Blender 3.x roadmap

In August of 2000, Blender 2.0 was released at the SIGGRAPH show. That means that the Blender 2 series has been running for a little over two decades! We certainly don’t want the 3 series to take that long. Starting with Blender 3.0 a new version numbering convention will be used, with a major release planned each two years. According to the new planning, in the coming two years eight minor Blender 3.x releases will be made, of which two of these as Long Time Support (LTS) versions.

This article aims to provide an outline for the Blender 3.x roadmap. It has been reviewed with several of the core contributors already. Feedback is very welcome as usual, this will be a living document for a while.

Blender 3.0 Beta
Blender 3.0 Beta

General focus: building on 2.8 UI and workflow, and improving it further.

The general guideline will be to keep Blender functionally compatible with 2.8x and later. Existing workflows or usability habits shouldn’t be broken without good reasons – with general agreement and clearly communicated in advance.

Before 3.0 gets released, all module teams will check on reviewing ongoing implementation and workflows, and come with design docs with requirements, outlining what we accept to be revised, and what we stay away from now. 

The module teams should make clear which changes will be happening, what the user benefits are, how we will keep compatibility of previously saved work, and (last but not least) how to get involved as a contributor. 

Most areas in Blender are quite stable, but in some areas bigger changes are being expected: for physics, ‘everything nodes’, sculpting/painting, texturing, and character rigging. None of these changes will be violating the roadmap as outlined for 2.8 though.

Core

The core module will be empowered to manage code standards and engineering practices everywhere in the Blender code more strictly (please write docs and provide tests!). Ongoing improvements of architecture and code will continue, aiming at better modularity and performances.

Anything that affects core Blender functionality such as ID management, Blender files, DNA data design, Python API, undo, dependency graph, overrides and APIs in general is meant to get good specs and functional docs, for contributors to know how to use it efficiently. No commit to this module will be allowed without review of Core Module owners.

Python scripts and Add-ons

The Python module is committed to keep the API work and compatible for all of the 3.x series. Some breaking changes to the API are inevitable, these will always be communicated at a minimum of 6 weeks before a release will happen at the Python release log page. The biggest change in one of the 3.x releases is that BGL will be entirely deprecated and replaced by the GPU module.

Modeling

Modeling tools in Blender will be maintained and keep working compatible. Speedup for managing massive datasets (large scenes or large models) remains a key topic for more development.

Sculpting / Painting

Currently a proposal for a hybrid sculpting/painting workflow is under review. This would eliminate the need for multires, and introduce a novel approach to combine traditional (triangle offset) sculpting with shader-based texture displacement. The benefit would be to achieve extreme detailed resolution, without need for massive polygon datasets, memory use and giant files.

Related to the proposal is to make Blender’s current editmodes atomic and flexible, allowing tool designers to combine multiple editmodes together for more efficient workflows.

Final decisions on the proposal and editmodes are pending a design and review workshop with contributors, which is tba.

Texturing

Blender’s procedural texturing system is in urgent need to be upgraded. Modern workflows offer node-based procedural textures that can be layered to do something similar to image textures – or better. In Blender we can do this by fully integrated support for these textures inside Eevee, Cycles, viewport drawing and painting tools.

Design is an open topic still.

Animation Tools

Character rigging is also in need of a serious upgrade. The ‘animation 2020’ project was postponed due to circumstances, but in ‘22 it should come back on the agenda. 

Keywords are: node based, layering, debugging tools, speed, crowds, mocap support, automated systems for muscles and physics.

Blender Institute will lead this project by involving industry veterans for design and implementation. The new rigging system is expected to use similar concepts for bones or poses, but it is not expected to be feature compatible. For that reason we expect old rigs and new rigs to co-exist for a while.

UI/UIX

The UI team will be reviewing the 2.8 workflow (toolbars, shortcuts, layouts), and come with a design doc to provide focus for developers and guidelines for everyone to understand in what way to contribute to sensible high quality and consistent UIs in Blender.

The UI module is currently in need of senior UIX engineers and designers. Until these roles have been assigned, UI module team members will mostly assist the other modules in keeping their UIs and workflow well balanced and consistent.

Assets

Asset browsing and managing is one of the big new 3.0 features. Work on this area in Blender will continue for a while. The expectation is that good asset tools in Blender will help configuring efficient workflows; for expert artists in productions and beginners alike. Part of the Asset and UI project is to publish (a series of) Blender Bundles; relatively large downloads (1+ GB) with presets, primitives and asset libraries.

Blender 101 (Application templates)

Being able to configure Blender for personal workflows is a key Blender target, part of the 2.8 project. 

To bring configuring even one step further, .blend files can be used to replace the standard startup file – creating ‘custom versions’ of Blender this way. These templates will allow to completely configure UI layouts, define custom keymaps and enable add-ons to run at startup. A smart Python scripter can design complete new applications this way, using vanilla Blender releases. As prototypes the team will release a couple of simple templates: a video-player review tool and the 100% dumbed down ‘Monkey Blender’.

A template used this way will be called “Blender Application Template” or just “Blender App”. Planning is to have this working as beta for Blender 3.1, after extensive testing and reviews by contributors and module teams.

