Blender 2.8 Beta and Roadmap

We are currently an estimated 2-3 weeks away from releasing a 2.80 Beta version, which will have all the main features that will be in the 2.80 stable release.

Currently, a few areas area still being completed:

  • Stable and well documented Python API for add-ons developers.
  • User interface for Collections in the 3D viewport.
  • Keymap for left-click select that works well with the new tool system, and a new keymap compatible with other software.
  • Making image empty objects a fully-featured replacement for background images.
  • Restore support for all the physics systems within the new dependency graph.
  • New subdivision surface and multi-resolution modifier implementations based on Pixar’s OpenSubiv.

Blender 2.80’s viewport and EEVEE work better on many graphics cards now, but there’s still a long way to go to find and work around the quirks of the various graphics drivers and optimize performance. Besides this, bug fixing and user interface polishing is ongoing; with the stable release expected in early 2019.

What’s Next

After the first stable release we will go back to doing frequent releases, and incrementally add more features and improvements in Blender 2.81, 2.82, and so on. We will be defining the roadmap for them in more detail as we get closer to starting work on them.

There’s a few features that won’t make it into 2.80 that we plan to add in following releases, particularly Asset Management and Overrides. Both projects are quite advanced in terms of development but they need testing and design. Furthermore, two significant projects that are in early stages of planning are the Interactive Mode, and the Everything Nodes project to make Blender’s animation, particles and modeling toolset more powerful.

Feedback

But we also want to hear from you, the community, to gather what the most important things are for us to prioritize. More EEVEE features? Better texture painting or sculpting tools? Improve performance in big scenes? Or better interopability with other software? Work on these future releases is still a while off, but you can let us know what your top priorities are in the comments.

354 comments 153,883 Views
  1. “Mantaflow” and “Fracture modifier” is not ready… But maybe for later versions? Or its because of future “interactive mode” such physics systems going to be discarded?

  2. thank you brecht for providing a more clear view what is going on with blender 2.80, probably some people were exited for the asset manager and overrides but it is what it is at this point so they’ll just have to wait.

    in my opinion, what the developers should prioritize right is the following order:
    1_improve performance in big scenes (i think it was talked about before unofficially but hope this will get it due diligence)
    2_better painting/sculpting tool and maybe a little more love to the texture editing
    3_better interopability with other software
    4_improving the physics and particles systems and removing some of the limits they currently have

    the rest is up to you people as far as i’m concerned
    good luck to you guys.

    • I would love to see more painting brushes and tools along with sculpting tools.

  3. Hello 🙂 do we know when mantaflow and UDIM texture are being implemented? would it be in the 2.8 beta or later versions like 2.81/2
    Again thank you for the amazing work.

  4. Thank you for this awesome update! Looking forward to this iteration in its fully functional capacity. Has been fun taking ‘er out for a spin already. You all have done a great deal of hard work and this is highly exciting!! Congratulations to all of the Blender devs!

  5. First thing after 2.8 release should be Texture Painting workflow with strong connection to Cycles nodes (more below). It’s not important for me but it just looks like logical step to utilize EEVEE in more fields than it is now.

    1. Particle brushes and dynamic brushes. I don’t now if it’s possible, but using textures made by nodes would be huge thing. Without baking them.
    2. Making more Cycles nodes compatible with EEVEE – mainly Bevel and AO for edge detection and wear&tear masks. With this goes patch for Noise and maybe more procedural weathering options. AO in current form is painful to use.
    3. Optimizing them for easier use in this job.
    4. Cursor along normals, started here: https://devtalk.blender.org/t/brush-cursor-implementation/1656/15
    5. Simplify baking. AO, Curvature, and other related to geometry helpful with painting process.
    6. Vector shapes in nodes and tools to utilize them, check here: https://blender.stackexchange.com/questions/102692/randomly-populating-procedural-shape-or-image-texture-in-cycles

    And yes those are similar things to SP and SD, because we are so close to them, yet so far.

  6. Well, as we speak I’m in the middle of a gig where I have to cooperate with other SW. So that’s a biggie, but Asset Management was my main *wish* to be completed for the release. Oh well. I’ll wait. I see more and more people moving away from the hobbyist blender user realm to the pro blender user world (I think this is a good thing) . That being said the linking system, asset handling and all the workflow easing features should be in place a.s.a.p in my opinion. I mean posting gifs on twitter about new features is nice, but working in a non-destructive and reusable manner and having a robust out-of-the-box import-export system will benefit everyone in the long run. Sometimes the non-sexy features will help more people in the long run. I like blender way too much to switch to anything, I ‘ve ditched maya in my personal work, now I’d like to step away from commercial SW and sail the Open seas of Open source, till death.

  7. NDI
    First of all, a big compliment for your work, you offer the world superior quality tools and I would always be an experienced evangelist to spread them. I work in entertainment on huge surfaces and we start using newtech’s Network Device Interface NDI technology to broadcast from one software to another, this allows for example to output full resolution a window after effect or unity in unreal, or vice versa. I see a huge potential for my industry to use Blender and Eevee in live mapping especially if interactivity is implemented I have already tried but my only blocker has been resolution. here is my request: implement NDI technology and allow blender to output a full viewport live resolution window without borders and without the tools as a rendering in order to broadcaster it in mapping surfaces or other programs.

    this is the SDK of NDI

    https://www.newtek.com/ndi/sdk/

    Thank you again for listening.

