I would strongly recommend you to register and post your questions on the BlenderArtists forum, as these two things are totally doable, which makes me think that maybe some of your other concerns might be solved already too.
QWER certainly works for me (this was one of the first things I customized, right after the Maya-style navigation) ; and even though I didn't quite understand how to work with them at first, smoothing groups/hard edges are definitely fonctional - I use them all the time myself.
Now I certainly wish there was a way to "doublesmooth" with them like in Max, but that's a very Max-specific trick that no other 3d app has anyways, so I cannot quite blame Blender for not having it ... yet !
a doublesmoothish workflow is there, i'll ask one of our guys tomorrow, he recently did this.
I got a decent result by removing all lights but the sky, setting the sky to a bright white color, applying a subsurface material to my mesh, changing the subsurface parameters, baking the subsurface direct pass to a 32-bit float buffer, saving that out as exr 32-bit, applying a diffuse material to my mesh, baking the diffuse direct pass to a 32-bit float buffer, and using the compositor to find the difference between the two bakes. But it would be a lot faster and probably more technically correct to just use Xnormal to bake it, and you'd be able to have antialiasing too, and much better batching support.
Yeah, but regular edge creasing totally defeats the purpose of working with edge hardness-driven subdivision ...
I agree that It doesn't seem like much of a difference in principle, but in practice it is night and day. Double smoothing based on smoothing groups is about 100 times more popular than creasing in Max for a reason
Yeah, but regular edge creasing totally defeats the purpose of working with edge hardness-driven subdivision ...
I agree that It doesn't seem like much of a difference in principle, but in practice it is night and day. Double smoothing based on smoothing groups is about 100 times more popular than creasing in Max for a reason
i'd agree, but recently i started doing it more with creases for various reasons.
First of all some clients want chamfered meshes, not doublesmoothed geometry, to have a base they can easily manipulate. While those meshes are not ideal for sculpting, for hardsurface this is great. So quadchamfer im combination with creases opens a whole new world of flexibility.
That said, creases in Max are horrible for various reasons, it's cumbersome to set them up by hand and they are not previewed in the edges. Which i think they are in Blender.
So what i usually do for doublesmooth is select faces i want a nice sharp edge around and assign a smoothinggroup via script.
But from there on creases should be able to do the exact same. Instead of assigning smoothing group the script should
assign SGS or Hardedges to get the visuals
select border
assign generic crease value to border
and another script which will:
apply smooth modifier
apply second smooth modifier
ideally instanced
this would make the workflow exactly the same as doublesmooth in max to but, but with the flexibility of making certain edges sharper or softer. And in blender it would visualise the edges for you, sadly not in max. but oh well.
Yeah, assigning creasing based on hard edges through scripting would probably be a very trivial thing to do for anyone familiar with Python. I will probably post a design doc about it on BlenderArtists in the future, as some kind soul would probably be more than willing to do it. As a matter of fact, a simple script automatically selecting all the hard edges of a model would be a great first step.
As for the viability of double smoothing in production, I agree - this is certainly not a one-size fits all solution, since at the end of the day it is not that finely controllable. It sure saved my butt a few times though !
Chiniara, to see lowpoly hard edges in the viewport you need :
- "Shading" to be set to "Smooth" in the left panel ;
- "Normals / Autosmooth" to be turned on and set to 180 in the data tab of the object (this is the part that I initially didn't understand, but it actually makes sense)
- and then in edit mode, perform "Edges - Mark Sharp".
Blender already have the doublesmooth workflow nailed down. All you have to do is crease the edges and then add 2 subdivision surface modifiers in your stack. Here's a quick example with a cube for you:
I was very surprised when someone told me that the second subsurf modifier ignores the creasing, but that is how it works. This behavior is not very widely known, but nothing really stops anyone from using the doublesmooth workflow in Blender.
