I also have a script which toggles the visibility of all my modifiers, so I suppose I could merge these case-specific actions to it If anyone thinks this is a bad idea for any particular reason though, do let me know. I am all for working smarter!
To you and anyone who rely heavily on Blender's modifiers, I highly recommend using "Modifier List" addon. It adds a whole new level of organization to Blender's modifiers, and makes it much easier to add modifier controllers (such as empties, lattice objects and vertex groups) automatically.
import bpy
for obj in bpy.context.selected_objects:
obj.data.uv_layers.new(name='lightmap')
Thank you for the script! Im still trying to find a quickier way to bulk edit different objects and packing their UV channels for lightmap automation without merging them... I came up with this updated version but since I suck at python its not getting me results.
import bpy
for obj in bpy.context.selected_objects:
obj.data.uv_layers.new(name='lightmap')
meshes = {obj.data for obj in bpy.context.selected_objects if obj.type == 'MESH'}
for mesh in meshes:
mesh.uv_layers['lightmap'].active = True
bpy.ops.uv.smart_project(island_margin=0.05)
mesh.uv_layers['UVMap'].active = True
bpy.ops.uv.unwrap(method='ANGLE_BASED', margin=0.001)
Anyway, here's a little series Im putting up sharing the process of creating a city inspired by the anime Evangelion.
@LuisCherubini Thoroughly enjoyed watching you model those buildings, thanks for sharing. What was the light setup you were using in that eevee scene? Resembled like an sunrise/afternoon-light hitting the city, kinda
I came up with this updated version [...] its not getting me results.
import bpy
for obj in bpy.context.selected_objects:
obj.data.uv_layers.new(name='lightmap')
meshes = {obj.data for obj in bpy.context.selected_objects if obj.type == 'MESH'}
for mesh in meshes:
mesh.uv_layers['lightmap'].active = True
bpy.ops.uv.smart_project(island_margin=0.05)
mesh.uv_layers['UVMap'].active = True
bpy.ops.uv.unwrap(method='ANGLE_BASED', margin=0.001)
If you want to use blender operators, you need to do it in the same way as if you use the GUI. You need to make sure your active object is set accordinly for each iteration of the loop. Sometimes selection as well.
So, iterate over objects, not meshes, then make them active, then use the ops.
A slightly smarter way that doesn't require to explicitely set the active, is by using a context override instead.
@MACHIN3 Could you (or anyone reading this) elaborate on that? There are a ton of scripts I use that I could apply this knowledge to. I understand I need to iterate through selected objects to be active so that I can execute whatever commands I want to, but is the code you shared doing that?
The for at the beginning is going through all the objects you selected, and the for mesh in meshes is doing it again...but in mesh edit mode? I don't get it.
I tried placing some commands in both fors, namely these ones:
...But the only resulting objects that were correctly modified was the active one in my selection, whereas I would want/expect it to be applied to all of them.
@Justo there is a distinction between the selected object and the active object. While the active object is both selected and active. The selected objects are just that, selected.
i believe luis is just seperating the objects which don't have the type mesh. Like lamps etc. Since those are not needed and can cause errors.
Most operators in blender operate only on the active selection just like you were to select it with the mouse in the viewport. And since there can be only one active object at a time you need to iterate the active selection.
Here is a simple script I made to iterate through all selected objects, make them active and apply low poly smoothing
import bpy
for ob in bpy.context.selected_objects:
bpy.context.view_layer.objects.active = ob
bpy.ops.object.shade_smooth()
bpy.context.object.data.use_auto_smooth = True
bpy.context.object.data.auto_smooth_angle = 3.14159
Amazing! I was missing that that view_layer.objects.active = ob line, thank you m00k!
Also, about setting code in here in the forums, it can be confusing sometimes how to do so without the page destroying your indentations haha. What I do is go to a new line, type crap like ojifdg, highlight that and make it into the code format, and then replace that with my actual code: this way the code spaces are preserved.
@LuisCherubini Thoroughly enjoyed watching you model those buildings, thanks for sharing. What was the light setup you were using in that eevee scene? Resembled like an sunrise/afternoon-light hitting the city, kinda
Its the second HDR on the look-dev mode, resembles London I think. A fill light to give the base sky (blueish), and a inclined sun lamp with warm temperature.
Is there any reason that these directional lines are so long on lights? Is there
any way to change the length? On some scenes - depending on object scale
- these lines shoot off screen and I have to zoom out really far just
to see the ends of them.
