Hey guys, I came across support loops recently and how some people use them and some dont. Someties you need them, mainly before subdividing mesh etc. But this left me with questions, why do you use them or dont, what are the cases ?
support loops simply define the hardness of the edge they are supporting (well, they are also used to have a consistent loop around edges, so it looks nice and has no smoothing artefacts)
sometimes you don't need them because there`s already geometry in place that acts as a support edge, or you want the edge to be super soft
some poeple also work with edgecrease in blender, which is basicly the same and replaces supportloops, but has some drawbacks
support loops simply define the hardness of the edge they are supporting (well, they are also used to have a consistent loop around edges, so it looks nice and has no smoothing artefacts)
sometimes you don't need them because there`s already geometry in place that acts as a support edge, or you want the edge to be super soft
some poeple also work with edgecrease in blender, which is basicly the same and replaces supportloops, but has some drawbacks
you need an edge to be super soft - I guess you speak about a ase in which you are further going to smooth mesh either by subdividing or working with it. But I also saw using support loops on a geometry that wont be touched - why ? In that case I find it useless to have it there,if topology is ok.
"Support loops" is terminology, so it might mean different things to different people.
In the context of the Subdivision Surface modifier -- in Catmull-Clark mode, engineered by Pixar -- the position of an original vertex (one of the vertices of the mesh before it is sent to the modifier) becomes a weighted average of the positions of its surrounding vertices.
This means that the closer together the vertices are, the more their output averaged positions are going to look like their input positions. That is, the original vertices of the mesh move less in areas where they are closer together.
Anyone know why this shader fails to compile? Line 16 has a "<=" that appears to have been incorrectly formatted as a "<" in the forum's text box, but once that's been sorted out I just get reports of a syntax error at an unspecified location.
Today a Monday material question: Node graphs are nice for custom or complex materials. But if I just want to setup the good old basic inputs (diffuse, normal, roughness, happiness) adding the same nodes over and over again is unnecessary time consuming. I don't want to waste time deciding where the nodes are places and if the uv data is routed correctly through. What's the go-to workflow for this?
edit: I guess what I like to do is use an existing material-node-graph and create instances from it and just change the linked textures
You could enable the materials library VX addon that comes with Blender. You can also insert your own libaries and materials here. That way you can store your often used setup
I'm lazy and want to keep add-ons to a minimum, so if it were me, I would create a node group in a blank material, then save the scene to a .blend file so it can be linked / appended as a template, where you import that blank material and have instant access to that preset node group from any material in your scene.
@RN I think you can actually link/append node groups without the material it is used in. But I would have to check again and I can't do that right now.
@f1r3w4rr10r that's cool, can confirm it, you import the group itself as a node tree:
Importing the material seems to import all node groups used by it, so you can use the material as a container of all your preset groups, as well as being able to import individual groups.
Is there a way -through some code maybe?- to fix the extra windows not being on top when the main one is?
Thanks
So far I'm plugging them on the side of the main window and extending it. But when you restart Blender it's back to being on one monitor only, so I gotta reposition it again every time.
Can anyone give me a quick rundown on what Blender does faster than max modeling wise? One thing I'm interested in is the fact you can work without the gizmo. It seems slightly faster than having to snipe axis clicks on the max gizmo. In max you can't rotate/scale without the gizmo tool as well. Really all I care about is how streamlined the tools are with modeling and if there is any clear advantage over max.
@tynew most of it pretty much comes down to what you already said. Hotkeys make a lot of things faster and the fact, that you do not have to care where in your viewport your mouse is, when you activate a transform that way.
Add to that, that most other modeling operations are also just a hotkey away, and I would say you already have a solid groundwork for being faster.
In my experience, a lot of it comes down to learning the hotkeys and using them.
@tynew personally I really like how Blender breaks down the Pivot types, origins and orientation. I don't recall if Max can replicate all of them, and if it does, I think it's hidden away in a more convoluted UI. There are a lot of cool things to play with, but you will probably find Blender's spline modeling tools lacking compared to Max', in case you used those a lot (I didn't).
Can anyone tell me how Diffuse BSDF and Gloss BSDF shader nodes relate to the values one tweaks in a Principled BSDF? Diffuse and Gloss have roughness values that are handled differently, for one thing. Is the amount of glossiness instead handled by the "Specular" principled value?
@f1r3w4rr10r Is there anything else for you that comes to mind? 99% of my modeling is already done in max with hotkey smart function scripts and stuff like keyhydra. I almost never touch the command panel or menu items for modeling already. Even have my tab and caps lock keys rebound to 50 or so functions. I'm looking in terms of unique features, such as working without a gizmo.
@Justo Interesting, could you elaborate more on the pivots/orientation? Specifically a use case where you can't achieve it in max?