Cycles

Cycles is now 10 years old, and had a big rewrite in the last 6 months, the Cycles X project. This included an essential architecture redesign, leading to significantly improved GPU rendering performance and interactivity in Blender 3.0. Following Blender release will build on this, adding new Cycles production features and improving handling of complex scenes.

Eevee

Our high-quality real-time 3D rendering engine is being upgraded as well. The new design brings screen space global illumination, more efficient shading, improved volume rendering, panoramic cameras, and tight grease pencil integration. It also paves the way for GPU ray-tracing through Vulkan.

A real-time compositing system is also planned, bringing compositing nodes into the 3D viewport. The new Eevee will be designed to efficiently output render passes and layers for interactive compositing, and other renderers will be able to plug into this system too.

Compositing

As promised at the 2019 Blender Conference, compositing is getting attention from developers again. The main goal is to keep things stable and solve performance bottlenecks. New is support for a stable coordinate system using a canvas principle.

VSE

The Blender Video editor is getting more developer attention, especially to improve the UI and usability. Noteworthy are the new preview images in strips. The VSE module will spend time on getting the roadmap and requirements clear, to enable the community to align with generally agreed design principles for VSE to contribute efficiently.

Viewport

The 3D viewport in Blender is where the action is; the 2.8 strategy to move more tools to 3D (not as buttons or panels, but by carefully designed 3D widgets or facegroups) will be ongoing. 

A big target for the 3.x series is to move Blender to entirely use the Vulkan open standard. Vulkan and Metal backends for Blender’s GPU API are being developed. We expect these to be ready to replace the OpenGL backend by the end of 2022.

CPU/GPU support

Blender now is being supported by all major silicon manufacturers: Intel, AMD, Nvidia and Apple. The developers closely work with the hardware industry to ensure an equal and fully compatible experience for artists who use Blender, disregarding the platform or operating system. Everyone can monitor or submit performance reports on Blender’s Open Data website, which will be updated when 3.0 is out.

AR/VR

Blender supports a real-time stereo viewport, for using the industry standard OpenXR library to communicate with VR devices. Support for VR controllers has been added for 3.0. A possible project for 3.x is to add real-time camera tracking for AR and virtual sets.

Everything Nodes

The Geometry Nodes of 2.9x are a big success, and can be seen as proof of concept for ‘everything nodes’ in Blender. Next on the planning is to introduce solvers as nodes, add point-based nodes (particles) and nodes for physics simulations. 

Hair systems are a separate topic; here simulation is secondary to artistic control and tools. Specs and design for a node based hair system is on the todo.
Object level (animation) nodes is an open topic as well.

Physics

Blender 3.x will keep using the old physics systems such as for Bullet, and modifiers for Mantaflow, Softbody and Cloth. An OpenVDB modifier will be added. For the rest, new development will go to Node based Solvers. See previous point.

Real-time mode

Physics simulation and VR viewing already require a new concept in Blender – “time”. Continuous passing of time can be treated independently of frames output for rendering or animation. Blender needs a formal design concept for this; enabling all tools and options for Blender to always work, whether playing in real-time or still. Design challenges are related to animation playback, dependencies, caching and optimization. Ideally the real-time viewport could be used for prototyping and designing real-time environments or experiences. Design topic: a node based editor for events, logic and states.

Grease Pencil

Gpencil projects for 3.x are: storyboarding workflow using Sequencer strips editing, better support for Eevee, asset browser support.
With Gpencil being used by the industry a lot, also for final rendering, we expect increasing development attention to this area.

NPR

Line art will get further performance tweaks and features. While speed is a major focus, also generating lines of higher quality will be on the roadmap.

Production pipelines and IO

Blender will remain committed to other industry standards such as OpenVDB, OpenEXR, OpenTimeline, OpenColorIO, OpenSubdiv, and OpenShadingLanguage. Support for MaterialX is on the list to be discussed.

Blender 3.x will see further integration with USD, also to enable hybrid pipelines. More (design and code) contributions are going to be needed by studios from the industry.

Work on a complete and 100% free/open source pipeline will continue as well, coordinated by the team in Blender Studio. Challenge for the coming years remains to ensure that individuals and small teams can create & manage complex industry grade 3D media projects using Blender and other FLOSS tools.

Blender and internet

A core principle of Blender is that it does not require the internet to work. The Blender developers remain strongly committed to providing a complete offline user experience. Optionally, however, Blender should be able to connect to the web for additional features to work. For example: signaling a user of new releases, updating add-ons, browsing asset repositories, sharing data with others or set-up collaborative environments.

Blender will start a new module for this (name to be defined, ‘meta’ is getting too many different meanings!). Module ownership and membership will be managed openly and accessible, making sure that only things will be added with general consensus that directly benefit users. 

A dedicated website and server on blender.org will be set up for proof of concepts and rolling out online features. By principle and by design only free & open services for everyone will be added here (only requiring an account to login). That means that possible commercial add-ons (like for storage, versioning, libraries) will need to be provided by external and independent vendors. The module can provide APIs for it, and help keep this safe and stable.