- "Normals / Autosmooth" to be turned on and set to 180 in the data tab of the object (this is the part that I initially didn't understand, but it actually makes sense)
I don't get the purpose of the Autosmooth option to be honest. I never used that.
I just set the mesh to smooth shading in the T panel and mark the edges hard that should be hard and finally add a edge split modifier. That always worked for me.
Still it bugs me not to understand what that auto smooth thing is useful for, since some people here mentioned it.
Blender already have the doublesmooth workflow nailed down. All you have to do is crease the edges and then add 2 subdivision surface modifiers in your stack. Here's a quick example with a cube for you:
I was very surprised when someone told me that the second subsurf modifier ignores the creasing, but that is how it works. This behavior is not very widely known, but nothing really stops anyone from using the doublesmooth workflow in Blender.
yeah but the thing is, in many cases setting up SGs is just way faster than selecting all border edges by hand, but all this could be done in script. In Blender and Max. it could be even case sensitive, if you select faces it will do it on the border edges if you select edges it will apply a standard value to those edges. making it more flexible than doublesmooth in max is for me right now.
I don't get the purpose of the Autosmooth option to be honest. I never used that.
I just set the mesh to smooth shading in the T panel and mark the edges hard that should be hard and finally add a edge split modifier. That always worked for me.
Still it bugs me not to understand what that auto smooth thing is useful for, since some people here mentioned it.
Autosmooth smooths your mesh depending on the angles between faces. If you make a cube and set autosmooth to maybe 80 or so it will make the edges hard automatically for it. If you add a sphere to the same object it will look just as smooth as you'd expect it to.
The only thing that's really confusing about the whole autosmooth thing is why it's not turned on by default, especially when it's got such a high default value that it won't break anyones model. I asked for it to be turned on by default a while back, but I don't think anyone showed any interest.
yeah but the thing is, in many cases setting up SGs is just way faster than selecting all border edges by hand, but all this could be done in script. In Blender and Max
Scripting would speed it up a little bit further but you could just use the 'Select Similar' feature while in edge mode, select already assigned hard edges or select them by face angles and from there apply your creasing. It sounds kind of convoluted but once you do it a few times it's pretty quick, a script would however make it even quicker.
I just set the mesh to smooth shading in the T panel and mark the edges hard that should be hard and finally add a edge split modifier. That always worked for me.
Still it bugs me not to understand what that auto smooth thing is useful for, since some people here mentioned it.
From what I understand, the "edge split" modifier workflow effectively ... splits the edges, which is not desirable in a majority of cases.
As for autosmooth, it can be a great presentation tool when set to medium values like 30/40. As a matter of fact in some cases it can even be used as a visual help when modeling, as it will basically indicate if a shape is strong enough or not. It is a bit of an exotic tip, but helpful nonetheless.
Nossiak : yeah, I am aware of that of course. But as said earlier, in practice things are different. Imagine starting from the model shown in the screenshot I linked above, and having to mark all these edges manually - it would take forever, and similarly, marking them as "creased" while modeling wouldn't be an option either since there is no way to preview that without actually subdividing the model. As Neox said, a script going back and forth between hard edges and creased edges would help.
[edit] Frankie : looks like "select similar - sharpness" does the trick ! It is a bit convoluted but it does work so no need for a script then (aside from the subtleties mentioned by Neox). I am still hoping that one day, the subdivision modifier will have the option of simply taking hard edges into account but that's probably just a matter of time anyways, and implementing Opensubdiv is understandably a much higher priority right now.
Interesting to hear about the presentation and visual use of the auto smooth option. I might play with that in the future.
As for the edge split modifier. The edges get only split if you apply that modifier. Otherwise it's non-destructible like all of Blender's modifiers. And in a game engine, the verts get split anyways where the hard edges appear. Or am I wrong?