Edit: seems that scaling them on Z is the only hacky workaround I can find. Very annoying default.
@xrg Not that one, this one when you open blender.
Oh, heh. It's been so long since I used Windows I forgot it even did that. They really should figure out a way to get rid of it entirely.
This post says it's in Window -> Toggle System Console. It may have moved in 2.8, though. You also might need to save the startup file (File->Defaults->Save Startup File) to make it permanent.
Would anyone have a fix for having your other blender windows be on top as well ? (and so brought back when you switch back to your regular Blender window) bugs me !
There are so many Blender threads being created nowadays in the Technical Talk forums, I love it...
Can someone enlighten me as for smart uses of median center vs bounding box center when using pivots? I understand their difference being the following, taken from here:
-The Median center is calculated as the average of all selected points ( vertices or object centers ) so it depends on the density distribution of the mesh -The bounding box : is the smallest box (Cube ) that can contain the mesh of an object while aligned to the local space -The bounding Box center will always be at the center of this bounding box regardless of the density distribution of the mesh inside
...That's all good, but I was wondering what specific cases would the median center be needed for. Like, for getting certain shapes or operations. Because I've never really had a use for it and feel like I can do without it. I just wondered if anyone heavily relied on it for certain ops, so I could learn something new.
@SnowInChina maybe this example illustrates it better (any irregular shape will do):
The Median center is the average of all points, so it will prioritize clusters of points because they pull the average towards themselves. It's at the point of most density.
The Bounding-Box center prioritizes the minimum/maximum of all points, so it's at center of the volume of the shape.
actually i wanted to show, that if you are working with circular objects, you should always work with median point, if they are geometrical circles
working with the bounding box does not guarantee you that you are actually working with the exact center, it depends on the vertex count of the circle as well as the rotation since the bounding box is calculated on the furthest away vertex and not on a mathematical circle, which is an issue for rotating parts (wobbling)
i assumed that was clear, next time i will specify
I hadn't notice that, that's a nice tip. If the shape is supposed to be "even" in all directions (a radial shape, like a circle), then manipulating it from the median center will probably be better than the bounding box.
With irregular shapes (maybe something organic), depending on what you want to do, it's best to test each center and see where they are first.
The Bounding-Box center prioritizes the minimum/maximum of all points, so it's at center of the volume of the shape.
that's not correct, its at the center of the bounding box (y1+y1)/2; (x1+x2)/2; (z1+z2)/2, not the shape, its not linked to the volume, and doesn't take into account all vertices, only the ones that have the lowest and highest x,y&z value
you can test it yourself, if you create a random shape and move vertices around without moving them past your furthest points in x,y&z, your center point will stay the same, best way to visualize that would be to take your object and build a cardboard box around it, as close as possible.
at some point you will touch your object and cannot move the sides any closer, that is your bounding box and from there you take the exact middle of all planes and shoot a line at a 90° angle inwards, where they meet is your center point
personally i almost never use the bounding box
edit: another way to think about is, when you visualize your model in a 3d space with the x y & z axis
the bounding box looks for the minimum and maximum points on each axis, this point is always a vertex, since you cannot have curved edges or faces in polymodeling.
so when you rotate your model all points get shifted and your bounding box changes (as far as i know the bounding box is local, so rotation in object mode actually rotates the bounding box with it, rotating in edit mode does not rotate the bounding box and your center point will shift on every model, except for some geometrical forms)
sorry to stress this, but i've seen a lot of models where they messed this up and it needed to be fixed later on
Here's something that should be simple but isn't :
Starting from a set of curves (either of the "2D" or "3D" kind in Blender terms), how does one generate a surface from that while still keeping the curves independantly editable, and with the surface updating itself either in real-time or at the very least at press of a button so that edits can be done easily ? Basically, how does one perform a lathe/loft ?
(of course it would be possible to convert these curves to mesh, then merge, then bridge/fill, but that's besides the point here unless that can be done in one click ...)
Also : is there a way to apply all modifiers of all currently selected objects, in one click ?
No worries - I use both myself as they have different strengths and weaknesses. On the topic of curve patching : looks like this is actually not possible at all, as all topics online refer to converting to mesh and adding faces from there - which I am of course aware of but not what I am looking for.
Still curious about collapsing modifiers on multiple objects at the same time.