@tynew some max users complain about the bevel/chamfer in blender for not being as ''smart'' as in 3ds max, you should test a bit the bevel modifier workflow and see how it is in comparison with the smoothing groups and quad chamfer.
@tynew. I have started out my 3d modeling with Max, but that was a really long time ago. So my memory of it is a bit foggy. What @Justo meant is that it is easy in Blender to change you pivot point. You can choose among different presets like object origin, individual object origins, median point or go fully custom with the 3d cursor. Since 2.8 this one also supports orientation, not just position.
@tynew : working without a gizmo is probably the biggest (and positive) paradigm shift indeed. There are a couple more things too :
• Off-hand transformation (there is no official name for that, that's just what i call it). The idea is that on top of not having to grab any onscreen gizmo, there is also no need to grab onto the active object or component(s), as your mouse cursor can be anywhere in the viewport at the time of, say, move or rotate being activated. The transformation is done according to the on-screen relationship between object and mouse cursor (with a little dashed line visualizing the link), also with shift key acting as .2 multiplier (I think ?) reducing the effect. It is somewhat similar in concept to the Zbrush transpose line but much, much better in its implementation. It becomes second nature very quickly and once you go there all other 3D programs and level editors start feeling like outdated antiques
• Pure modeling won't be much different from Max as the tools/features are similar. Skin weighting and morph target handling is an order of magnitude more powerful than Max though, as the data is stored very cleanly (vertex groups) and doesn't get destroyed when editing a model/splitting/merging parts. Huge time saver.
• The 3D cursor, not only as a way to set custom transformation centers on the fly but also as an aid to viewport navigation - by having it snap to surfaces and then pressing a key to focus to cursor. Navigating inside nostrils has never been easier ! But more seriously this is great and completely eliminates 3D viewport motion sickness.
• Overall (a little less true in 2.8 than 2.78/2.79, but still valid compared to Max) : the UI is way more responsive than Max/Maya (or at least the last few versions I've tried). Buttons don't have any micro stutter, UI elements are rock solid and never glitch in/out, panel resizing is butter smooth, and so on. There is just less "crap built over old crap" and it makes for a very responsive and robust experience.
• Updates/installation management is a breeze. No need for any installer, user preferences are all stored in a folder, and so on. Meaning that testing out new features will never feel like a drag - it's a matter of minutes.
• Now I would say that the biggest downside is scene undo, which is objectively broken (extremely slow) at this time.
• Also, while the new viewport is extremely powerful, overall all viewports (even raw shading) tend to refresh mush slower than previous versions when working with animated skeletal meshes - which imho makes 2.8 not suitable at all for animators at this time.
My unique real gripe with Blender 2.8 / 2.81 is the poor perfomance with subdivision modelling.
I miss some modifiers from Max or Maya, but i think it's a matter of changuing the modelling philosophy. I remember when i started using modo 1.0, i missed all from Max. Right now i do things in a different way.
I'm heavy customed to press space in order to drop tools in modo, and "alt" + Ctrl or Shift to control Viewports. In modo i can do what i call poly-sculpting, i don't use a mouse with modo. I mean, using the sculpting tools, i don't need to use any gizmo or anything, just some key shortcuts and that's all. with Blender i'm still having issues to work as fast as in modo + Cintiq.
Do you guys use Blender with a cintiq/Intuos for modelling? I do, and i'm having a hard time
@Udjani I definitely will try it out. Currently the chamfer mod in max 2020.2 is seemingly the best out of any DCC app right now. Although I don't mind moving to blender for its other features.
@pior Fantastic thank you for the write-up. Handling transformation like that definitely feels more fluid for a 3D app. Anything relating to sniping clicks/elements and forcing UI popups/spinners can really break the flow when you work. For example using the mouse wheel in blender to dynamically add chamfer segments or loops. This sort of stuff should really be native in max by now.
Yeah I saw your cool post on copying half a head and still having it animate! Very cool.
Hopefully, the viewport performance improves, I generally deal with 4+ mil items with hard surface work. I'm definitely excited to learn blender considering how well it's supported and has a very active community. Plus having a single program that can do a majority of tasks reasonably well, eg sculpting and animating makes it a very convincing option.
If I wanted free rigged models to start a sculpt in Blender, what options are there? I was looking for something that looked simple and clean, since the sculpt I was planning to do was a very minimalistic, stylized kid.
Anyone know why this shader fails to compile? Line 16 has a "<=" that appears to have been incorrectly formatted as a "<"
@Zablorg < is the HTML entity for "<" (less-than), no equal. So replacing that piece with less-than and refreshing the script in the Script node makes it run (screen from 2.79):
@Justo unless you're pressed for time, I personally feel that modeling
and rigging your own base meshes is way too educational to skip on it.
You should really make your own, if you can.
Besides topology reference images, you can also have reference models such as the "build-a-bod" models (which are Blender-friendly rips of Daz3D Genesis models, anatomically complete and accurate), and those from the fork of that Manuel Bastioni Lab thing.