But I'm always open to hear about better workflows. That's why I'm reading here
Yeah, but having to keep a modifier on a model just for visualization purposes is a bit of a hindrance - for instance you would have to think about it each time you want to merge parts, split models, and so on ... The way I see things is : "the less brain noise, the better". As a matter of fact, after trying the "regular" hard edges method outlined above I am convinced that you will feel the same too
I also know for a fact that at least in Maya (and maybe Max too ?), merging parts with crease information on them causes issues (it has been addressed to an extend in recent releases of Maya, but iirc there are still some lingering problems) whereas the bold on/off nature of hard edges seems to be better handled and never "dropped". I am totally willing to admit that this is part of the reason why I tend to be wary of edge creasing in general - once you loose valuable work (even just once) because of an unstable tool, it is hard to go back to it and trust it again, even in a different program.
I personally see hard edges not so much as an "end step" (in which case, any trick, even convoluted, would be valid) but more as a useful thing to play with and rely on during modeling, especially under tight time constraints. For instance the model shown above was to be used as a blockout/preview so slapping hard edges on it was a way for me to know that things are "good enough for now", and I can move on to the next piece. And of course, being able to instantly turn this visual information into sharp edges on a subdivided model is super useful for later sculpting.
Yeah, but having to keep a modifier on a model just for visualization purposes is a bit of a hindrance - for instance you would have to think about it each time you want to merge parts, split models, and so on ... The way I see things is : "the less brain noise, the better". As a matter of fact, after trying the "regular" hard edges method outlined above I am convinced that you will feel the same too
I also know for a fact that at least in Maya (and maybe Max too ?), merging parts with crease information on them causes issues (it has been addressed to an extend in recent releases of Maya, but iirc there are still some lingering problems) whereas the bold on/off nature of hard edges seems to be better handled and never "dropped". I am totally willing to admit that this is part of the reason why I tend to be wary of edge creasing in general - once you loose valuable work (even just once) because of an unstable tool, it is hard to go back to it and trust it again, even in a different program.
I personally see hard edges not so much as an "end step" (in which case, any trick, even convoluted, would be valid) but more as a useful thing to play with and rely on during modeling, especially under tight time constraints. For instance the model shown above was to be used as a blockout/preview so slapping hard edges on it was a way for me to know that things are "good enough for now", and I can move on to the next piece. And of course, being able to instantly turn this visual information into sharp edges on a subdivided model is super useful for later sculpting.
But then again, everybody works differently !
Thanks for the explanation. I find that really interesting how people tackle that in different ways. I'll give your method definitely a spin next time. I'm always open to different approaches.
The current system depends on the scene depth buffer to generate information for the effects. As such, any drawing that writes to the depth buffer, such as widgets and wires will contribute in the effects.
The SSAO effect relies on the surface normals to function, however since the current system is not true deferred shading, those get composited from the depth buffer and smooth shading is not taken into account. Also there will be artifacts at mesh edges due to the composited nature of the normals.
Multisampling will not work in this first iteration - it requires too much memory to do properly.
From what I gather, the viewport internals are still work in progress, improving as they go.
also dof should dof the whole thing not just within the silhouette, this way it's rather useless?
The dof is not blurring the BG because I cut it out in photoshop xD, lol.
the dof is associated to the active camera, that means that it works seamless between viewport and render. the focuse basically is made using an empty (null) object.
For the AA, is not in the viewport yet, but you can force by your videocard and it works if you make a "playblast".
They're working on figuring out a way to avoid destroying animation takes that people don't want destroyed, which is really good news in my opinion because remembering to click that F for every take to force it to be saved is really a pain and I forget to do it too often.
I hope that whatever the change is will also apply to materials.
Thanks, but this clears only your selection, I was looking for a way to have some faces selected and while working on vertices/edges, have those faces still there when I switch back to face mode.
SonicBlue : I am confused, this is exactly how it works by default here ...
Would you mind describing your issue further ? We might be talking about different things.
SonicBlue : I am confused, this is exactly how it works by default here ...