After 3.0 is out, Blender Foundation will set up this project at blender.org to start initial design discussions and to recruit contributor members.

Research

As a free and open source project we also need to keep looking for topics that will define the future of 3D tools. We especially welcome contributions from research labs or universities in this area.

A big topic we need to tackle once is AI tools – useful ‘assistants’ to speedup repetitive tasks in the creation process.

Mission

In the last year Blender worked on defining a clear mission statement. It’s what drives the blender.org projects and keeps the focus:

We make the world’s best 3D technology available as open source tools for artists.

Or in short: we defend and further The Freedom to Create.

Exciting times are ahead!

Thanks for reading,

Amsterdam October 2021
Ton Roosendaal, Blender Foundation chairman.

113 comments
  1. How about screen space bevel for eevee?
    Or could there be parallax occlusion mapping with beveling for eevee?
    Caustics for eevee? Realtime global illumination? Ray tracing support for eevee?

    And how about those path guide caustics will it support volumetric caustics?
    How about deformation motion blur support for openvdb data – mantaflow doesn’t support defromation motion blur when using openvdb and deformation motion blur is supported only for liquid when using uni cache – but fire and smoke doesn’t look to have deformation motion blur support and openvdb files doesn’t seem to have deformation motion blur either.

    And is there any plans for gpu accelerated simulations in future – like mantaflow with gpu acceleration. Granulars/sand? Multiple liquid setups with different viscosity and density setups with different shaders and possibility to mix liquids and their shader data (like colors – oil color, water color etc. and making lighter liquid with lower density to flow over the heavier / high density liquid.

  2. >To bring configuring even one step further, .blend files can be used to replace the standard startup file – creating ‘custom versions’ of Blender this way. These templates will allow to completely configure UI layouts, define custom keymaps and enable add-ons to run at startup.

    These should not be “blend” files, the name should be different to distinguish them.
    Blend file – is a source you can trust. You can’t trust files that changes your application.

  3. Please consider leaving the OpenCL rendering support for Cycles in Blender 3.X UNTIL full Cycles Vulkan is an option, since users with cards like the Vega64 aren’t RDNA cards.. and I have an AMD 3950X CPU, so I can’t use the other rendering options for Cycles either in the options menu. My only option for 3.x seems to be EEVEE, which is a great option, unless I just don’t use 3.x and only use a version with OpenCL. OpenCL rendering may be dated or not updating fast enough for Blender’s liking, but it DOES work for now, even if you don’t update it. The decision to remove it seems like gatekeeping based on newer graphics hardware, which is insanely expensive right now.

    • To quote the release notes: “The combination of the limited Cycles kernel implementation, driver bugs, and stalled OpenCL standard has made maintenance too difficult.” But there is good news too: “We are working with AMD to add support for Linux and investigate earlier generation graphics cards, for the Blender 3.1 release.” Source: https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycles

  4. Are more modifiers going to be implemented into Geometry Nodes?
    Bevel, solidify and displace modifiers are I would want the most using nodes.

  5. Animation Layers and Viewport Speedup for complex modeles, now that’s what I came to hear. That’s It! I hope It comes out soon, I am freaking hyped for that!

  6. “customizable rounded corners and margin” and all that thumbnail stuff.. blender is now complete. Who cares if blender forgets to save the painted textures or the linked image.. it is probably a feature that it feels sudden and like a bug when baking happens to work for one time… great priorities!

    • Blender don’t forget – user does.
      This is a feature that allow to keep massive project files clean during intensive teamwork.

  7. Great features! But what about the UV Editor? in the comparison to other DC applications the currents state of the UV Editor is very poor. Is there some plans to improve it?

  8. I will hope that the scant mention of NPR is more than a footnote. Freestyle needs to be updated to work closer to “realtime” with more attention to “artistic” line flow. Having lines respond or adjust to object flow without undue ‘tweaking.’ If Grease Pencil is the standin for NPR work, it needs more line quality modifiers. Several years ago I contacted Pencil+ but they remain firmly planted in the Maya/3DMax world with no plans to port their tools to any other platform. I also contacted the developers of the Watercolor/Line tool MNPRX, but they too had no plans to port over. These are creative tools that need sponsorship in the Blender Community. There should be a movie project that uses this visual language to tell a story. All else it excellent work, and my comments are not here to detract from the enormous effort and fine work of the Blender Team, just an appeal from a different quadrant. Works by Miyazaki and Chomet and Paperman remain benchmarks for my reference.

    • The comparison of naming something like Line tool MNPRX seems somewhat unfair to me, which is a kind of addon paid for 75-100 dollars. It is a miracle that Blender has tools to make NPR for free and that they are part of the same program without being a separate addon. One can ask for things but taking into account the correct comparison and the context. I start reading the comments and 99% ask for different things each with their own priorities, some even with a certain anger that is not justified, and I swear it is depressing.

      Blender grows huge every year and some people never get enough of it. For my part, I only ask that the program continue to improve at the rate it has been doing with each version. Thanks Blender for being so big and being available to everyone, in a world where it is increasingly common to see programs that try to get money each month.