    Translated with http://www.DeepL.com/Translator

    • Implementing their proprietary SDK/middleware within blender is likely out of the question purely because of issues related to licenses, since they are very likely incompatible.

      • Blender 2.8 is a truly amazing update, thank-you! I also use Blender to create content for live productions. Syphon (on Mac) support and Spout (PC) with eevee would be incredible, supporting the streaming of video frames to live show software like Qlab, Resolume, VDMX, Max / Jitter etc. perhaps these could be supported by addons at some stage?

    • What about access to spout 2.0? that too would be interesting.

    • I wish ndi was emplemented…
      I have been doing vifeo mapping and imteractive installations with blender..
      My work around.
      -I took a wile to know this one. Alt+Ctl+space will give you a fullscreen with out widgets clean for capture.
      -Vision mappio (mapping tool.. And free for linux) lets you capture blender window
      -in linux you can make a virtual screen and place blender window there then just tell mappio to capture that widow

  8. Top priorities: Improve performance in big scenes and Asset Management.

    One of the things why i work in blender is its performance in big scenes. Asset management is also very important for a fast workflow for me, so will be nice if its supported soon.

  9. Great work so far on 2.8. I think in sculpting, biggest tool I think you need in sculpting is an angle tool. For example, if you need a 90 degree angle, it would be great to just select an angle tool and define the planes you want smoothed. I would like to see an overlay in the 3d view port of what brush you have selected. That way you can be sure to have the correct brush without having to resort to switching to panels, etc. Still waiting on the animation add-ons to work like rigify as well as the ability to import old rigs into 2.8, but that may be wishful thinking.

    Again great work so far. The UI is starting to come so keep tinkering with it.

  10. Top priorities: Improve performance in big scenes

  11. Great work !
    For us here (a french TV serie studio), the most important need is Cryptomatte (and AOVs) !
    Thanks for all this work !

    • You rock!
      I would like easy export of ID masks , Z-depth, light and diffuse passes for cycles, like Vray have. – I work mostly with single image visualisation.
      It would be nice to have better nurbs tools, like the way MoI 3D or rhino works with extruded and revolved curves etc.
      I am looking forward to everything nodes, i hope it can also work with nurbs.
      Evee is amazing. I would like to have a VR mode for Cardboard, Vive and Oculus like you can do with Sketchfab on Firefox.

    • Please release the 2.80 beta as much as quickly. I cant give textures to my models because blender get close itself whenever I get in to the materials section and render section. It closing every time . Plzzz release beta version fast.

  12. Are nodes currently out of Blender? I cannot find the Node Editor in 2.8 and for me Sverchok is an important need.

    • The shader and compositor node editors now appear as their own dedicated editor in the list. For addons that add their own nodes there should be a general Node Editor or their own entry in the editor list.

    • I agree sverchok is really impotant

  13. Great jobs, congrats to all of you 🙂

    In my point of view, I wish some 3D realtime time workflow features (a lot has been done these years for movie, of course with the help of open movies), but this may comes with interactive mode?
    Here what’s come in my mind:
    – scene statistics (through the outliner why not: classify by poly/tri/vertexcount values)
    – normal editing
    – UV active channel switching more convenient? (maybe already doable with the new “favorites” menu?)
    – texture baking workflow enhancement (currently default workflow is a pain in a scene with hundreds of objects to bake, I feel in love with the Baketool plugin)

    By the way, i’m so happy about multiobject editing, that’s mean UV sharing made a lot easier, thanks!

    • @V!nc3r On the normals: You might want to check out the new Weighted Normals and Normals Edit modifiers. Maya’s normals have always been superior to every other package I’ve seen, but with the addition of those two modifiers in Blender 2.8, I’ve been able to achieve the same results and even more flexibility in Blender 2.8.

  14. Hello, after Blender 2.80 could you please implement the multiview branch ?

    Branch:
    https://developer.blender.org/diffusion/B/repository/soc-2016-multiview/

    Project:
    http://hlzz.github.io/blender1/
    http://hlzz.github.io/blender2/
    http://hlzz.github.io/blender3/

    Documentation:
    https://en.blender.org/index.php/User:Tianwei37/User_Documentation

    I hope that the code is compatible with 2.80.
    Thank you.

  15. Great work so far guys!

    In our VFX studio we work hard to fit Blender into our pipeline as much as possible.
    Top priorities for us:
    Overrides + Override Interface, Linked Asset/Scene Interface –> The interface should allow to see at a glace what is linked into the scene and what overrides were applied to the linked assets. From this interface the assets need to be exchangeable, the overrides sortable and deletable.

    This is our most important request, because once this is working not only our studio would switch to Blender completely 😉

    Keep up the good work guys, you are great!

  16. Image sequencer ..that would be great for us green screen people ! At the moment it only ever sequences to one frame and will push on no further so it wont animate a frame sequence …….It was working in March/April Eevee but it was obviously switched off for reasons best known to you legends ..Would love to see it back on in 2.8 Beta..have a whole city built that i want to incorporate a real person into, so i would love this to be switched back on, it would enable our music video with our female singer to happen ..I’m not sure it’s a popular fix or even wanted for most Blender users ..but it would be the whole world to us !!