Would you mind describing your issue further ? We might be talking about different things.
Sure, let's say you are selecting some faces, then you see that they are not straight, switching to vertex mode to set your 3D cursor and then to edge mode to select the affected edges that you want to straight. At this point, when you came back to faces mode, you lost your selection, the first one you made.
Same thing to make them planar, I know that you can first set the 3D cursor, and then make the selection, but well, if you didn't notice this first it's a problem to select them back again.
The workaround would be to set a selection (Selection Tag in Cinema 4D or Cluster in XSI), but I couldn't find a way to do that too.
Haaa I see what you mean now. Maybe there is an option for it somewhere ?
The thing is, many users would rather have the other behavior (which is default at the moment) which is to maintain the current selection across the different selection modes. However if you clearly document what you want, with a video and/or mockups, the devs could very well implement it as an option, or, someone one the BlenderArtists forums could maybe script it for you.
Normally i would say "vertex group" to keep selection...but it doesn't work well with Face selection (if you have a checker of face selected for example).
You can use however various tricks like assign an other material to the selected face then use the Shift-G : selected by materials. Or split the selected face with Y then L to reselect.
That's not the best, the better way would be to have a Face Group inside Blender...I saw it once but i don't know what happens to him
Remembering what you have selected while moving to other modes is not that essential, it would have been better if Blender had the usual functions you expect from a 3D software, usually to convert your selection to another there is a specific command, and by default you can select what you want in the different modes.
I'll try to make a video, even though it's already clear what the "problem" is
Yeah man totally, showing things clearly with a mockup of what you want is usually the best thing.
Just be careful to not label it as a "problem" or something that Blender does "wrong", because even thought there *are* some bugs and issues in Blender, this isn't one of them - this is just you asking for an interesting feature. Presenting it as such is usually the best way to get traction on getting something implemented.
Of course, I put that into quotes because it's a problem I have with the software, not a Blender problem itself, my workflow is based on that, so I'd expect to behave in a way that is not intrinsic to it.
I never see it, and I'm not saying its ok to be broken but in that particular situation a better workflow might be to make and select the end vert then select the start vert and press j to "connect vertex path". Or use ctrl R for loop cut then connect the part it wont do with j
Yeah, I did that part with Loop Cut tool, I just triggered the Knife bug since I was recording another video for a post on Blenderartist (that is still waiting approval) and took the chance to film it, it usually happens while cutting many consecutive edges. A possible workaround is to cut few piece per time, but it's easy to forgot, I tend to cut a lot with the knife in Cinema 4D.
Thanks for the suggestion of the Connect Vertex Path ;D
There are other ways to cut edges? The CTRL+R way is not optimal if you are cutting long polygons and the model is not perfectly framed in the view.
(actually I think it changed to be connect vertex path in the latest test build and in the current release its just called connect)
Also if you ring select edges(so all the ones you want to cut next to each other like | | | | | are selected) you can do "subdivide" and it auto joins in the center maybe that helps? Although it's a bit awkward to make the selection efficiently and I can't convert "pick shortest path" from face mode to the edge ring selection needed.
It works, it works, I was just curious about other tools that I may not know.
I also know the Edge ring split since it was used a lot by Max users, even though I never used this technique since Cinema 4D knife is quite capable, it can easily split a polygon at a defined percentage.
What I find strange is that I can't cut where I pointed the mouse, the Loop Cut always starts at 50% and you have to move it afterwards, instead of cutting directly where you want and eventually move the cut to another position.
Hey guys, sorry for not reading much in this thread, but it's freakishly huge to trawl through. I already googled most of these things and haven't found very satisfactory answers.
So I'm learning Blender, I like it a lot more than maya and it's way easier to get into than Modo.
But:
I don't like how picking the Maya Preset also changes all the shortcuts, making it almost impossible to google certain shortcuts since everybody always names the default ones. Can't I just have Maya-like mouse interaction and navigation, but all the rest default?