  9. I’ll be another to say it, but the animation tools and workflow are and have been seriously lacking. I love the idea of blender in principle but as an animator in the industry the animation tools, particularly the graph editor are well beyond 20 years out of date. I tried opening up blender and animating but once I got to the clean up phase it was too frustrating for me to continue. I’d be more than happy to donate or buy a shirt if I new the money would help update those areas, but at the moment Blender seems to be more focused on gimmicks and being a jack of all trades while a master of none.

  10. Great roadmap, but I personally think that there are a couple of “core” areas that would desperately need some improvments.
    First and foremost UVing in Blender is still a huge pain in the a** compared to 3dsMax, Maya or let alone RizomUV.
    The controls need basically a complete overhaul (UV-UI is terrible, you’re searching for features all the time – especially if not used to Blender’s UVing), simple features as “align elements” “straighten elements” etc all have to be achieved via third-party plugins if you want them to work properly and easily accessible. The relax-allgorithm in use gives average (at max) results thus your packing will also only be “somehow useable” compared to a clean and nice unwrap from other software.

    Rigging / Armatures are also super frickly compared to other software.
    Especially weight painting is pure horror with the current controls and behaviours.

    Import/Export (especially with fbx) is also a big pain when working with different tools (as for e.g. Zbrush). Scaling issues are always prominent.

    Outliner – Layer (Collection-Management)
    Selecting a complete hierarchy when selecting the parent, more info on whether objects are linked or not, etc all that should be normalized to industry standards since right now they’re causing quite some headache.

    Also, since it is getting more and more important in character art: improvement of the hair system would be great. while not too bad to start with, Blender’s hair system still lacks behind by far in comparison to XGEN, Yeti or others.

    In general I think focusing on improving core features before adding new “fun” stuff and then continue improving those should be a priority.

    Still looking forward to Blender 3.0! It is on the right path to get on top of the 3d market, but, as said, it needs to get it’s core features to work properly FIRST.

  11. How about adding a Gantt chart widget?
    This is useful for production management and even 4D/5D BIM features in BlenderBIM.
    https://www.autodesk.com/products/shotgrid/features
    https://github.com/IfcOpenShell/IfcOpenShell/discussions/1779

  12. In the future, I expect the interaction of the workflow of creating characters and environments, creating and interacting with real AI Artificial Intelligence (Quantum Electronic Human Mind) . Connection to the development and improvement of 3D programs , and quantum Artificial Intelligence through cloud connections . Creating hyperrealistic worlds based on quantum computing .

  13. I love everything that you guys did

  14. Yeah it’s amazing

  15. Is this really a list for the next two years? It looks like work is planned that easily take the next 20 years.

    But maybe there are non-published branches in git where 90% of the planned features is already done, and only needs the next two years to implement the missing 10%. Let me dream!

  16. improving of Storyboarding is great but more preproduction features are also needed such as Screenwriting workspace, Script breakdown workspace, Shot blocking/Previz workspace, etc…

  17. I mainly wish the fbx pipeline for rigged characters was improved, I can barely get working rigs imported into unreal in any manageable way. The scale is always forced to be exported at 100x and other madness is always going on with the bones.

    I’ve looked into many other ways of importing rigs but unfortunately the pipelines are all built around fbx

  18. please insert 360 degress camera into eevee for vr renderings

    • Yep, panoramic render fot Eevee would be great for people doing VR. 15000×75000 pictures are so long to render with Cycles, even with CyclesX. Eevee would help so much….

  19. Are there any plans on tackling long-standing bugs / issues within blender? There are some outstanding issues that have been confirmed, but not tackled for a year or two.

  20. What is needed is a deep core rewrite to handle more data and future proof designs.
    A BlenderCoreX basically…. You understood the need of such a rewrite for the renderer, time to understand that Blender itself needs the same thing desperately now…

    • “What is needed is a deep core rewrite . . . . You understood the need of such a rewrite for the renderer, time to understand that Blender itself needs the same thing desperately now…”

      I disagree completely. The reason Blender is so fast right now is that the core was written at a time when programmers knew how to write fast programs that are easy to read, maintain, and debug. Blender now has a simple, clean architecture that doesn’t get bogged down in horrible, slow, buggy nonsense from the likes of poorly designed languages and external libraries. One of the worst things that could happen to Blender is to rewrite the core with anything remotely resembling “modern” or “future proof” code.

      What is so “desperate” about the need to rewrite? Why is it even needed in the first place? Where are data bottlenecks? What does it mean to handle a “future proof design,” and what leads you to believe Blender can’t be future-proof already?

      Seriously, just look at how advanced and flexible Blender is after decades of open-source development.

      _If anything, every software company in the world should model their flagship programs on the current core of Blender._

      • Neither agree or disagree with rewrite, but I disagree with simplicity being always better. Some things are inherently complex, and user’s growing needs add to that. Simplicity for the sake of simplicity might not solve a problem well.

        Rewriting the core in JavaScript, that would be a bad idea.

  21. > We especially welcome contributions from research labs or universities in this area.