But you use these just as reference: you'd make your own from scratch with box-modeling or sculpting-retopo.
Today a Monday material question: Node graphs are nice for custom or complex materials. But if I just want to setup the good old basic inputs (diffuse, normal, roughness, happiness) adding the same nodes over and over again is unnecessary time consuming. I don't want to waste time deciding where the nodes are places and if the uv data is routed correctly through. What's the go-to workflow for this?
edit: I guess what I like to do is use an existing material-node-graph and create instances from it and just change the linked textures
If you're using nodes at all, look into 'Node Wrangler', an official, included-by-default addon. It adds a lot, but most importantly for you: a hotkey (shift-ctrl-T) and menu-item that allow one-click setup of PBR materials.
Hey @rollin ! Well, I can certainly confirm that this code here totally works. Excellent, thanks !
Now I just need to dig a bit deeper into what it takes for PME to call full on python scripts, as I find that manually creating UI panels is just too cumbersome. Totally doable mind you, but just not fluidly enough imho - as everytime I find myself doing so to neatly store useful little presets and shortcuts I always end up reverting to PME, for its ease of use and speed. Looks like it does need to be an external .py for it to call it though, so I suppose it has to be formatted like any other addon. But hey at the very least it will save me the hassle of creating a panel
You'll be glad to hear that external scripts in PME don't need all of the addon boilerplate stuff, they just need the lines that do the actual work.
Going through the Blender Conference videos now (they're all up on the Youtube account!), and I was wondering if anyone could explain how this (and the next example) is done?
I understand that the DataTransfer modifier is probably involved, but I was hoping for some more specific instructions to help me out
My unique real gripe with Blender 2.8 / 2.81 is the poor perfomance with subdivision modelling.
I miss some modifiers from Max or Maya, but i think it's a matter of changuing the modelling philosophy. I remember when i started using modo 1.0, i missed all from Max. Right now i do things in a different way.
I'm heavy customed to press space in order to drop tools in modo, and "alt" + Ctrl or Shift to control Viewports. In modo i can do what i call poly-sculpting, i don't use a mouse with modo. I mean, using the sculpting tools, i don't need to use any gizmo or anything, just some key shortcuts and that's all. with Blender i'm still having issues to work as fast as in modo + Cintiq.
Do you guys use Blender with a cintiq/Intuos for modelling? I do, and i'm having a hard time
In what way are you having a hard time? I'm too coming from Modo, and have to say I like Blender 2.8 more and more. I'm not as fast there as in Modo yet (been working with Modo over 10 years, professional modeler), just digging into Blender 2.8 for a couple of weeks now. But getting faster by the day with practice, and I must say that I really like the fresh air. There is surprisingly little functionality in Blender that I actually miss compared to what I do in Modo most of the time. One of those things certainly is the workplane, that strange 3D cursor in Blender is not an adequate replacement. If they implement a workplane into Blender, I might actually consider switching completely.
Also the ability to place any menu entry as a button into the UI in Modo, some functionality in Blender requires too much searching in menus, or the use of a shortcut. The Blender UI does not feel flexible enough, or not flexible enough for the standard user. You can probably do a lot with scripting ect, but thats not something I'm into.
I agree that working with a pen tablet kind of feels suboptimal in Blender in a couple of places, a couple of things/ addons are geared towards using a mouse + mouse wheel as input device. I'm using an old Intuos4 pen tablet. Trying to model with a mouse feels like trying to sculpt delicate details with a brick to me.
I find it very difficult to work different meshes UVs at the same time. You got em all selected, say you want to work on your Second channel, only the last selected object will have its second channel edited, all the other selected objects will end up broken / editing UV channel 1.
@Blaizer : this all really comes down to setting up your inputs to replicate what you would like. I mean ... it's perfectly normal to not like any navigation/interaction scheme different from one's preferred one, so in that sense your question/approach is a bit backwards : of course it will feel odd an inefficient if you don't set it up the way you want it in the first place
Now practically speaking, my advice for that would be to first not worry at all about any of this and just using the app for about a week or so with everything default just for the sake of digging into the features, also messing around with the various settings and playing around with editing the keymap manually without worrying about breaking stuff (and you likely will). Then, starting fresh again (by wiping out the preferences folder in Roaming) and taking as long as you need to carefully set up viewport and interaction controls to be exactly the way you want them. And then keep on exploring from there.
Here's an example on how to set this up manually (done with an older Blender version and replicating the Maya-style inputs but the concept still applies)
As for tablet use : well, here too I would say that this is a bit of a strange question. Of course it can be used with a tablet - just the same way every other program can be I personally constantly switch between mouse and stylus for both 2D and 3D and it works just as well in Blender as in any other app. So I would assume that here too this is just a matter of tweaking the inputs to what you'd like them to be - but only *after* spending some time with the defaults, because you will likely be very pleasantly surprised with some of the very clever interaction paradigms that Blender offers (gizmo free editing, off-hand manipulator line, 3D cursor, and so on).