  17. Please finish the cycles baker to at least be on par with the removed BI baker.
    Displacement baking and the often useful bias option are still absent, and I’m not sure if the high memory usage when baking from many source objects has been resolved.

  18. >Better texture painting or sculpting tools?

    Yes! Let Allegorithmic make room!

    >Or better interopability with other software?

    Yess! Especially with Unreal Engine 4!

    And some more things I’d like to ask: value switching button in nodes (for example, switch two colors in Mix RGB node) and the ability to rearrange UV maps in the stack of Object info properties.

  19. Hi, thanks for the quick and concise summary! Blender 2.8 is looking great!

    I would find the following features important:
    – interoperability with other software. More varied and robust input/output capabilities. Possibility to import nurbs file formats (ie. with tessalation), there was a great GSOC project for this that was almost done, but then abandoned and never integrated into Blender.
    – improved asset management.
    – increasing performance on big scenes.
    – Expanding the Outliner with more scene-management features.

  20. Awesome!

    Please don’t forget about Cycles now that there is Eevee. Cycles needs a lot of love – adaptive sampling, mipmapping, performance and memory improvements across the board (displacement is so slow compared to Redshift, Octane.) out of core textures/geometry, Custom AOVs, Cryptomatte, etc.

    Some other random ideas:
    – It would be nice to have some new, improved modifiers with easier visual controls – it would be huge appeal to motion graphics community. Just look at C4D modifiers for inspiration – Collision, Jiggle. Some sort of falloff system for modifiers also would be great – for dynamic control of affected area.

    – Layer system for Weight painting, Vertex color painting and texture painting. Super useful for skeleton weight painting – take a look at https://www.ngskintools.com/ for Maya.

    – Custom geometry attribute import throgh Alembic – accessible in shader nodes.
    – More procedural nodes for Cycles and Eevee

    – Object level render opacity parameter – allows fading objects without touching material shader. Idea is taken from Octane render (Object visibility slider) – it is super useful for motion graphics where you need to fade in/out objects a lot. Currently in Blender you need to create individual material for each object which makes things super messy. Also in Octane it treats object as a single “layer” – not showing internal parts of the object (not like when you just mix transparency in shader).

    – C++ API for addons would change everything, but I know it will never happen 🙂

    • You can use Weight Mix Modifier for layering vertex groups. Really powerful stuff.

  21. is interactive mode going to be in 2.81 or 2.80?

    • It will not be in 2.80. It’s not defined yet which later release it would end up in, could be there is some initial version in 2.81, or it might be later.

  22. Thanks for the great work you’ve done! I’m really excited for 2.8 😀

    Now, if you have to prioritize the next steps I’d go with this first:
    – Asset management and overrides.
    – Better performance with big scenes and/or dense meshes.
    – Improved interoperability with other software.
    – Full UDIM, Alembic and OpenVDB support.
    – Custom AOV’s and cryptomatte support for cycles.
    – Everything nodes!

    And after that please give the whole sculpting tools some love (both dyntopo and multiresolution), it’s been quiet for a long time now.

    • I’m really excited for 2.8
      Vote for
      1. Better performance with big scenes and/or dense meshes.
      2. Full UDIM, Alembic and OpenVDB support.
      3. Custom AOV’s and cryptomatte support for cycles.
      Thanks

  23. I think EEVEE features are pretty good already, if we doing high end render we should use Cycles anyways. Texture painting is just simple too good in Substance Painter, so we can’t expect Blender could be a real competitive. Improvements in big scenes and dens meshes are good idea, but scuplting improvements are the best idea since I’m sure if Blender developers would work much on this field less and less Blender user would open Zbrush, and more and more Zbrush user would consider the free alternative, I belive Blender can compete on that field!

  24. Can’t wait for the Beta version!

    One request: As an addon developer, I’d like more control over memory management in Blender. Using the Python API, I’d like to be able to delete an object AND free up that space in memory.
    My addon (https://github.com/neverhood311/Stop-motion-OBJ) imports lots of meshes into the scene. The number of meshes a user can import is limited by what they can fit in memory. It fills up really fast for large, complex meshes.

    • Just use bpy.data.meshes.remove(yourmesh). This will free the memory afaik.

      • Yes, it should be possible to unlink and remove both the object and mesh. Note they will remain in the undo system memory, but a memory limit can be configured for that.

      • bpy.data.meshes.remove(myMesh) definitely reduces memory consumption, but doesn’t free all the memory that was allocated by importing the mesh. For example, I opened Blender (2.79 and 2.8) and the memory settled around 110MB. I imported an 89MB STL file and the memory rose to 500+ MB. I then removed the mesh from bpy.data.meshes, and the memory dropped to around 240MB. Definitely better, but still a far cry from what it was before importing the mesh.

  25. i know you guys are having so much at hand right now, so performance and stability is the key here, especially now with EEVEE..and for future projects i think improving UV/Texture painting, sculpting(which should be a bigger project), more Grease Pencil features like for storyboarding…etc.
    also i can’t wait for the road-map for animation 2020.Good Luck!