    What a change of heart…

  22. Something I would love to see is the ability of multiple people working in a blender environment at the same time, realtime. Being able to see their viewport camera like avatars in a multiplayer game. then you maybe can slave yourself to their camera view so you can see in Realtime what they are looking at, and maybe use greasepencil to make notes and give feedback Realtime and be able to witness whatever they are doing. You might be working on modeling a house while another user is constructing a forest or modeling something else and then being able to bring it together.

    It would remove the need to wrangle with version control software to manage work on the same scene going back and forth, back and forth.

    Just a thought.

    I am exited to see where blender is at at the end of 2022 and later. This package is getting better by the day and I love that I can do almost everything within a single program and it does each thing to a good enough standard which for 90% of the cases is exactly what you need.

    I would like to see a bit better physics simulations through. Cloth sim is pretty good but it could be muuuuch better. First off I would love it if that would be faster, additionally I would love to be able to drive more permanent creases trough animation, so for example you create clothing for your character and then you let it run trough a running animation on your character to receive creases just like real cloth would get since real cloth has a bit of a memory property that is often ignored in simulations which would create more believable folds in your cloth.
    Would also be great to be able to manipulate your cloth sim semi Realtime like marvelous designer, where I can pick up corners, and drag them around allowing me to loosen up problem areas or manipulate my cloth sim as I want it to be. Or maybe allow me to solidify simulated parts

    anyway, enough ranting from my side. exiting stuff!

    • I remember Ubisoft has been developing such a tool for inhouse use and planned to release it.
      Not sure if it covers all your needs.

      I am not sure if I’m allowed to post links here but it’s called “Mixer” from Ubisoft, they have a github repository available.

    • Mixer is a Blender addon for real time collaboration in Blender. It allows multiple Blender users to work on the same scene at the same time and is developed by the R&D department of Ubisoft Animation Studio.

      https://github.com/ubisoft/mixer

  23. Simple selection fixes please.
    If two verts are co planar please make drag select be able to select both for easy welding.
    Please make it that I can select verts etc on the other side of an object without having to go into XRay mode, adjust the opacity to 1, do my drag select, turn the opacity back. switch out of Xray mode.
    It may not sound like much, but every other software can do it as default behaviour. If you don’t want to select the other side of an object, that is what back face cull is for!

  24. Generating Ragdoll Physics puppets would be interesting with geometry nodes – generating puppets like particles and throw them into something. Everything nodes will support rigging, particles and physics i guess so something like puppet spagetti would be possible in future? Is there ragdoll physics considered? And how about dynamic mesh generation/modification (edit vertex index numbering/order)? Or generate motion vectors for motion blur? (at least fire and liquid should be supported in deformed motion blur level with openvdb – without using uni chace but also possibility to generate motion vectors and support motion blur to third part addons should be supported – also for 3rd party render engines – this could include vertex motion / velocity data) Also missing this: “lagging”/”lazy” parent feature – like spring/bouncy/wiggle bones that could be used to “simulate” elastic things like tentacles.

  25. Before updating anything relating to Animation tools or Character rigging please consider overhauling the armature and bones system first.

    I cannot articulate the misery of so many years spent with bones that only/always/must point down their own +Y axis. Interop with other packages becomes a total no-go for me with anything that must be rigged. As a tech artist the conversation with artist simply goes like this:
    “Will it need rigging?”
    “Yes? Don’t make it in Blender.”
    “No? Use whatever DCC you like, have you tried Blender? It’s great!”

  26. This is a bit much to ask for… but I really wish the UI was waaay more user friendly. Maybe redone. For one, there are a million tools dispersed in hard to find places if you don’t already know where they are. And then there’s the button mapping… it’s impossible to remap buttons because there may well be thousands of button mappings. Want to switch 1 button? Well good luck, it will almost definitely conflict with another.

    A single button should map to as many different but related functions as possible depending on context. Not jamming in a new ctr-shift-alt-space-Q to extrude a point or whatever.

  27. Why is the proprietary metal API being added? This would be like if Blender added DirectX support. Stick with the open source Vulcan backend.

    • Vulkan is not supported by Apple sadly. If you want to have a low level graphics and compute API on Apple hardware you are forced to use Metal API. On Windows we have Vulkan as open source alternative to DX12. You can make Vulkan apps work on Apple by using moltenVK though but it looks like its not popular.

      • False.
        Valve bought MoltenVk that is a Vulkan to Metal driver, to make it open source and freely available.
        Since the 2 APIs are very similar, performance is very good.

        Blender for Mac could totally be made with cross-platform Vulkan + MoltenVk.
        The only reason why Apple sponsors Blender dev is that they don’t want people to witness a famous proof of concept where skipping Metal and for a Vulkan + MoltenVk solution works very well.

    • Probably for the same reasons CUDA and everything NVidia is favored. These companies are funding and dedicating their own developers time, support, etc while the lands of OpenCL didn’t to name one example…
      I’m not quite against it if it isn’t halting the Blender devs at all either, except review and merging process. The massive performance improvements that the image denoisers, RTX backed accelerations, etc made waves and helped Blender get more of the spotlight.