@wilson66 I'm very customed since year 2006 (or earlier, dunno), to work only with a tablet or digital pen (Cintiq right now), and only with the sculpting tools of modo (since they appeared). I don't use a mouse for work, nor gizmos, just the pen tablet + a few shortcuts i created (i almost model without UI, just a custom viewport with flatness 120%).
My workflow is similar to working in Zbrush + "Zmodeler next gen" --> Modo sculpt + better polygonal tools (what i call "polysculpt modelling").
I wish i could select faces, edges and points in Blender Sculpt mode, edit the cage mesh/subdivision model and so on. For me, using blender right now it's like going back in time, decades, to my old days of Max and Maya. My whole workflow is totally broken using Blender, as in Max or Maya.
@Pior I ended changing the keymap and navigation to something like Maya, like i did with 2.79 but worse, needing to change more keys than expected because it ends broking a lot. Blender 2.8 changed all, it's like a slap on the face. https://www.youtube.com/watch?v=Pb4UqxQormA
I found that swithching to "alt", i was broking other key bindings involving the "alt" key such as "alt + middle mouse + mouse move" (to easily rotate to top and side views without touching the wheel gizmo).
I really need to see what i can, and what i can't change. I restored all to default, and i'm with the Blender Manual studiying all the new features, improvements and changes. But i'm seeing that i won't use it for what i'm customed to do very fast, like modelling and uv mapping. Anyways, i like a lot Eevee and its render times.
Well, if anything the fact that you are comfortable with Maya-style navigation is already good info. I was under the assumption that you were attempting to recreate some kind of default Modo paradigm with a lot of things specific to that - but Maya-style is indeed 100% possible. That's what I've been using with of course a number of personal edits on top and it really is great.
As mentioned a few times I also personally rely very heavily on the 3D cursor for navigation, to precisely set the camera orbit center exactly where I want it. This is imho a great example of of Blender-only feature that a user hoping to replicate the input paradigm of another app wouldn't think of (I certainly didn't) yet it yields great benefits. Having perfect control on the way the camera behaves has a tremendous impact on flow/staying in the zone while working, and that seems to be exactly what you are are looking for.
So yeah my recommendation would still be to keep playing around without fear of breaking things and with no expectations of results for a little longer - and then starting fresh again after about a month or so of indepth exploring. All the grievances you mention are imho non-issues once everything is setup to your likings.
A practical example : many users seem to rely on the so-called "Specials" menu (previously W and now right click, if I remember correctly). That's something I personally never had any need for since there are other/faster ways to access these tools, so I had to take the time to disable the new 2.8 right click menu in all contexts to make way for what I use right click for (instant cursor placement on surface). It sure took a bit of time and could be seen as annoying, but it is also very easy to do thanks to the way the keymap editing screens gives access to everything.
TLDR : I can confirm that the fluid workflow you describe is 100% possible
Do you guys use Blender with a cintiq/Intuos for modelling? I do, and i'm having a hard time
Its been fine on my end, though I use the official Industry Standard keymap option, which is basically a maya+modo like keymap. Mix it with pie menus and it works great!
The pie menus are good, but i still can't work like in modo. The middle mouse/wheel key for navigation makes impossible to use a cintiq in edit mode, just in sculpt mode. And it's totally different compared to modo, whereas the spacebar is "drop tool". How do i edit polygons in sculpt mode? the help Blender help is useless
I want to work like i shown years ago in this video. ~30 min timelapse modelling, using the sculpt tools + modelling operations with key shortcuts.
@Blaizer I'm not sure if you can rebind the mouse stuff for Blender yet, but you could probably use a workaround. I'm pretty lazy so I just used a program called x-mouse button control and you can swap your mouse keys per program. Gave me a lot of relief changing the pan/rotate to rmb from mmb in max.
The pie menus are good, but i still can't work like in modo. The middle mouse/wheel key for navigation makes impossible to use a cintiq in edit mode, just in sculpt mode. And it's totally different compared to modo, whereas the spacebar is "drop tool". How do i edit polygons in sculpt mode? the help Blender help is useless
I want to work like i shown years ago in this video. ~30 min timelapse modelling, using the sculpt tools + modelling operations with key shortcuts.
I came from Modo as well. I am not sure what issue you are having exactly. Middle mouse works fine for me on a pen. I use the industry standard keymap (which you have to select from the properties or splash screen), and then Alt + pen tip for left click, alt + side pen button for middle and right clicks. No issues at all with navigation.