  26. Go back BGE and Blender Render (Internal)…

  27. Awesome news, I’m looking forward to the beta. ?

    Personally, my two biggest quirks with Eevee right now are:
    – the way it handles shadow maps. I’m surprised how noisy, and problematic they are – both ESM and VSM often don’t look great. But I assume there’s no easy fix for that. ?
    – Light brightness isn’t consistent between Eevee and Cycles. It’s pretty easy fix, but I wish I didn’t have to do it every time.

    Other than that, it’s pretty much exactly what I was hoping for. Good job dev team!

  28. Better texture painting, pleasepleaseplease! I want my workflow to be as Blender-centered as possible. 🙂
    PLEASE AUTO SAVE IMAGES! Or at least warn about the potential loss of data when an image has been modified and not yet saved. Something more than that little asterisk, please! Or even an option to toggle auto saving of images for us dumb-dumbs.
    Also, any kind of documentation for 2.7 users to help transition to 2.8 would also be great. Something like “If you expected THIS from Blender 2.7, do THIS in 2.8” .
    Keep up the hard work, we love Blender!

  29. I’m not seeing the asset manager or did I miss some important news about that?

  30. Improve performance in big scenes
    Everything nodes
    Better painting

  31. Will displacement rendering + adaptive subdivision be moved out of experimental for Cycles? How about EEVEE support? Thanks!

    • Cycles displacement is already a supported feature now, we hope to do the same for subdivision as well soon. EEVEE support for these features has not been worked on yet.

  32. Please, we need “shadow catcher”, “shadeless” materials, and more “render passes” for EEVEE. If it’s going to replace Internal Render, we’re going to lose the ability of making fast VFX combining real footages to rendered animations. There’s no way for a lot of us to use Cycles for long animations, because of render times. Thank you!

  33. Thanks for the update Bretch!

    I would vote for:
    1- Improve performance in big scenes
    2- Asset Management and Overrides

  34. About animated images / texture sequences have already been discussed above. I support !!! Very necessary and useful option. Well, I would be very grateful for the “ChromaKey” node in the shaders. That would be just amazing!

    Next one: I’m not sure if this is a bug: a point source of light and a spotlight project non-existent shadows if they are at some distance from (or behind) the camera, provided that the value of the “bias” is very small. Perhaps this is the limitations of the engine, but I would like to know, can I somehow fix this?

  35. Will full motion blur in EVEE make it to 2.8?

  36. Thanks for the exciting update, Brecht.

    My vote for most eagerly anticipated wish is an improved Sculpt Mode. There are a number of things I really miss in Blender sculpting when compared to ZBrush:

    • An alternate smooth brush mode that preserves volume.
    • A general smoothing tool to auto-smooth a complete mesh or a masked area (in stead of having to add and apply a Smooth modifier each time).
    • A general mask blur option, in stead of having to smooth the edges of each mask manually.
    • Easy extraction of masked areas into a new geometry layer with a desired amount of thickness (now you need a separate add-on for that).
    • Topological sculpting (like the Move Topological brush in ZBrush). Now you have to mask surrounding geometry each time.
    • Polygon grouping and smoothing of the polygon group edges.
    • A solid auto-retopology tool (like ZBrush ZRemesher or 3D-Coat Autopo).
    • Easy cutting using a custom curve or shape.
    • A tool to easily Boolean all intersecting parts of the same mesh (like the Remesh by Boolean deformer in ZBrush).

    Many thanks.

  37. 1. Everything Nodes
    2. Performance in Big Scenes
    3. Asset Management

    Amazing job! Blender is turning into a beast of a software!

  38. 1) GLTF support.
    2) FBX support. Make it even better than it was in 2.79 by enabling the Principled node to work with FBX. Blender should automatically recognize the textures attached to the Principled node, and assign those textures to the correct material slots (roughness, etc) in the exported FBX. This would be amazing for game developers.

  39. UDIM Support is the most important feature right now

  40. Very good work so far!. It’s a pity that some of the main features have to be delayed, but that’s development, right?.
    I would love to see implemented, as very needed features:

    UDIM, they are a HUGE timesavers in terms of lookdev and texturing workflow!

    TEXTURE BAKING directly from the node editor.

    BETTER TEXTURE PAINT SYSTEM. There are multiples petitions in right-click-select

    OVERRIDES and IMPROVED REFERENCE MANAGER SYSTEM are seriously needed too, from the perspective of professional work in a studio (I can say that it was two of the key motives the studio I work for abandoned the idea of migrating to blender for the time being).

    MORE COMPLETE RENDER VIEWER (like arnold’s new renderview, with a shader debugging mode, etc)

    BETTER POLYGON MANAGEMENT for sculpting.

    LIGHTING: arnold’s-like BLOCKERS, and a good LIGHTING LAYER SYSTEM/ better VIEWLAYER SYSTEM (one that lets the user see the settings of all renderlayers in a more intuitive manner, now it is kind of obfuscated)

    BETTER BIG SCENES MANAGEMENT (although I think that will be improved over time, since it is still alpha state)

    But if I had to vote just for one, I, personally, would ask for UDIMS. They are so-much-better to work with!

    Even so, again, great work!, you guys are great!

  41. Wow, hard options to choose at the end! But I guess better interoperability with other softwares would help us to keep our workflow up. Like Beniamino said about to keep fast VFX that will probably be edited somewhere else; or FBX Unity / Unreal support + baking options in my case.