  28. That is a great roadmap, great vision, but I think the general motto of the next blender versions should be – “Groundbreaking view-port performance and optimization, ability to work with limitless amount of objects in the scene. Infinite creativity”.

    I think the team should allocate way more resources on performance and optimization rather than on new features. Blender is quite feature rich already and is perfect for small to mid production purposes, what it requires is a huge operation speed improvement. Some optimization work done for object and edit modes in 3.0 is a great start, but not quite ambitious and could be improved in further iterations of blender, IMO, this should be a number one priority.

    Blender has so many features to create big, creatively complex projects, but lacks horsepower in viewport department.

    • I agree, at the moment blender is great for small to medium productions, but as soon as you scale things up to a professional level, things start to fall apart pretty fast. the amount of hard crashes you get as you try to work with larger scenes that would work effortlessly in any other professional 3d applications.

    • I totally second that! It couldn’t have been said better! This needs to be a top priority IMHO!
      Please check this video and the ensuing comments :
      https://www.youtube.com/watch?v=B79bGW7F8N0

      The design/dev team should really check out where other 3D packages are at regarding instances performance. I’m thinking about Mercenaries Guerilla that can handle millions of instances of a single instance node : http://guerillarender.com/doc/2.0/User%20Guide_Procedurals_Instances.html
      I think they are achieving that by relying heavily on the Alembic format and topology matching
      Blender needs that kind of performance !!!!!!!!!!!

  29. For god sake , please make displacement modifier work with procedural textures , cycles displace node is too god and working but not anybody is working on rendering and animation , someones using blender for 3d printing , also please extend geometry nodes the way that it can be used for more purposes such as parametric modeling and more understandable like sorcar addon, since right now Geo nodes is using for distribution only .

  30. Awesome!!

  31. “So many things”…
    If the developers do not change the development cycle from 3 to 6 months, then we will continue to receive unfinished features.

  32. Please let eevee become lumen or better, the world needs real realtime!

  33. Some OCIO nodes in the compositor is something I’ve been waiting for a long time.
    Currently, the view and the output are tied together in the color management settings, which is kind of convenient and easy for newcomers, but lacks flexibility and hinders some workflows.
    Having the ability to override the current color management system and chain color transforms in the Compositor would provide a solid an powerful way to control colour in the output.

    As an example, here are some of the things that are currently not possible within the compositor:
    – freedom to load external OCIO configurations to allow extra color transforms on the fly.
    – “bake” transforms (like cm does currently when exporting to display referred formats, can be also useful if you need display linear with filmic-blender baked in an EXR, for instance)
    – ability to chain colour transforms, like going from the linear reference to some log/perceptual transfer, apply some LUT designed for that transfer and flip back to display linear. Useful in color grading, as grading linear can be difficult in some portions of the dynamic range of the image.
    – Properly manage external assets that might not be in a colorspace supported by the included OCIO configuration.
    – Control to produce specific display-referred elements that have to be composited over scene-referred CG (titles, logos, etc. that are tricky to composite in pure scene-referred workflows).

    Addressing these issues would make producing display-referred images more flexible and robust, specially for people who want to use Blender end-to-end to produce final pieces.
    That flexibility would be also welcome by people who have Blender as part of their pipeline with other software, as it would give them fine-tuning for EXRs, tiff, and dpx, etc.

  34. eevee panoramic!

    cant wait to render my 360 vr video in eevee!

  35. Really excited to see the compositor improvements! Thanks for all you do.

  36. 1) I wait Vray for Blender. 2) I need edit spline and edit poly modifiers. Please

  37. Have Blender foundation asked if michal1000w would be interested to work for blender foundation?
    https://devtalk.blender.org/t/realtime-gpu-smoke-simulation/13142/43

    And have blender development team considered other simulation engines’ integration like
    https://github.com/doyubkim/fluid-engine-dev
    https://dual.sphysics.org/
    https://projectchrono.org/
    https://github.com/taichi-dev/taichi_elements
    or will these “Everything Nodes
    The Geometry Nodes of 2.9x are a big success, and can be seen as proof of concept for ‘everything nodes’ in Blender. Next on the planning is to introduce solvers as nodes, add point-based nodes (particles) and nodes for physics simulations. ” be new solvers?

    Granular(sand,snow) liquids with multiple meshes and shaders? (different colors? viscosity and density?) and possibility to mix these different density liquids and possibility to mix colors from different liquid sources? mix them with sand/snow? make them wet?(wetmaps? drying?) possibility to make bubbles with gas solver and liquid solver mix?, crowd simulaitons (cars, birds, insects, humans – fight and run, escape etc)? muscles? Fracture modifier? and possibility to fracture dynamic objects (like walking character fracturing while its walking)?

    How about caustics? or eevee’s global illumination that would be based on similar technique as unreal’s new Lumen? Or rendering this “cctv” effect (rendering camera view in shader – shader that could render selected camera’s view).

    Motion Graphics kind of tools for everything nodes – falloff based animations/position/Inheritance effects? Follow object’s original position for simulations (how strong it will follow original position and will simulation be affected only for rotation or position / possibility to control what key frame channels are affected by dynamics/simulation and which are controlled by animation key frames and how strong).