It is true, that once you get used to dropping tools in Modo, it becomes hard to adjust to other applications that don't use that approach. In Blender you can edit tool or active tool properties but once you "drop" the tool by clicking onto something else not related to it then you cannot edit its parameters. In the lower left you will find a tool properties associated with the active tool. GUI wise its not ideal but its more of an annoyance than it is a game breaker.
As for sculpt mode and edit mode, they are two different modes. I often flip between them by using the 1-3 key and the 5 key depending on component or sculpt. You can also toggle proportional editing to move components around without the sculpt tool. However yes, the two modes are completely separated, where as in Modo everything is technically consolidated.
As for sculpt mode and edit mode, they are two different modes. I often flip between them by using the 1-3 key and the 5 key depending on component or sculpt. You can also toggle proportional editing to move components around without the sculpt tool. However yes, the two modes are completely separated, where as in Modo everything is technically consolidated.
That's my issue. Sculpt mode and Edit mode are totally different, like separated tools, and i can't use the sculpt tools with edit tools or viceversa. I can't do a simple bevel or extrude while sculpting (inflating polygons, pinching, etc.). In modo, i can do such simple things since luxology added sculpting features (back in 2006). I don't need to flip between modes in Modo, i have all things in a single viewport in full-screen mode. I don't like to go back to year 2005, with a maya or max mouse only workflow (edit mode).
If you watch closely the video i posted, i'm sure you'll understand me better. I just use the sculpt brush for all, and a few keymappings. I had to record that video to explain the polysculpting modelling workflow.
Well, that's precisely why I believe it is pointless (and dare I say counter-productive) to look at it that way. If the tools, workflows and default hotkeys were exactly the same as Modo ... it would be called Modo.
Instead if you are curious about the app just dive in casually without expectations and explore/understand the systems fully for about 2 weeks to a month. A practical example : you seem to be getting stuck with the assumption that since Blender has two distinct modes for editing and sculpting, then that means that you cannot replicate your hybrid modeling/sculpting workflow (which imho is nothing unique really, and I don't mean it in a snarky way : I personally cannot understand how anyone could work on organic models any other way as this is objectively the best and most efficient approach).
At the surface level this is of course true : Edit mode and Sculpt mode are separate. But in practice switching modes is probably just as fast as selecting the sculpt brush in Modo ; and switching mode + selecting the grab brush can be done in one sweep with a macro anyways.
Of course it won't be behaving exactly like what you are used to, and even performance might differ - but that's an obvious as these are two different pieces of software.
So yeah I honestly think you are looking at it from the wrong end, especially since you can very well use both apps in tandem for their respective strengths. If anything, you might not find exactly the workflow you want ... but you might be able to precisely document the things that you feel are missing, and these missing things could end up being implemented eventually based on your very own feedback. Think of it as an investment in the future evolution of the software.
As for things like working in full screen mode with a stylus : as said, this is 100% possible since everything can be mapped to anything you want. The buttons/hotkeys might not be there by default but it is perfectly possible to assign hotkeys to anything you want using the keymap editor, or diving further by creating menus and macros for anything you want using PME (Pie Menu Editor, which is not just for Pie Menus but also for custom hotkeys/macros).
On a different topic : does anyone know of any addon allowing to replicate the basic BoxCutter workflow (cutting in boolean shapes using rectangles/circles/polylines in screen space) but without all of the overhead of the recent versions BoxCutter ?
The 2.8+ iterations of BC are as powerful as ever but the addon is becoming seriously bloated compared to what the tool once was (it has a UI button to "activate it" but no button to switch it off ; and it now has dozens of options and modes). It is reaching a point where I have to constantly disable/enable the add-on for it not to get in the way and that defeats the whole purpose. I've tried a few other ones but I would welcome any suggestion. Thanks !
In the default keymap the W key can "drop tool." I don't know how easy/difficult it would be to switch it to spacebar, since I rarely edit keybindings. I also don't know if the Industry Compatible keymap has anything similar.
The 2.8+ iterations of BC are as powerful as ever but the addon is becoming seriously bloated compared to what the tool once was (it has a UI button to "activate it" but no button to switch it off ; and it now has dozens of options and modes). It is reaching a point where I have to constantly disable/enable the add-on for it not to get in the way and that defeats the whole purpose. I've tried a few other ones but I would welcome any suggestion.
I feel you on that. I find surface snapping behavior enabled by default to be less consistent than simple screen-space orientation for cutting...
Well, that's precisely why I believe it is pointless (and dare I say counter-productive) to look at it that way. If the tools, workflows and default hotkeys were exactly the same as Modo ... it would be called Modo.
Pior, Pixologic added to Zbrush the Zmodeler "brush", "subdivision surface modelling in Zbrush", and Zbrush it's not called "Modo" precisely. They saved us a lot of time adding features we already have in Modo or Max.
