tho, heavy scene navigation is a very welcome feature Even Maya choke up when I try to fly around checking things in final render scenes, and the poly count wasn't anywhere near 100M.
Replies
Looks like the remesh modifier. Great for rounding off mechanical type geometry created with boolTool.
If you open the timeline, in the header you can enable/disable auto keying. In addition, you can enable the button next to it (active keying set only) and also pick "whole character" keying set. Now only the rig bones will get auto keyed and only their channels which haven't been locked.
I've triangulated my low poly model, duplicated it to a cage. Inflated the cage, and exported both to .obj.
Sometimes I see Blender flickering the shading at times when sculpting or editing the model. But there shouldn't be any difference between the low poly and the same duplicated low poly cage.
I can't use Blender for baking normals now, because the high poly from Zbrush is about 1 GB. It definitely wouldn't handle the HP in the viewport.
The model is originally made with quads and triangles, no n-gons. And remove doubles removes 0 vertices. So, it's a solid piece of mesh.
EDIT: I think I figured it out, probably. I have tick the "Keep Vertex Order" in Blender's OBJ export setting for both low polies, then it seems to work. I'll have to test more...
thanks! if you are interested in nanomesh kind of stuff take a look at THIS ADDON
Set viewport draw type to bounding box, import the highpoly, select it in the outliner and name it appropriately, and then in the Properties panel->Object tab (i.e. the yellow cube)->Display rollout set Maximum Draw Type to Bounding Box for the HP. This will let you see your lowpoly meshes easily once you set the overall draw mode back up to Solid or whatever.
Now the speed of the baking might be a bit of a different story, but if you can get a nice high-res bake running overnight it shouldn't be a problem.
haha so close to the way you do it in max
Two solutions for that :
1 - bake in Xnormal, which is a specialized tool. It bakes in seconds/minutes.
2 - reduce the model to be imported into Blender using decimation master. It is extremely effective !
(I personally rely on both concepts : baking in Xnormal, using a decimated high. That way bakes are really, really fast!)
If you don't have Zbrush for whatever reason and you don't want to use the decimate modifier, you can also do the decimation using Meshlab.
It's hard to say exactly how long a particular bake would take. You don't need eleventy shizillion Cycles samples for a normal map, but you might for a lightmap. On the other hand Blender seems to take quite a while to fill up an 8k image.
The main issue (now that the memory issue is ameliorated) that I have with Blender baking is that there's no real antialiasing, so you basically have to bake things at 4k or 8k and scale down to get a good result. Other than that it's made quite a bit of progress since the old days. I still think it could be improved, though, and xNormal is a testament to that.
I hope this helps - and let me know if you spot any errors !
lamp1
lamp2
...
...
lamp99
Is there a I way I can do something similar to Windows or Maya: left click lamp1, shift left click lamp50 for example to select the lamp1-lamp50?
Hi direwolf, you can use in the outliner
shortcut B (boxselect)
the lights are highlighted, now use 'rightclick'
in the menu you can then use 'select'
now those lights are selected in the 3D Viewport.
Pior maybe you can help me out on something that I've been trying to figure out and banging my head against the wall about the GoB export settings from Blender to Zbrush. I'm still pretty new to Blender. I'm trying to export a new retopoed mesh back to Zbrush from Blender using the Gob addon but it seems that the mesh itself won't align with the original mesh. I thought that by pressing either the "Insert" or "Append" Subtool that the mesh itself would correct itself by following the same offset measurements as the currently selected subtool but it doesn't seem to do the trick.
I've searched for any solutions on Google and all I can find mostly that have the closest solution to the problem is the GoB thread in Zbrushcentral. I read on that thread that you have to update Goz in order to correct the offset problems in Goz. The thing is that I'm working with R7 which I think Goz is already updated. Am I missing some export settings or maybe my files are put in the wrong places? I am at a loss here.
I've tried everything that I could think to fix the offset problems with GoB export to Zbrush. One solution that I found is to "Unify" all the Subtools but that would mean that the scale of all the Subtools would adhere to Zbrush's scale which in turn make everything tiny when you do export it out to other 3d applications. I'm using the "Julie.ztl" by the way to conduct my tests. The only other solution that I found is to physically move the retopoed mesh in Zbrush using the Transpose tool and align it as best as I can to the original mesh and then Goz it back to Blender and apply the "Shrinkwrap Update" in order for the retopoed mesh to stick onto the hi poly reference mesh.
B > hi light 20 lights, scroll down > B > hi light more lights > scroll down > over and over a few times to get all the lights.
Is there a faster way than this?