    Wrangles: stress map/stretch map/tension map. Velocity weight map (how fast vertices and edges and faces are moving).

    I also wonder could 2d-views/animations be projected as shaders – and controlled by drivers (for example: to make 2d face expressions in 2d mode and then project them into 3d polygon object? and controlled with driver objects). This would make something like lego face animations easier.
    And this makes me think: could texture editor have same tools as 2d animation “mode”/”layout” – animating vectors in “image editor” / “uv editor” would be very handy / flexible thing.
    I think possibilities to use these also for displacement kind of things which reminds me that will there be any parallax occlusion kind of displacement in eevee?

    • Or how about that “hybrid sculpting/painting workflow”
      could it be used for dynamic paint modifier also (and will there be dynamic paint in geometry nodes).
      And how about dynamic fracturing that could fracture liquid mesh (form the end of liquid tail) will these kind of things be possible in everything/function/particles and physics nodes?

      And how about manifold / destructive extrude – will it be fixed? Developed more? Or topology tools?
      Or multi-threading for modifier? Or GPU based sculpt – simulation (cloth, liquid, rigid) – will it be focused in the future – even in long time period? I am not trying to hurry these things just wondering are these considered?

    • I agree with Juha. Just the other day I ran into the limitation of the fluid simulations having to be in multiple domains for their colors to be changed. Even then, the fluids cannot interact. Please please please Blender, put your effort into the Physics system. It is lacking essential features.

  38. Good plan!
    But the animation tools are really really in a state of absolute emergency

  39. Easy use of multi language text part from English, will be very much useful, thanks

  40. Thank you developers for your hard work! Dummies like me are very grateful for your continued improvement and your integrity making this software better and better. I’m very appreciative of you!

  41. Great roadmap.
    I wish for a nodal rendering workflow, where instead of a monolithic user interface properties editor setup, we would have a new rendering editor with:
    -a scene input node (where we choose what from the outliner we want to import)
    -a renderer properties node (engine choice, sampling settings, camera, AOVs…)
    -an output properties node (folder, file names, framerange…)
    -an object properties node (visibility types, custom render overrides…)
    -a material assign node
    -a collection node (to create groups of procedurally selected goes)
    -some merge and isolate nodes
    -expression node to drive attributes through python/osl.

    …Cf: the AMAZING Gaffer (https://www.gafferhq.org/), the not too bad Katana.

    • An addon call renderstacknode is somthing that match your descriotion

      • Thanks for the suggestion but I don’t base projects on addons from I don’t know who.

        It needs to be a Blender officially supported feature and very optimized and polished.

        If a proof of concept can be made by a solo python coder, I have no doubt Blender can officially craft something similar natively.

  42. I am new to blender I’m 71 years old I’m just now learning how to use blender and I find this program amazing thank you to all you who contribute and volunteer their time to this wonderful program without you it would not exist I thank you from the bottom of my heart for your time in contributing to everything that blender is keep up the good work have a fantastic day and please remember to StayWell

  43. I’m definitely glad to see a lot of these features. Though there should be a few more things I think. For animation tools, the addition of proper animation layers and keyframe layers is long over due.
    For cycles, I honestly think Blender should drop OpenCL support in favor of Vulkan compute shaders. Vulkan implementations have been pretty consistent across platforms where OpenCL for years has been the complete inverse and nothing short of a headache.
    For texturing and painting, Blender feels like it’s been only a few steps away from having some Substance designer/painter like features for a while now. It’d be really cool if it got the same love the sculpting tools have.
    For geometry nodes, I feel a bit awkward towards this. I like the idea, but it’s not even remotely gonna replace Houdini for me, and the current implementation feels like someone just cut a sliver of features out of Softimage ICE. I dont mean to rag on a brand new feature, it just lacks the intuitiveness competing tools have.
    But honestly before anymore features, just fix bugs please! The Ctrl+Z bug haunts me like it’s the grim reaper.

    To end on a positive note, Blender is absolutely amazing though. Thankyou everyone who’s worked on it! :)

  44. It looks like the animation & rigging project keeps getting pushed every year and there is no real concrete plan,just the same wishful stuff repeated, basically you guys have missed a golden opportunity.
    While other companies are making huge strides like Maya & Houdini, it would take even another decade just to catch up and establish a good foundation with new innovation & improvements , most of the animation community don’t even give a blender a look besides the very few studios out there who are blender based ,which is a shame since there was some good potential for blender to take the lead.
    I guess things have to be dragged for so long before anything can be really accomplished.

    • Animation layers are the thing i wish blender would implement like Maya… or improve the NLA system’s UX to be used for that.

    • After reading this post (and the “Asset Creation Workflow” post), I realized that I, as a rigger and animator, should just go back to Maya. I’m sure Blender is a fantastic tool for modelers and sculptors, but it’s quite clear that animation is not the priority for the devs.