Once you are customed to work with a tablet pencil, using a mouse is like something really primitive, not evolved and not efficient for me (for god's sake, we are in the end of 2019!). I tried "Alt" key combinations for viewport navigation because: 1 - it's like the standard (Maya) and 2 - it's a good choice using a tablet or Cintiq. Not because we use the "alt" key in Modo.
Believe me, when i discovered the possibilities of polysculpting modelling, i said to myself: i won't go back. Going back to the old and prehistoric modelling methods would be really counter-productive in my opinion.
And you may believe it's pointless, it's just your opinion. But for me, is totally the contrary as you can see clearly in the video i posted (working in expert mode).
In any case, being able to use Blender sculpt tools in edit mode would be a better functionality, and nothing counter-productive. Not all people use the sculpt tools for modelling in modo, and less, the procedural tools or the hundreds of feautures packed within the app. But they are there.
Switching modes is not an option for me. Like i said, if i can't do a simple extrude or bevel while sculpting, the tool mode is useless for me. I was hoping to see a workaround, but it's impossible right now in version 2.8.
I gave blender a few weeks and a whole month, and i already know its flaws very well. It can't be used for serious subdivision modelling like Modo, Max or Maya. I don't understand why the perfomance is so horrible compared to ver. 2.79 (and it was terrible aswell, very slow). I only hope they solve at least the horrible viewport perfomance in version 2.82.
Anyways, like i said, Eevee is the deal for me. I only take that from blender.
Replies
That is, the original vertices of the mesh move less in areas where they are closer together.
And if you don't want the vertices to move at all you can set their edges as creased. There's some other info in the Blender manual too: https://docs.blender.org/manual/en/latest/modeling/modifiers/generate/subdivision_surface.html#edge-loops
Anyone know why this shader fails to compile? Line 16 has a "<=" that appears to have been incorrectly formatted as a "<" in the forum's text box, but once that's been sorted out I just get reports of a syntax error at an unspecified location.
Node graphs are nice for custom or complex materials. But if I just want to setup the good old basic inputs (diffuse, normal, roughness, happiness) adding the same nodes over and over again is unnecessary time consuming. I don't want to waste time deciding where the nodes are places and if the uv data is routed correctly through.
What's the go-to workflow for this?
edit: I guess what I like to do is use an existing material-node-graph and create instances from it and just change the linked textures
Being able to use the GameDev tools as well, sweet Jeeesus
*edit* 11:15 brings a blenderman like me a tear to my eye
Also, has anyone had a crack at my previous post?
https://gumroad.com/alexbel#NbMya
Otherwise have a look around in:
https://blenderartists.org/t/list-of-addons-that-work-with-2-8/1132912#heading--UV-Editing
Is there anything else for you that comes to mind? 99% of my modeling is already done in max with hotkey smart function scripts and stuff like keyhydra. I almost never touch the command panel or menu items for modeling already. Even have my tab and caps lock keys rebound to 50 or so functions. I'm looking in terms of unique features, such as working without a gizmo.
@Justo
Interesting, could you elaborate more on the pivots/orientation? Specifically a use case where you can't achieve it in max?
• Off-hand transformation (there is no official name for that, that's just what i call it). The idea is that on top of not having to grab any onscreen gizmo, there is also no need to grab onto the active object or component(s), as your mouse cursor can be anywhere in the viewport at the time of, say, move or rotate being activated. The transformation is done according to the on-screen relationship between object and mouse cursor (with a little dashed line visualizing the link), also with shift key acting as .2 multiplier (I think ?) reducing the effect. It is somewhat similar in concept to the Zbrush transpose line but much, much better in its implementation. It becomes second nature very quickly and once you go there all other 3D programs and level editors start feeling like outdated antiques
• Pure modeling won't be much different from Max as the tools/features are similar. Skin weighting and morph target handling is an order of magnitude more powerful than Max though, as the data is stored very cleanly (vertex groups) and doesn't get destroyed when editing a model/splitting/merging parts. Huge time saver.
• The 3D cursor, not only as a way to set custom transformation centers on the fly but also as an aid to viewport navigation - by having it snap to surfaces and then pressing a key to focus to cursor. Navigating inside nostrils has never been easier ! But more seriously this is great and completely eliminates 3D viewport motion sickness.
• Overall (a little less true in 2.8 than 2.78/2.79, but still valid compared to Max) : the UI is way more responsive than Max/Maya (or at least the last few versions I've tried). Buttons don't have any micro stutter, UI elements are rock solid and never glitch in/out, panel resizing is butter smooth, and so on. There is just less "crap built over old crap" and it makes for a very responsive and robust experience.
• Updates/installation management is a breeze. No need for any installer, user preferences are all stored in a folder, and so on. Meaning that testing out new features will never feel like a drag - it's a matter of minutes.
• Now I would say that the biggest downside is scene undo, which is objectively broken (extremely slow) at this time.