  42. All of the above suggestions are fine, but most of what I wish:
    – Make object creation less destructive; there’s no reason not to be able to dynamically change the number of sides in a cylinder etc. as long as I’ve not edited it. The current “create primitive -> tweak settings -> done” workflow is too destructive. Workarounds in the community are not valid excuses to continue to live in this workflow. Let us continually change all object creation parameters up until we enter object mode for that primitive at least.
    – Cleanup and remove the need for the Loop Tools Add-On. Much of that Add On should be built in natively and or at least de-duplicated with what Blender already offers. Cleanup and sort out all the other shipped Add-Ons as well. Most are complete junk or could stand for some better standardization or implementation into Blender proper.
    – There’s no excuse to keep “Easy Lattice” in the completely unrelated, mostly crap, Add Advanced Objects Add-On. Creating a lattice FFD immediately surrounding the selected object should be a basic task built into Blender
    – Add ability to visually create a Loop-Cut anywhere in a quad. Do not force the cut to start or be triggered to start only in the Middle of the quad.
    – Implement a “Set Flow” operator for edge-loops natively (see other software for what it does)

  43. Game Engine for EEVEE !!!!!

  44. Any news on RTX development for Cycles – when do you plan to add HW accelerated ray-tracing?

  45. Better snapping and modelling. Cleaned-up UI. Some UI elements are still meh.

  46. I am proud of current development, it really makes a HUGE difference. If I would wish anything more it would be:
    – make EEVEE visually even more stunning. I mean: give us robust light baking system like in Uity / Unreal to get GI right
    – continue with Everything Nodes – that’s great direction, especially when we see Niagara system in Unreal and Visueal Effects Graph in Unity – Blender will need similar solution if it wants to be on par when it comes to VFX and physics
    – gain userbase: make Blender more compatible with external engines (Unity and Unreak shaders out of the box, Houdini implementation, Substance Live Link in Blender that would work smoothly like Live Links in other game engines, etc.)

    Basically: real-time rendering is the future. And not distant one. Go for highest visual fidelity, nodes and compatibility. AND CONGRATS FOR THE HARD WORK!!

  47. Since I work in a planetarium, my number one feature request would be to support the panoramic camera types (Fisheye, particularly) in the viewport/Eevee.

  48. Full support for open vfx standards like Pixars USD, Alembic and OpenVDB so that Blender can get integrated into more professional production pipelines.
    This would most likely increase Blenders market-share and the Blender-community.

  49. i come from maya and i would say some important features are missing and hard to do in blender
    like permanent changing of the Pivot Point,snapping..etc
    if u want to make a vert/edge/face as the origin for pivoting and on multiple objects is a nightmare.
    in maya it’s just one click makes the center point editable and you can click anywhere (in object mode) vert/edge/face, u can use gizmo to move it and place it easily,
    probably making the origin editable as a pivot point will be very useful which leads to making better snapping tools.
    also real object grouping and better display layer system than just out-liner which clutter with information and not Artist friendly.
    and i got many more ideas but in due time.
    Thanks for reading 🙂

  50. ASSET MANAGEMENT!. With ‘collections’ finally implemented this would be the next priority. Glad to see so many users with the same opinion. We should all be unified for the better.

  51. It’s nice to have an idea of when the Python API will be stable for us addon developers.

    – Better performance with big scenes. My addon creates lots of objects for large scenes (cities) and quickly Blender’s performance drops exponentially it seems. Past a few thousand objects, it grinds to a halt when adding new objects. I don’t know if 2.8 fixes that.

    – Improved interoperability with other software, especially FBX must be rock solid, as easy as possible, supports materials, latest FBX versions etc…

    • I love the new update to your add-on! I would love to know if that would be a thing too 😛

      • Thank you 🙂 I’m adding objects instancing for the next version, so the more objects Blender can handle, the bigger and more detailed the cities!
        Sorry, what would be a thing?

        • sorry I meant this part of your question: Better performance with large scenes, I almost killed my works computer 😛
          but that’s awesome news about objects instancing, looking forward to more of your tutorials on your youtube

          • I’m planning to make new tutorials yes 🙂 What tutorials you would like to see more?

  52. 1. increasing “big scene” and high poly viewport performance should be the top priority over everything else IMO. Doesn’t matter what shiny features you add to the program if it lags and makes me want to pull my hair out
    2. Sculpting at this point has been neglected for far too long and it’s behind any other sculpting software that exists currently. There isn’t even a workaround for certain things like trying to implement sculpting layers.
    3. Better texture painting and texture management like UDIMs and such. The node system should introduce some “filtering” options like blurring of certain textures in the node/shader editor.
    4. Better Baking workflow. The way it’s done now seems really unintuitive

  53. I know this is pretty minor, but shortcut management and theming are a complete nightmare. Also more control over the UI and what add-on panels go where, etc, and some sort of permissions system for add-ons would be really nice (e.g. permission to change UI, to change shortcuts, etc).

    Otherwise, I would like to see improvements to the sculpting tools and texture painting (I always find it really confusing).

    Also I know it’s extremely unlikely, but I wish Blender had some decent CAD tools.