I'm pretty bummed out about only having one single autosave file; for some reason I deleted a whole bunch of objects and transformed another bunch completely wrong, yet the autosave only has like one version from 10 minutes ago which is no help at all. In Max you could have many, allowing you to go back to a point before stuff was broken. Seems like there's no way to do this?
Can I please for the love of god turn off that stupid 'confirm to delete' dialog?
Some tool issues:
Knife tool does not snap to vertices on other objects and tends to generally be hard to use (just touching an edge without clicking = cut) compared to Max's Cut Tool. Any tips?
I really miss Shift-Move to extrude edges, like Max. Why is it that only Max does this properly? Doing and extrude and then moving or snapping is just plain slower.
Is there a proper Target Weld?
Can the slide loop tool be used a bit more interactive? In max it previews under your cursor and you click to commit it. It's kind of annoying to have it require 2 clicks in Blender.
How do I apply properties to multiple objects at once? The properties panel only edits like the last selected one, which is not very useful. Do I need to use separate tools/actions for this every time, like 'move to layer' ?
Xoliul, for your first question, I think someone posted something about that a couple pages back (maybe pior?). I believe it was a maya navigation but not hotkey setup.
I don't like how picking the Maya Preset also changes all the shortcuts, making it almost impossible to google certain shortcuts since everybody always names the default ones. Can't I just have Maya-like mouse interaction and navigation, but all the rest default?
I'm pretty bummed out about only having one single autosave file; for some reason I deleted a whole bunch of objects and transformed another bunch completely wrong, yet the autosave only has like one version from 10 minutes ago which is no help at all. In Max you could have many, allowing you to go back to a point before stuff was broken. Seems like there's no way to do this?
Can I please for the love of god turn off that stupid 'confirm to delete' dialog?
Some tool issues:
Knife tool does not snap to vertices on other objects and tends to generally be hard to use (just touching an edge without clicking = cut) compared to Max's Cut Tool. Any tips?
I really miss Shift-Move to extrude edges, like Max. Why is it that only Max does this properly? Doing and extrude and then moving or snapping is just plain slower.
Is there a proper Target Weld?
Can the slide loop tool be used a bit more interactive? In max it previews under your cursor and you click to commit it. It's kind of annoying to have it require 2 clicks in Blender.
How do I apply properties to multiple objects at once? The properties panel only edits like the last selected one, which is not very useful. Do I need to use separate tools/actions for this every time, like 'move to layer' ?
This has apparently been discussed a while ago https://developer.blender.org/T37422 but no changes have been made as far as I know, so currently you can't disable the prompts without making changes to the code (someone correct me here)
Knife tool does not snap to other objects, but "just touching an edge = cut" seems weird as I can't replicate that kinda behaviour, can you elaborate on that?
You can do that. Press E to extrude and MMB to free the extrusion axis (if it's locked to the normal), then use Ctrl to activate snapping if it's not already active.
Alt + M aka Merge is the shortcut you want.
The first Loop cut mode is for you to select how many loops you want, with the Mouse Wheel. Then you click and slide the loop/s into position. EDIT: I misread what you wanted to do.
You want Ctrl + L aka Make Links. This works with your active selection (Select the object that you want the properties copied from last, that's your active selection) and links the selections properties to all the other selected objects.
* Do Save Versions cover the auto-save requirement for you? Not quite the same thing.
* Confirm to delete is dumb. Press D immediately after X is the closest thing I can think of (chaining hotkeys)
* Selecting a component and then control-clicking is Blender's weird equivalent of Max's shift extrude.
* Target-welding is done by enabling AutoMerge Editing (under the Mesh menu) and using vertex snap. Vertices become welded automatically. Alternatively, you can chain A > ctrl-V > D to weld all overlapping vertices together.
For applying properties to multiple objects you can also right click some properties and pick "apply to selected."