I have the problem where I can't do anything in the viewport. The outliner does not react except the eye-icons work. Keymap in the preferences is an empty screen. Hovering over one of the buttons on the right screen gives "Internal Error" Toolttips The remaining interface seems to work normally but I haven't thoroughly checked.
Testing with Rc2 atm.
Anyone had similar stuff happening?
Hello me, so your problem is that your python install(ations) are bit strange and blender 2.8 is using the wrong one. The reason the debug batch files (like 'blender_debug_gpu_glitchworkaround.cmd') is because they clear the windows 'PYTHONPATH=' variable. So I made a minimum batch file which does just that and allows to be called with parameters which will get forwarded to blender
@echo off
echo Starting blender with PYTHONPATH=''
set PYTHONPATH=
%~dp0\blender %*
EDIT: I have to say it's a bit weird for me right now in the current 2.8 release. I don't get the menu to adjust spans after creating a surface from curves.
I have to say Blender is really great and 2.8 did the amazing thing that it just clicked - for me at least. So I thought I do a small project from modelling, unwrapping, rigging, skinning, scripting. You know, the usual stuff I do with max. The aim is to improve my overall speed with the tool.
But - of course - it wouldn't be Blender without some headaches..
But there's one thing I can't stand and I couldn't find a workaround yet: You can't see the weighting on the vertices as you can in max. This makes it impossible to weight in wireframe mode and it's impossible to see how the weights on backfaces or otherwise obstructed parts behave.
And as I want to get into scripting too I just picked this as a reason to start building my first tool:
But there's one thing I can't stand and I couldn't find a workaround yet: You can't see the weighting on the vertices as you can in max. This makes it impossible to weight in wireframe mode and it's impossible to see how the weights on backfaces or otherwise obstructed parts behave.
And as I want to get into scripting too I just picked this as a reason to start building my first tool:
Take That blender nonsense (to be continued)
Vertex weights should show up on the "n"-panel item tab, it's hidden if you don't select a vertex and it only show the last selected vertex, dosn't work for box select etc. It's quite weird or I just don't understand it correctly, but I would expect to see averaged values and being able to make changes for the whole selection not just the last one.
Yes the default weight shading in wireless view is next to useless. Vertex or edge colouring would be very useful.
Seems someone was trying to develop weight shaded wireframe view back in '14. It almost got committed but not quite. You can look at the blender developer article here:
Seems someone was trying to develop weight shaded wireframe view back in '14. It almost got committed but not quite. You can look at the blender developer article here:
This looks nice. So how does this blender development work. Do I have to vote for this feature? Also it seems to be not thought through yet. The wires are colored, which is nice but the verts need it more desperately. However, this is done in the C core of Blender? This would obviously be nicer but so long I'll use the python workaround
You prey that Blender devs find it necessary enough and has a desire to maintain the code. The Right Click Select (https://blender.community/c/rightclickselect/) suggestion board supposedly is the place to give some weight to our needs in Blender, but I'm not sure if core Blender devs go there at all. You can also try to check https://devtalk.blender.org/
I also highly recommend you talking with the rigging community from Blender in twitter. They seem to be pretty friendly and help each-other quite a lot. @Muream, @TheLapisSea, @JeannotLandry and @chadvernon are some I can remember from the top of my head.
Vertex weights should show up on the "n"-panel item tab, it's hidden if you don't select a vertex and it only show the last selected vertex, dosn't work for box select etc. It's quite weird or I just don't understand it correctly, but I would expect to see averaged values and being able to make changes for the whole selection not just the last one.
Yes the default weight shading in wireless view is next to useless. Vertex or edge colouring would be very useful.
Sorry I totally missed your replay. Ok wait, ... I'll have to prepare:
Wooooooooooot??? If I box-select a single vertex it doesn't show up but if I single-click??? Thx blender that you are still there to astonish me...
Anyway, thx for helping me out!
------------------- Edit: does someone know how to access the default weight ColorRamp? I know I can access bpy.context.preferences.view.weight_color_range but this doesn't return the red-green-blue ramp if bpy.context.preferences.view.use_weight_color_range is disabled
Wooooooooooot??? If I box-select a single vertex it doesn't show up but if I single-click??? Thx blender that you are still there to astonish me...
I know it will probably not make you feel any better, but the reason for this seems to be that: box selecting stuff "only" selects stuff while clicking directly on a single thing makes that thing immediately the *active* selection
Wooooooooooot??? If I box-select a single vertex it doesn't show up but if I single-click??? Thx blender that you are still there to astonish me...