  54. Thanks guys for this amazing work!

    I’m already fascinated on what Blender can do, however I still think Blender lacks some complexity in some cases, for example, there’s no option to control the curve of the proportional editing, the snap lacks a lot of options, there’s no scene statistics, no physical camera, best architectural modeling tools. Blender is a fantastic software, but is about time to add those complete features.

    Also, scene and asset management. This was also my TOP priority feature. Blender really really needs this management tools, Is really important. I got really excited about it.

    And finally export-import tools and interoperability, most artists work on multiple platforms in those days.

    Nobody here doubt it how great Blender is, but after all this adding of tools Blender needs to set it down and polish everything.

    Thank you very much! o/

  55. Nvidia RTX hardware support for cycles!!! That will be such a game changer! Please make this a priority! Thank you! 🙂

  56. Yes! better texture painting or sculpting tools.

  57. i’ll ask for one thing only a Professional Retopo Tools built in blender, something like this one or even better…it’s a shame that blender doesn’t have this for many years.
    https://www.youtube.com/watch?v=6UKUxAeGqqI

  58. + Cryptomatte
    + AOVs
    + EEVEE passes
    Better Pass generation (easy to use/understand, please remove the layer masking system, it´s cumbersome, not human readable or change it so that it would be intuitive (exclude/include)).
    + CYCLES speed
    + Big Scene handling data

    I think this should be top prio until March 2019

    • I completely agree!

      I work for a Design/3D animation company and we use Blender fulltime, the compositing we do could be greatly improved in terms of quality and speed with the use of Cryptomatte, AOVs and EEVEE passes.
      Given the state of cryptomatte i really hope to see it in 2.8 (some parts are already merged i think)

      Cycles speed is ofcourse a welcome improvement but a never ending endeavor.

      Great work on 2.8 devs!!

  59. I’m happy with the developments, very happy. 🙂 I have some wishes though.. 🙂 Would it be possible to make an easy way to solo-view a collection? Like right clicking the ‘disable in viewport’ and select ‘solo’ or something in that direction? Or is that already a way of doing something like that?

    • ah…. i just figure that one out.. forget my comment right up there..

  60. What happened to Blender 101? Personally (and working with a lot of young students learning blender) I’m quite happy to have one interface for all, two intrefaces would require students to learn, well, two interfaces…. However, there are issues with the current interface that will cause issues for new users:

    https://twitter.com/peterejkemp/status/1050360844343750656

    • I will also chip in to say that it would be great to get a update (a developer blog article perhaps ?) on Blender 101. 🙂
      (also since 101 was a target for the 2.8 code-quest).

      If it will take time to get all the internal architecture in place to make 101 fully realised. Then maybe some estimates from the developers, when they think ‘sub-components’ will be implemented during the 2.8x release cycle.

  61. Grease Pencil is within reach of being able to do TV/web series production, but is missing a few things that seem outside the scope of addons:
    -Masking
    -Lattice modifier
    -Surface Deform modifier

    (Also, RightClickSelect seems to be displaying Blender ID users’ email address rather than their nickname on user pages)

  62. Would be nice to know if you had selected head or a tail of a bone in B bone or stick.
    They look the same to me.

  63. I would like to see more Eevee features, specifically some way to achieve genuine shadows from an HDRI background. One idea for how to implement this would be to borrow the “portals” concept from Cycles and modify it so that when the user adds an area light to an Eevee scene and sets it to be a portal, image based light and shadows from the HDRI are only calculated from that point and not the whole 360degrees of the background. This should be relatively easy to do and it would make interior scenes with Eevee look a lot better.

    Also, true soft shadows that automatically vary with the size/shape of the light would be great; I don’t care if it is slow to render, having the option would be great.

  64. Please make a keymap that works. I dont mind using 2.8 or 2.79 but I can’t hide stuff under 2.79 keyconfig and in 2.8 I can’t start the timeline with alt + A. I wish there was a happy medium where z works like it should and timelines can be played. But overall I’ll be pleased with whatever you guys do. I think Eevee will be quite a bit of fun in the meantime.

  65. Thanks for the news about your highly appreciated hard work! Like some entries above, – I would also be very happy to finally see the animated image sequences or movie file textures implanted!

    With the new OPEN GL engine and SSD that would make realtime playback possible?!

  66. I wonder how many of the people posting these requests are contributing to the development fund… I can’t give much right now, but I’m on it.

  67. Yesterday Ton talked about all the effort needed to “Make the Video Sequence Editor Great Again”(as an example), but maybe smaller steps like taking an interest in user submitted patches, could help move things along? This summer there have been submitting VSE patches without gaining much interest from the foundation devs. I know that the VSE development wasn’t on the list of targets for the code-quest, but now it is (almost) finished I hope it will be prioritized to spend the necessary time to help new voluntary (VSE) devs along.