There's no "proper" target weld but you can use Alt+M to merge vertices or if you need to weld lots you can activate vertex snapping and use right click and drag tweaking to move vertices onto each other, then use Remove Doubles (I usually find this in the spacebar search) to weld them.
Idk about the autosaves but Blender automatically saves .blend1, .blend2 and .blend3 files when you save manually, giving you more stuff to go back to. Just rename the file to "xxx.blend" and you'll be good to go.
You shouldn't be sliding the loop with a tool, you should be sliding it with a hotkey. In this case the hotkey is "gg".
For quick extrudes use Ctrl+Click. Then tweak the verts into place with right mouse click and drag, and left click to confirm (or right click to cancel if you change your mind).
Make sure you enable Loop Tools and F2 addons. They can really speed up the modeling process. F2 lets you create a face that naturally extends a surface from a selection of one or two verts, and Loop Tools has some handy things like loop spacing and relaxing.
Replies
a doublesmoothish workflow is there, i'll ask one of our guys tomorrow, he recently did this.
I got a decent result by removing all lights but the sky, setting the sky to a bright white color, applying a subsurface material to my mesh, changing the subsurface parameters, baking the subsurface direct pass to a 32-bit float buffer, saving that out as exr 32-bit, applying a diffuse material to my mesh, baking the diffuse direct pass to a 32-bit float buffer, and using the compositor to find the difference between the two bakes. But it would be a lot faster and probably more technically correct to just use Xnormal to bake it, and you'd be able to have antialiasing too, and much better batching support.
I agree that It doesn't seem like much of a difference in principle, but in practice it is night and day. Double smoothing based on smoothing groups is about 100 times more popular than creasing in Max for a reason
i'd agree, but recently i started doing it more with creases for various reasons.
First of all some clients want chamfered meshes, not doublesmoothed geometry, to have a base they can easily manipulate. While those meshes are not ideal for sculpting, for hardsurface this is great. So quadchamfer im combination with creases opens a whole new world of flexibility.
That said, creases in Max are horrible for various reasons, it's cumbersome to set them up by hand and they are not previewed in the edges. Which i think they are in Blender.
So what i usually do for doublesmooth is select faces i want a nice sharp edge around and assign a smoothinggroup via script.
But from there on creases should be able to do the exact same. Instead of assigning smoothing group the script should
assign SGS or Hardedges to get the visuals
select border
assign generic crease value to border
and another script which will:
apply smooth modifier
apply second smooth modifier
ideally instanced
this would make the workflow exactly the same as doublesmooth in max to but, but with the flexibility of making certain edges sharper or softer. And in blender it would visualise the edges for you, sadly not in max. but oh well.
As for the viability of double smoothing in production, I agree - this is certainly not a one-size fits all solution, since at the end of the day it is not that finely controllable. It sure saved my butt a few times though !
Chiniara, to see lowpoly hard edges in the viewport you need :
- "Shading" to be set to "Smooth" in the left panel ;
- "Normals / Autosmooth" to be turned on and set to 180 in the data tab of the object (this is the part that I initially didn't understand, but it actually makes sense)
- and then in edit mode, perform "Edges - Mark Sharp".
I hope this helps !
I was very surprised when someone told me that the second subsurf modifier ignores the creasing, but that is how it works. This behavior is not very widely known, but nothing really stops anyone from using the doublesmooth workflow in Blender.
I don't get the purpose of the Autosmooth option to be honest. I never used that.
I just set the mesh to smooth shading in the T panel and mark the edges hard that should be hard and finally add a edge split modifier. That always worked for me.
Still it bugs me not to understand what that auto smooth thing is useful for, since some people here mentioned it.
yeah but the thing is, in many cases setting up SGs is just way faster than selecting all border edges by hand, but all this could be done in script. In Blender and Max. it could be even case sensitive, if you select faces it will do it on the border edges if you select edges it will apply a standard value to those edges. making it more flexible than doublesmooth in max is for me right now.