I know it will probably not make you feel any better, but the reason for this seems to be that: box selecting stuff "only" selects stuff while clicking directly on a single thing makes that thing immediately the *active* selection
I had to laugh so hard when I saw this today:
Blender is a tool allowing artists to create content, and not a coders API! That's quite obvious, but it provides emphasis on functionality for things that first-most *work* and improve the creation process. Of course Blender should be very flexible to configure and extend to serve many workflow issues, for that high-end and accessible (scripting) APIs have to be available. (https://wiki.blender.org/wiki/Reference/UIParadigms)
Who ever wrote that must had a good time while doing so
@rollin Blender is certainly not without its flaws. There is an increasing monthly budget for hiring developers though, and slowly but surely the wishes of the community are transformed into features via sites like https://blender.community/c/rightclickselect/
If you ever feel like missing a feature in Blender, please submit it and hopefully it will be implemented sooner or later.
box select a single vertex just works as expected.. it just does not set the active element - which makes sense for every selection with more than one element.. but yeah might be an idea to handle the single element case differently - its not like they are not up for ideas
In my previous post I just wanted to point out that the vertex weight values in the panel are just for the active element and not all selected element. Box selection was not the point, just an example.
I know this might not be what most people want to hear but the best way to get a feature implemented in blender.... is to develop it yourself! Often times a simple python script will do the trick. Learning to code will pay dividends even in other programs.
Blender is open-source after all! Which is both it's strength and maybe it's weakness
@SnowInChina@RN Thank you for your input on median points vs bounding boxes! I never really got the difference until now. I suppose I'll try working always with Median Point rather than BB unless I need the latter for something specific.
Something I came across today: about reference images in Blender. As I understand it, the quickest way to bring an image to Blender is by dragging images from your Explorer to the viewport, which will automatically create the images. This is super convenient and all, but...is there no way to modify these created images, inside Blender? Say for example, if I wanted to divide it in two and cut it in half, to move around the drawings separately or put them in different views. These "Image" object-types dont seem to be able to be edited like meshes, after all.
... if I wanted to divide it in two and cut it in half, to move around the drawings separately or put them in different views. These "Image" object-types dont seem to be able to be edited like meshes, after all.
For that you should use the "Image as planes" option
Coming from Max and being completely new to blender, are there any settings that will make life easier for me? I just started with a basic 2.8 tutorial and it bothers me that selecting transformation tools using the default hotkeys involves so many button presses. I'm a bit hesitant to go changing a lot of stuff without more experience but I was hoping that maybe there's a proven path to transitioning.
Coming from Max and being completely new to blender, are there any settings that will make life easier for me? I just started with a basic 2.8 tutorial and it bothers me that selecting transformation tools using the default hotkeys involves so many button presses. I'm a bit hesitant to go changing a lot of stuff without more experience but I was hoping that maybe there's a proven path to transitioning.
Use the industry standard key map. I'm from max too and the only thing you might should drop is max "non-standard" way of camera navigation. Maya won
(even though you can setup blender to behave like max I wouldn't suggest bc then it gets into the way of selecting and other default hotkeys)
Replies
import bpy for obj in bpy.context.selected_objects: obj.data.uv_layers.new(name='lightmap') meshes = {obj.data for obj in bpy.context.selected_objects if obj.type == 'MESH'} for mesh in meshes: mesh.uv_layers['lightmap'].active = True bpy.ops.uv.smart_project(island_margin=0.05) mesh.uv_layers['UVMap'].active = True bpy.ops.uv.unwrap(method='ANGLE_BASED', margin=0.001)Anyway, here's a little series Im putting up sharing the process of creating a city inspired by the anime Evangelion.
The for at the beginning is going through all the objects you selected, and the for mesh in meshes is doing it again...but in mesh edit mode? I don't get it.
I tried placing some commands in both fors, namely these ones:
...But the only resulting objects that were correctly modified was the active one in my selection, whereas I would want/expect it to be applied to all of them.
i believe luis is just seperating the objects which don't have the type mesh. Like lamps etc. Since those are not needed and can cause errors.
Most operators in blender operate only on the active selection just like you were to select it with the mouse in the viewport. And since there can be only one active object at a time you need to iterate the active selection.
Here is a simple script I made to iterate through all selected objects, make them active and apply low poly smoothing
Also, about setting code in here in the forums, it can be confusing sometimes how to do so without the page destroying your indentations haha. What I do is go to a new line, type crap like ojifdg, highlight that and make it into the code format, and then replace that with my actual code: this way the code spaces are preserved.