    Examples:
    VSE SMPTE Charts: https://developer.blender.org/D3578
    Added generated Mask to VSE: https://developer.blender.org/D3622
    Color-coded strips in the VSE: https://developer.blender.org/D3579
    Sequence proxies by running parallel threads: https://developer.blender.org/D3484
    https://devtalk.blender.org/t/vse-proxy-improvement/1696
    Add font selection to VSE text strips: https://developer.blender.org/D3579
    Scale & stretch & cut sound waveform: https://developer.blender.org/D3496
    Add proxy support for all sequences: https://developer.blender.org/D3496

    And a vital modal handler function has been removed from the API, which will cripple one of the most comprehensive and advanced VSE toolsets – but no reply on the question of how the future looks for this add-on: https://devtalk.blender.org/t/porting-my-addons-to-2-8-missing-scene-update-post-handler/2465 or here: https://developer.blender.org/T47811

    As written I know there are good reasons for things being prioritized this way, but I fear if people who invest a lot of time on improving the VSE, aren’t taken more seriously, no-one will voluntary bother with working for the VSE anymore, leaving it up to the foundation to deal with all the mentioned troubles of “MVSEGA”…

  68. I would like better performance on Texture painting. But that probably means ditching the current projecting system. Anytime there are lots of faces behind the face im painting, it slow downs really bad, which makes little sense considering the face im painting is right on front, but the projecting method works backwards, first looking at all the faces, behind the brush, and then discarding the ones that aren’t on front. Hire the armorpaint developer :]

  69. Amazing work! But I am not certain if this “interactive mode” is really going to be useful, apart from “abusing” it as a game engine, which I have no doubt abot people are going to do. Why not go full game-engine then and include Armory? You wouldn’t need to really keep supporting it, since you kinda neglected BGE as well and it was still praised. You “only” need to update it every release with the Armory-community contributions and Blender will be able to compete with Godot, Unity and such! Also the people interested in Armory and so contibuting to the project will vastly increase once it is installed by default.

  70. There are two main things to make interesting for Blender 2.8x :

    1) VFX ==>> Bright more tool for 2D planar tracking more advanced as Mocha PRO or make 3D tracking as PFTRACK, we need it to do more effective in faking, add 3D in video more complex, etc…

    2) Game engine : High quality as AAA to make very modern in games for all kinds with support in Python or C++

    Everything else stuffs are very good too for work when peoples need them :
    – better for big scenes
    – everything in nodes
    – make EEVEE to approach to Cycles in terms visual quality
    – add to Cycles for BDPT
    – grease pencil to support 2.5D that make look like 3D

  71. I think the Blender now needs a new tool for sunsets, a Sky Generator with sea, sun(moon), horizon and clouds in the sky…Thousands of photographers every day go to the beaches to photograph the sun in the sea and the wonderful colors in the sky. This is the number one item in photos worldwide…
    The Hdri tool is good for one render image per camera view but not for a sunset animation or multiple changes in the clouds.
    I think that a quality addon in Blender is now needed along with the EEVEE for
    a animated sun and quality animations
    In other words, a new Bryce within the Blender(see my site).
    Thanks for your work.

  72. It is necessary to create a new portfolio site for CG Blender artists to promote our work, similar to the Behance of Adobe.

  73. Hi, I am also in support of better texturing tools

  74. Give the sculpting and texture painting some love. Besides Dyntopo, Blender’s sculpting mode was already too lackluster 5 years ago, when it got it’s last major update.

    Also, some kind of poll would be a better way to access it, maybe a poll for Blender Dev Fund and Blender Cloud donors.

  75. Along with suggestions above, I’d like to see some editor/object mode improvements like better snapping tools. I come from an XSI background, and the lack of simple, adjustable, on-the-fly snapping tools in Blender is frustrating.

    Also, more AOV’s/Baking output channels (like displacement for baked maps)…. and better noise reduction in Cycles 🙂

    But still, these are kind of niggles. Blender is amazing overall, so, thanks for all the work you Dev’s do 🙂

  76. Prioritize the bugs that make it unusable right now:

    Still crashing when using Intel drivers when resizing or closing windows.

    Rendering is still bad on Intel drivers – Some accumulation buffer or something is not getting cleared out, it introduces random pixellated garbage on the edges of items.

    Import addons not working (obj/fbx) – they’re failing that you can’t iterate over a float.

    • Note: Updating to the October intel Gfx drivers fixes the rendering issue.

      Intel users still get the “crash on resize” though.

  77. Parralax for eevee would be nice

  78. Please and improved paint skin weights tool. Is really a nightmare in Blender.
    Something like a volume paint around the cursor so you dont accidentally paint any other part of the geo, would be great.
    Thanks!!

  79. NURBS with actual surface updating when moving control points.
    Thank you 🙂 !

  80. Blender 2.81-2.82
    +Grease Pencil Actions
    +Grease Pencil Libraries (In sorts like armature pose mode.)
    Texture Painting slots and easier manipulation.

    ——————
    Udims has already been discussed 🙂

  81. Just wanting to put my two cents down from a production view (infrastructure viz) with Blender 2.8… Havent used it all that much but have enough to make us not switch on release.

    — Major issues

    — Pie menus, This should always be an option to enable / disable. having it on by default (and no easy way to disable it) actually slows down artists workflows. Moving your mouse is significantly harder then pressing a key (or key combination ) on the keyboard (z to toggle wireframe mode for example)

    — Dupli-collection and how that works. Our current pipeline is to have certain objects reside on layer 10 inside of a group at 0,0,0. This layer is never rendered. Empties are then populated around the scene with that object/s, which aids in speed in the viewport as well as speeding up scene creation. With blender2.8, this functionality is unusable. Creating a duplicollection works fine, but then there is no way of hiding off the original objects whilst still being able to render them.