- Parent them to an empty, curve or whatever. You can then select Children/ChildrenRecorsive (whole children tree)
- Use the default Layers system (There is an Addon that is shipped with Blender that allows Layer Naming and such)
- Use the Super Grouper Addon
there is specifically for the outliner this small addon;
http://blenderaddonlist.blogspot.de/2013/11/addon-outliner-extend-keys.html
which makes your hoped function working like you wished.
left click lamp1, shift double-left click lamp50 , the lamps lamp1-lamp50 are now selected.
del pls
>>>DOWNLOAD LINK<<<
(The "face option" is still not working with multiple faces selected)
If you select a face and set it to Face, it points the normal of the nearby face to that direction (If I've understood correctly the function, it should do the same as selecting the face edges and setting them to hard), then if you want to correct the mess you have done, the Smooth and Flat will produce crazy results as this:
Here looks like you have marked all the edges as hard-edges, than you assign the "face" option to all the faces, but this is meanless, once you setted the hard edges the vertex normals works correctly by default. the face option is intend to be used to help to fake bevels as I did in my example before.
If you want to set hard edges with this addon you should play with the "split normals" feature.
My suggestion is to use only this addon or only the Blender tools to edit the vertex normals, not both together in the same time.
cheers!
If I deactive the Normal Editor, the editing is lost.
yes the addon overrides the object normals, it works more or less like a modifier.
On a cube, activate the Normal Editor, select a face, and click on one of the 3 options (it doesn't matter), you don't like the result, and click on Restore. Now your face is Smooth, even though your cube was on Flat.
If you then click on Flat, to put it back as its previous state, nothing happens, same with Smooth, but if you select all and click on one of these options, the weird stuff I showed in the previous post happens.
The only way to reverse this mess is to select all the faces, click on Tree (you may need to do this twice, disabling and enabling the Normal Editor), then click on Smooth, and finally on Flat. In this order.
myclay Thanks!!
I actually have this addon enabled in my setup, but haven't used this function. Awesome
Also, this addon by SynaGl0w allows to setup rigid bodies physics that will result in a seamless texture (or that's what I understand how it works at least :P)
http://blenderartists.org/forum/showthread.php?377518-Addon-Piles-N-Tiles-Seamless-Tools-for-Blender
Yeah you basically set up your rigid bodies how you want them, then add them to the seamless area where each rigid body gets surrounded by passive copies of itself. The passive copies are offset the dimensions of the seamless area and mirror the active with constraints.
Basically if an active rigid body goes too far away from the others, it interacts with passive copies instead of other actives, while the other actives interact with its passive copies.
There are some examples of what it can do over here (though most were made for tests as the tools developed): http://blenderartists.org/forum/showthread.php?357526-Seamless-Physics-Meshes-and-Textures/page3
The size of the seamless area can be changed at any time, even after its full of objects (although you will have to rerun the simulation). The tools also allow an orthographic camera to be automatically created and scaled with the seamless area as well. Also this is not solely for rigid bodies, it can be used on almost any object. It also comes with node groups for cycles to ease material creation for seamless objects.
There is documentation included with the addon that I hope explains it well enough.
@Edit
One of the biggest problem is that if I overlaping two the same UV-islands (mirrored), I have big problem with normal map in game engine. Even If I import this to 3ds max it seems that Blender merged that. I need to move mirroed part to second UV block and then it works good.
Currently the polycount limitation is quite low. There's some workaround by enabling VBO and disabling double sided but to sump it up, no blender can't handle a lot more.
But the developement is currently focused on that, Antony did some great job recently and it will go further : [ame]https://www.youtube.com/watch?v=0hKkJtybris[/ame]
can you actually edit such a high density mesh though without having the app drop to a snail's pace? or is it just some caching to not have the viewport lag at every step? the latter can be found in at least a certain autodesk 3d app and is only partially useful. <bicker bicker bicker>
when you enter editmode you are doomed
Was my thought as well. "Yes, but can you edit them in real time?"
what good is cake if you can't eat it?
no. just be able to edit complex geometry and have dev resources focused on speeding up the actual editing tools. you never know when you need the extra performance. in comparison being able to rotate stupid amounts of polygons in the viewport with simple materials thrown on is less useful in general.
there seems that underlying and necessary code re factoring has begun to
be able to edit such massive amounts of data...(129M triangles) albeit not showing yet in pictures due to its early state.
I guess it depends on the workflow - I personally rely quite a bit on having dense (not edit-mode friendly, but at least sculpt-friendly) models in my Blender scenes, so any boost in that regard is welcome ! That being said, 2.74 already behaves quite well when dealing with decimated models from Zbrush, so it is quite workable even now.
Can't wait to see where all this is going. Exciting times !