Its the second HDR on the look-dev mode, resembles London I think. A fill light to give the base sky (blueish), and a inclined sun lamp with warm temperature.
Can someone enlighten me as for smart uses of median center vs bounding box center when using pivots? I understand their difference being the following, taken from here:
-The bounding box : is the smallest box (Cube ) that can contain the mesh of an object while aligned to the local space
-The bounding Box center will always be at the center of this bounding box regardless of the density distribution of the mesh inside
...That's all good, but I was wondering what specific cases would the median center be needed for. Like, for getting certain shapes or operations. Because I've never really had a use for it and feel like I can do without it. I just wondered if anyone heavily relied on it for certain ops, so I could learn something new.
bounding box
Starting from a set of curves (either of the "2D" or "3D" kind in Blender terms), how does one generate a surface from that while still keeping the curves independantly editable, and with the surface updating itself either in real-time or at the very least at press of a button so that edits can be done easily ? Basically, how does one perform a lathe/loft ?
(of course it would be possible to convert these curves to mesh, then merge, then bridge/fill, but that's besides the point here unless that can be done in one click ...)
Also : is there a way to apply all modifiers of all currently selected objects, in one click ?
On the topic of curve patching : looks like this is actually not possible at all, as all topics online refer to converting to mesh and adding faces from there - which I am of course aware of but not what I am looking for.
Still curious about collapsing modifiers on multiple objects at the same time.
so your problem is that your python install(ations) are bit strange and blender 2.8 is using the wrong one.
The reason the debug batch files (like 'blender_debug_gpu_glitchworkaround.cmd') is because they clear the windows 'PYTHONPATH=' variable.
So I made a minimum batch file which does just that and allows to be called with parameters which will get forwarded to blender
It says from curve or whatever but you can use it on a mesh with modifiers just as well. It will apply all your modifiers.
Also Curves2mesh sure has some great potential. Thanks !
So I thought I do a small project from modelling, unwrapping, rigging, skinning, scripting. You know, the usual stuff I do with max.
The aim is to improve my overall speed with the tool.
But - of course - it wouldn't be Blender without some headaches..
Weight painting for example! Imo this sucks in vanilla Blender atm.
(Found your thread pior
I found this addon though: https://gumroad.com/l/weight_paint_tools
Looks quite nice for the usual symmetry stuff
Do you guys know some more?
Also I seem to be unable to find the Vertex Weights panel: Has anyone seen it in 2.8?
But there's one thing I can't stand and I couldn't find a workaround yet:
You can't see the weighting on the vertices as you can in max. This makes it impossible to weight in wireframe mode and it's impossible to see how the weights on backfaces or otherwise obstructed parts behave.
And as I want to get into scripting too I just picked this as a reason to start building my first tool:
It's quite weird or I just don't understand it correctly, but I would expect to see averaged values and being able to make changes for the whole selection not just the last one.
Yes the default weight shading in wireless view is next to useless. Vertex or edge colouring would be very useful.
https://developer.blender.org/T39054
Also it seems to be not thought through yet. The wires are colored, which is nice but the verts need it more desperately.
However, this is done in the C core of Blender? This would obviously be nicer but so long I'll use the python workaround
Ok wait, ... I'll have to prepare:
Wooooooooooot??? If I box-select a single vertex it doesn't show up but if I single-click??? Thx blender that you are still there to astonish me...
Anyway, thx for helping me out!
-------------------
Edit: does someone know how to access the default weight ColorRamp?
I know I can access bpy.context.preferences.view.weight_color_range but this doesn't return the red-green-blue ramp if bpy.context.preferences.view.use_weight_color_range is disabled
Who ever wrote that must had a good time while doing so
Blender is open-source after all! Which is both it's strength and maybe it's weakness
I have 'maxed' blender a bit
try working always with Median Point rather than BB unless I need the latter for something specific.
Something I came across today: about reference images in Blender. As I understand it, the quickest way to bring an image to Blender is by dragging images from your Explorer to the viewport, which will automatically create the images. This is super convenient and all, but...is there no way to modify these created images, inside Blender? Say for example, if I wanted to divide it in two and cut it in half, to move around the drawings separately or put them in different views. These "Image" object-types dont seem to be able to be edited like meshes, after all.
(even though you can setup blender to behave like max I wouldn't suggest bc then it gets into the way of selecting and other default hotkeys)