    — Minor EEVEE issues

    — EEVEE with sun. The sun position of EEVEE is dependent on where it is located. This means for large scene flythroughs it is impractical to use EEVEE.

    — EEVEE shader compilation. Not sure if this is the right word for it, but on production level scenes, just getting into eevees rendered mode takes 2-5 minutes worth of time to actually compile all the shaders.

    — Feature Request

    There was a patch a while back for persistent data for cycles. This allowed cycles to render animations significantly faster, by only computing what has been changed between frames. This initial patch was dismissed as it would be more ideal for blender2.8. Blender 2.8 is now on the cards so it would be great to see that patch revived.

    Thanks for the hard work!

    Carlo

    • Regarding collection instances, that should still be possible? You can put all those collections inside a parent collection that is hidden. Or put them in a different scene.

      • Thanks for the follow up, didnt realise that you could link between scenes with dupli-collections nor did i try putting them in sub groups. Thanks for the tips!

  82. No particular order:
    * Non photorealistic render definitely deserve some love
    * particule system (check in game engine like unity3d)
    * Interpolatory between software like the couple After effects/Cinema4d
    * Improve workflow for freestyle (I didn’t try since 2.8 but last time was a painful experience)

  83. top priorities: Sculpt and Texture paint improvements!!

  84. Improve the VSE pls. Love u guys.

    • To all the people who have suggested improving the VSE: Why? Awhy would anyone use Blender to edit video when you could use a much more capable dedicated video editor? The excellent DaVinci Resolve is even free.

      • Actually it’s not a free and open source software.
        I use Blender also because of the freedom, in fact it’s free as in “free speech”, not as in “free beer”.
        Beside that the VSE, in particular with some add-on created by the community, it’s very effective for a lot of Blender users.

      • Because it is in the VSE all the elements generated in the other areas of Blender becomes “blended” together. No VSE = no blender 😉

        Davinci Resolve is an entirely different beast. Ex. it will only run properly on dedicated hardware and it can’t be used to cut unrendered mise-en-scènes in an integrated workflow in Blender. Unrendered, real-time movie-making will be a great part of the future and Davinci can’t do that, but Blender will be able to do this, with the right tools.

        OpenToonz, Krita, Synfig or Storyboarder didn’t stop Grease Pencil from becoming an ingenious part of Blender today. Why should Davinci Resolve be an argument for not developing a similar ingenious sequence editor in Blender?

        I know the lack of manpower is one of the reasons for the current state of the VSE, but as mentioned above, there are voluntary coders, patches waiting and here is an 8 pages long Google document of VSE add-ons, showcasing the interest and effort people are putting into improving the VSE: https://docs.google.com/document/d/e/2PACX-1vSBzP9e0JbKlXa-oHJYxsUoFpvGsiPfeX5MXRdnQmKB55COr4J8JGw6orlJb6r0kjR0RMM3Jv4Vz_zX/pub

      • What about to Hitfilm, it is cheap and it brights VFX… It is good VSE ?

  85. thanks to developers , you are awesome =) Udim and Asset manager is almost done as I know , it will be in 2.81 or 2.82 so I think no needed to ask developers about it=)
    For my opinion in order of importance :
    1) Improve performance in big scenes and in high poly meshes in edit mode .
    2) sculpt mode ( topological brushes, better masking , hiding geometry by face maps or something similar ). I have created post some time ago on right click select (and it got some votes)
    3) Paint mode ( at least layer system, and it already will be awesome )

    I agree with dorian borremans about ” Parralax for eevee would be nice”

    Maybe new code Quest for sculpt and paint tools , I think people will support the initiative.
    Some time ago Zacharias Reinhardt proposed it in his video.
    Any way thanks again to developers , also thanks to Pablo Vazquez and his blender today live , we watch you) keep doing it 🙂

  86. 1. I think it is very important to make volumetric shapes work somehow properly in Eevee.

    2. Lens flares?

    3. All nodes working to be able to create more poweful fx with particles and dynamics.

    4. A way to emulate caustics and transparent colored shadows in Eevee?

  87. The Awesome Developer Team In The World …thank you for great new and powerful blender project

    > im ask about camera features like cycles camera panoramic (Equirectangular , Fishyeye and Mirror
    Ball)
    when to put it in blender 2.8 ( EEVEE ) . im wait it from the first build version 🙂
    > when i try to render interior scene by 1600*1200 get empty window after make some calculations
    and when i reduce it to very small res get the image but without rug particles hair i get the resolution
    and particles hair when i render by OpenGl
    > also the velvet shader not exist in material shaders in EEVEE
    > i want to suggest new feature if you let me … i suggest to ability to rescale any object by fixed value in
    in individual axis and all direction … this feature very very useful – i found it in sketchup
    thank you and regards

  88. 1. Sculpt tools refinement please!, and better performance of high poly models
    2. Rework UI. UI changes in 2.8 were good but arent all for the better, ie empty topbar and extreme right aligned elements. Nested/grouped sculpting brushes are annoying. Tools like inset or loopcut using right click feel counter intuitive.
    3. Nurbs or T splines.. I can dream.

  89. First, please fix the intel issues where it crashes whenever you close the preferences or resize any windows.