The only thing that's really confusing about the whole autosmooth thing is why it's not turned on by default, especially when it's got such a high default value that it won't break anyones model. I asked for it to be turned on by default a while back, but I don't think anyone showed any interest.
Scripting would speed it up a little bit further but you could just use the 'Select Similar' feature while in edge mode, select already assigned hard edges or select them by face angles and from there apply your creasing. It sounds kind of convoluted but once you do it a few times it's pretty quick, a script would however make it even quicker.
From what I understand, the "edge split" modifier workflow effectively ... splits the edges, which is not desirable in a majority of cases.
As for autosmooth, it can be a great presentation tool when set to medium values like 30/40. As a matter of fact in some cases it can even be used as a visual help when modeling, as it will basically indicate if a shape is strong enough or not. It is a bit of an exotic tip, but helpful nonetheless.
Nossiak : yeah, I am aware of that of course. But as said earlier, in practice things are different. Imagine starting from the model shown in the screenshot I linked above, and having to mark all these edges manually - it would take forever, and similarly, marking them as "creased" while modeling wouldn't be an option either since there is no way to preview that without actually subdividing the model. As Neox said, a script going back and forth between hard edges and creased edges would help.
[edit] Frankie : looks like "select similar - sharpness" does the trick ! It is a bit convoluted but it does work so no need for a script then (aside from the subtleties mentioned by Neox). I am still hoping that one day, the subdivision modifier will have the option of simply taking hard edges into account but that's probably just a matter of time anyways, and implementing Opensubdiv is understandably a much higher priority right now.
As for the edge split modifier. The edges get only split if you apply that modifier. Otherwise it's non-destructible like all of Blender's modifiers. And in a game engine, the verts get split anyways where the hard edges appear. Or am I wrong?
But I'm always open to hear about better workflows. That's why I'm reading here
I also know for a fact that at least in Maya (and maybe Max too ?), merging parts with crease information on them causes issues (it has been addressed to an extend in recent releases of Maya, but iirc there are still some lingering problems) whereas the bold on/off nature of hard edges seems to be better handled and never "dropped". I am totally willing to admit that this is part of the reason why I tend to be wary of edge creasing in general - once you loose valuable work (even just once) because of an unstable tool, it is hard to go back to it and trust it again, even in a different program.
I personally see hard edges not so much as an "end step" (in which case, any trick, even convoluted, would be valid) but more as a useful thing to play with and rely on during modeling, especially under tight time constraints. For instance the model shown above was to be used as a blockout/preview so slapping hard edges on it was a way for me to know that things are "good enough for now", and I can move on to the next piece. And of course, being able to instantly turn this visual information into sharp edges on a subdivided model is super useful for later sculpting.
But then again, everybody works differently !
Thanks for the explanation. I find that really interesting how people tackle that in different ways. I'll give your method definitely a spin next time. I'm always open to different approaches.
You find it only in the recent builds.
Wow that's some sexy stuff.
also dof should dof the whole thing not just within the silhouette, this way it's rather useless?
http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.74/UI
From what I gather, the viewport internals are still work in progress, improving as they go.
The dof is not blurring the BG because I cut it out in photoshop xD, lol.
the dof is associated to the active camera, that means that it works seamless between viewport and render. the focuse basically is made using an empty (null) object.
For the AA, is not in the viewport yet, but you can force by your videocard and it works if you make a "playblast".
enjoy.
I haven't dabbled with animating yet, so I don't understand any of this. What's the big deal?
I hope that whatever the change is will also apply to materials.
Thanks, but this clears only your selection, I was looking for a way to have some faces selected and while working on vertices/edges, have those faces still there when I switch back to face mode.
Would you mind describing your issue further ? We might be talking about different things.