• Also, while the new viewport is extremely powerful, overall all viewports (even raw shading) tend to refresh mush slower than previous versions when working with animated skeletal meshes - which imho makes 2.8 not suitable at all for animators at this time.
I miss some modifiers from Max or Maya, but i think it's a matter of changuing the modelling philosophy. I remember when i started using modo 1.0, i missed all from Max. Right now i do things in a different way.
I'm heavy customed to press space in order to drop tools in modo, and "alt" + Ctrl or Shift to control Viewports. In modo i can do what i call poly-sculpting, i don't use a mouse with modo. I mean, using the sculpting tools, i don't need to use any gizmo or anything, just some key shortcuts and that's all. with Blender i'm still having issues to work as fast as in modo + Cintiq.
Do you guys use Blender with a cintiq/Intuos for modelling? I do, and i'm having a hard time
I definitely will try it out. Currently the chamfer mod in max 2020.2 is seemingly the best out of any DCC app right now. Although I don't mind moving to blender for its other features.
@pior
Fantastic thank you for the write-up. Handling transformation like that definitely feels more fluid for a 3D app. Anything relating to sniping clicks/elements and forcing UI popups/spinners can really break the flow when you work. For example using the mouse wheel in blender to dynamically add chamfer segments or loops. This sort of stuff should really be native in max by now.
Yeah I saw your cool post on copying half a head and still having it animate! Very cool.
Hopefully, the viewport performance improves, I generally deal with 4+ mil items with hard surface work. I'm definitely excited to learn blender considering how well it's supported and has a very active community. Plus having a single program that can do a majority of tasks reasonably well, eg sculpting and animating makes it a very convincing option.
Also the ability to place any menu entry as a button into the UI in Modo, some functionality in Blender requires too much searching in menus, or the use of a shortcut. The Blender UI does not feel flexible enough, or not flexible enough for the standard user. You can probably do a lot with scripting ect, but thats not something I'm into.
I agree that working with a pen tablet kind of feels suboptimal in Blender in a couple of places, a couple of things/ addons are geared towards using a mouse + mouse wheel as input device. I'm using an old Intuos4 pen tablet. Trying to model with a mouse feels like trying to sculpt delicate details with a brick to me.
Now practically speaking, my advice for that would be to first not worry at all about any of this and just using the app for about a week or so with everything default just for the sake of digging into the features, also messing around with the various settings and playing around with editing the keymap manually without worrying about breaking stuff (and you likely will). Then, starting fresh again (by wiping out the preferences folder in Roaming) and taking as long as you need to carefully set up viewport and interaction controls to be exactly the way you want them. And then keep on exploring from there.
Here's an example on how to set this up manually (done with an older Blender version and replicating the Maya-style inputs but the concept still applies)
https://youtu.be/12fqTUyDts0
As for tablet use : well, here too I would say that this is a bit of a strange question. Of course it can be used with a tablet - just the same way every other program can be
My workflow is similar to working in Zbrush + "Zmodeler next gen" --> Modo sculpt + better polygonal tools (what i call "polysculpt modelling").
I wish i could select faces, edges and points in Blender Sculpt mode, edit the cage mesh/subdivision model and so on. For me, using blender right now it's like going back in time, decades, to my old days of Max and Maya. My whole workflow is totally broken using Blender, as in Max or Maya.
@Pior I ended changing the keymap and navigation to something like Maya, like i did with 2.79 but worse, needing to change more keys than expected because it ends broking a lot. Blender 2.8 changed all, it's like a slap on the face.
https://www.youtube.com/watch?v=Pb4UqxQormA
I found that swithching to "alt", i was broking other key bindings involving the "alt" key such as "alt + middle mouse + mouse move" (to easily rotate to top and side views without touching the wheel gizmo).
I really need to see what i can, and what i can't change. I restored all to default, and i'm with the Blender Manual studiying all the new features, improvements and changes. But i'm seeing that i won't use it for what i'm customed to do very fast, like modelling and uv mapping. Anyways, i like a lot Eevee and its render times.
As mentioned a few times I also personally rely very heavily on the 3D cursor for navigation, to precisely set the camera orbit center exactly where I want it. This is imho a great example of of Blender-only feature that a user hoping to replicate the input paradigm of another app wouldn't think of (I certainly didn't) yet it yields great benefits. Having perfect control on the way the camera behaves has a tremendous impact on flow/staying in the zone while working, and that seems to be exactly what you are are looking for.
So yeah my recommendation would still be to keep playing around without fear of breaking things and with no expectations of results for a little longer - and then starting fresh again after about a month or so of indepth exploring. All the grievances you mention are imho non-issues once everything is setup to your likings.
A practical example : many users seem to rely on the so-called "Specials" menu (previously W and now right click, if I remember correctly). That's something I personally never had any need for since there are other/faster ways to access these tools, so I had to take the time to disable the new 2.8 right click menu in all contexts to make way for what I use right click for (instant cursor placement on surface). It sure took a bit of time and could be seen as annoying, but it is also very easy to do thanks to the way the keymap editing screens gives access to everything.
TLDR : I can confirm that the fluid workflow you describe is 100% possible
I want to work like i shown years ago in this video. ~30 min timelapse modelling, using the sculpt tools + modelling operations with key shortcuts.
I'm not sure if you can rebind the mouse stuff for Blender yet, but you could probably use a workaround. I'm pretty lazy so I just used a program called x-mouse button control and you can swap your mouse keys per program. Gave me a lot of relief changing the pan/rotate to rmb from mmb in max.
It is true, that once you get used to dropping tools in Modo, it becomes hard to adjust to other applications that don't use that approach. In Blender you can edit tool or active tool properties but once you "drop" the tool by clicking onto something else not related to it then you cannot edit its parameters. In the lower left you will find a tool properties associated with the active tool. GUI wise its not ideal but its more of an annoyance than it is a game breaker.
As for sculpt mode and edit mode, they are two different modes. I often flip between them by using the 1-3 key and the 5 key depending on component or sculpt. You can also toggle proportional editing to move components around without the sculpt tool. However yes, the two modes are completely separated, where as in Modo everything is technically consolidated.
If you watch closely the video i posted, i'm sure you'll understand me better. I just use the sculpt brush for all, and a few keymappings. I had to record that video to explain the polysculpting modelling workflow.
Instead if you are curious about the app just dive in casually without expectations and explore/understand the systems fully for about 2 weeks to a month. A practical example : you seem to be getting stuck with the assumption that since Blender has two distinct modes for editing and sculpting, then that means that you cannot replicate your hybrid modeling/sculpting workflow (which imho is nothing unique really, and I don't mean it in a snarky way : I personally cannot understand how anyone could work on organic models any other way as this is objectively the best and most efficient approach).
At the surface level this is of course true : Edit mode and Sculpt mode are separate. But in practice switching modes is probably just as fast as selecting the sculpt brush in Modo ; and switching mode + selecting the grab brush can be done in one sweep with a macro anyways.
Of course it won't be behaving exactly like what you are used to, and even performance might differ - but that's an obvious as these are two different pieces of software.
So yeah I honestly think you are looking at it from the wrong end, especially since you can very well use both apps in tandem for their respective strengths. If anything, you might not find exactly the workflow you want ... but you might be able to precisely document the things that you feel are missing, and these missing things could end up being implemented eventually based on your very own feedback. Think of it as an investment in the future evolution of the software.
As for things like working in full screen mode with a stylus : as said, this is 100% possible since everything can be mapped to anything you want. The buttons/hotkeys might not be there by default but it is perfectly possible to assign hotkeys to anything you want using the keymap editor, or diving further by creating menus and macros for anything you want using PME (Pie Menu Editor, which is not just for Pie Menus but also for custom hotkeys/macros).
The 2.8+ iterations of BC are as powerful as ever but the addon is becoming seriously bloated compared to what the tool once was (it has a UI button to "activate it" but no button to switch it off ; and it now has dozens of options and modes). It is reaching a point where I have to constantly disable/enable the add-on for it not to get in the way and that defeats the whole purpose. I've tried a few other ones but I would welcome any suggestion. Thanks !
I feel you on that. I find surface snapping behavior enabled by default to be less consistent than simple screen-space orientation for cutting...
Once you are customed to work with a tablet pencil, using a mouse is like something really primitive, not evolved and not efficient for me (for god's sake, we are in the end of 2019!). I tried "Alt" key combinations for viewport navigation because: 1 - it's like the standard (Maya) and 2 - it's a good choice using a tablet or Cintiq. Not because we use the "alt" key in Modo.
Believe me, when i discovered the possibilities of polysculpting modelling, i said to myself: i won't go back. Going back to the old and prehistoric modelling methods would be really counter-productive in my opinion.
And you may believe it's pointless, it's just your opinion. But for me, is totally the contrary as you can see clearly in the video i posted (working in expert mode).
In any case, being able to use Blender sculpt tools in edit mode would be a better functionality, and nothing counter-productive. Not all people use the sculpt tools for modelling in modo, and less, the procedural tools or the hundreds of feautures packed within the app. But they are there.
Switching modes is not an option for me. Like i said, if i can't do a simple extrude or bevel while sculpting, the tool mode is useless for me. I was hoping to see a workaround, but it's impossible right now in version 2.8.
I gave blender a few weeks and a whole month, and i already know its flaws very well. It can't be used for serious subdivision modelling like Modo, Max or Maya. I don't understand why the perfomance is so horrible compared to ver. 2.79 (and it was terrible aswell, very slow). I only hope they solve at least the horrible viewport perfomance in version 2.82.
Anyways, like i said, Eevee is the deal for me. I only take that from blender.