Sure, let's say you are selecting some faces, then you see that they are not straight, switching to vertex mode to set your 3D cursor and then to edge mode to select the affected edges that you want to straight. At this point, when you came back to faces mode, you lost your selection, the first one you made.
Same thing to make them planar, I know that you can first set the 3D cursor, and then make the selection, but well, if you didn't notice this first it's a problem to select them back again.
The workaround would be to set a selection (Selection Tag in Cinema 4D or Cluster in XSI), but I couldn't find a way to do that too.
The thing is, many users would rather have the other behavior (which is default at the moment) which is to maintain the current selection across the different selection modes. However if you clearly document what you want, with a video and/or mockups, the devs could very well implement it as an option, or, someone one the BlenderArtists forums could maybe script it for you.
You can use however various tricks like assign an other material to the selected face then use the Shift-G : selected by materials. Or split the selected face with Y then L to reselect.
That's not the best, the better way would be to have a Face Group inside Blender...I saw it once but i don't know what happens to him
I'll try to make a video, even though it's already clear what the "problem" is
Just be careful to not label it as a "problem" or something that Blender does "wrong", because even thought there *are* some bugs and issues in Blender, this isn't one of them - this is just you asking for an interesting feature. Presenting it as such is usually the best way to get traction on getting something implemented.
Release notes.
Also I appreciate the developer humor when talking about bug fixes:
"As for every Blender release, no bugs were fixed. Creating new ones is much more fun!"
[ame]
this happens in 2.73 ad 2.74
Thanks for the suggestion of the Connect Vertex Path ;D
There are other ways to cut edges? The CTRL+R way is not optimal if you are cutting long polygons and the model is not perfectly framed in the view.
https://dl.dropboxusercontent.com/u/1693140/connectvertexedges.gif
(actually I think it changed to be connect vertex path in the latest test build and in the current release its just called connect)
Also if you ring select edges(so all the ones you want to cut next to each other like | | | | | are selected) you can do "subdivide" and it auto joins in the center maybe that helps? Although it's a bit awkward to make the selection efficiently and I can't convert "pick shortest path" from face mode to the edge ring selection needed.
I also know the Edge ring split since it was used a lot by Max users, even though I never used this technique since Cinema 4D knife is quite capable, it can easily split a polygon at a defined percentage.
What I find strange is that I can't cut where I pointed the mouse, the Loop Cut always starts at 50% and you have to move it afterwards, instead of cutting directly where you want and eventually move the cut to another position.
So I'm learning Blender, I like it a lot more than maya and it's way easier to get into than Modo.
But:
EDIT: I misread what you wanted to do.
Fast reply is fast, hope it helps.
* Confirm to delete is dumb. Press D immediately after X is the closest thing I can think of (chaining hotkeys)
* Selecting a component and then control-clicking is Blender's weird equivalent of Max's shift extrude.
* Target-welding is done by enabling AutoMerge Editing (under the Mesh menu) and using vertex snap. Vertices become welded automatically. Alternatively, you can chain A > ctrl-V > D to weld all overlapping vertices together.
There's no "proper" target weld but you can use Alt+M to merge vertices or if you need to weld lots you can activate vertex snapping and use right click and drag tweaking to move vertices onto each other, then use Remove Doubles (I usually find this in the spacebar search) to weld them.
Idk about the autosaves but Blender automatically saves .blend1, .blend2 and .blend3 files when you save manually, giving you more stuff to go back to. Just rename the file to "xxx.blend" and you'll be good to go.
You shouldn't be sliding the loop with a tool, you should be sliding it with a hotkey. In this case the hotkey is "gg".
For quick extrudes use Ctrl+Click. Then tweak the verts into place with right mouse click and drag, and left click to confirm (or right click to cancel if you change your mind).
Make sure you enable Loop Tools and F2 addons. They can really speed up the modeling process. F2 lets you create a face that naturally extends a surface from a selection of one or two verts, and Loop Tools has some handy things like loop spacing and relaxing.