      • Plz try making it require more low PC specs. No much ram or cup needed for fast and good animation

      • Ton: “Animation and rigging is priority in Blender 3.x”
        Random internet guy:”AfTeR ReAdInG tHIs POst, I ReAaliZed thAt I, As a RIGGger and AnImAtOr, shOUld just go back to MAYa. “

  45. Could you adapt Unreal’s Lumen lighting pipeline for Eevee?
    Could you improve and add new features to particles system ?

    • This us a very interesting idea. Wish someone will read this. How ever that also Jean blender gave to develop a game engine. To replace the one they canceled with 2.8

  46. Just a correction to my previous comment here, this is Parallax Occlusion Mapping, there’s already a patched version available for 2.91, it’s so close to done, someone just needs to get it merged into Blender properly:

    https://devtalk.blender.org/t/parallax-occlusion-mapping/15774

    • There’s even a patch on developer.blender.org, right here:
      https://developer.blender.org/D9792

      It just needed to be merged. I can’t understand why it wouldn’t be.

      • This patch was never finished, see the [WIP] (work in progress) in its title.

        • Finished or not, Parallax Occlusion Mapping must be a priority…

          Whats the point of eevee without it? eevee’s goal is “fast output + good quality”, POM is exactly for that… high details, fast rendering output…

          Its actually is THE feature that lets us achieve what eevee promised… there is literally no alternative for it.

          is anyone going to tackle this limitation?

        • Why was my reply deleted?

          Again Parallax Occlusion Mapping is a must have feature! finished or not, it should be the No1 priority for eevee…

          Eevee’s promise is “fast render + high quality” and it absolutely cant be done without Parallax Occlusion Mapping… there is literally no alternative for what POM provides the artist… High quality result in low time, which what eevee was promised to be, so shouldnt this be a must feature?

          no offence but do the developers know what Parallax Occlusion Mapping is for? if they did this would be implemented long time ago already.

  47. So nice to see all of these plans! But please please please work on the animation interface and workflow because right now it is put to shame by maya (which is an abomination of a software). I animated simple bouncing ball rigs recently and boy was it a journey. Pose mode needs to simply go away. The dope sheet is absolutely horrible and the graph needs a complete UI reboot.
    I love Blender with all my heart, been using it for 15 years now, and I have a dream that one day it can be used in animation industry-wide but some big changes need to happen for that! We’re moving in the right direction I think. I’ll keep supporting the Blender devs as best I can! You guys rock!

    • I’m only posing and animating simple rigs, but since I got the hang of it I really like pose mode :)
      It’s a distinct feature of Blender and works well.
      What’s your take on it? How would you make it simpler but still effective?

  48. You forgot to talk about Opensubdiv and a lot of features for eevee like Mesh shader ?

  49. While I love what the contributors have accompkished so far and am greatly anticipating some of those features I am disappointed with the lack of explicit mentions of baking and PBR texturing.

    Blenders current baking system is an absolute nightmare to use: you need to modify shaders, select stuff in the proper order and manually bake every different texture. Overall it feels like a hack that was added in as an afterthought instead of the core feature it is for everyone who wants to export their content to another tool like a game engine. There are a few plugins that aim to improve this, but none of them work 100% of the time and none have the clean ui and ux I have come to expect from blender. IMO this system needs to be redesigned from the grounds up.

    While I am happy that texturing was mentioned I think that the mentioned improvements are not enough. I am aware that I shouldn’t expect something compareable to Substance Painter, but better support for PBR materials would be great.

    Would greatly appreciate it if some dev could show these two areas a bit more love.

  50. Bravo! How exciting to see Blender progress, and the road map sure does look ambitious. With the increase in funding, support, development, etc. and Ton’s leadership I’m confident the team will reach their goals in 2022 and beyond; and as such I will continue to support Blender!

    Would very much like to see performance improvements with animation playback, compositing, and heavy datasets.

    One thing to note, with Facebook now changing their name to Meta perhaps Blender should re-consider that naming so to not associate with the evil empire.

    Looking forward to the premiere of Sprite Fright… it’s only hours away. :)

  51. I would like to see touch support since it seems like there’s already an effort to tackle UI.

    A large number of laptops have touch and being able to use it for demos, and more actively for sculpting and drawing would be great. I think this also relates to better track pad support.

  52. So happy to hear that panoramic camera support is finally coming to Eevee. It always felt strange that the option was there in 2.8x and 2.9x but it didn’t actually work as a panoramic camera. Rendering 360 degree videos will work so much faster with Eevee!

  53. I just wish the text object actually incorporates proper text shaping. Even if you’re not going to support colored fonts (eg. colored emojis), there are other areas that Blender Texts completely fail at.

    – Right to left
    – Vertical writing (which involves different positioning for some glyphs, not just rotating everything)
    – Ligatures

    This is something that competing software handle reasonably well. If a Text object type is going to be provided, it should not be a half-baked solution.

    I want to use Blender as much as reasonable, and not have to open Inkscape just to typeset some text properly.

  54. Thanks alot for Blender developers who are struggling hard to push our 3d software but if possible make hardhops & boxcutter default and improve on video editing part of sound effects it doesn’t edit audio like flangers, echo such effects are missing if it can please help me with that tutorial and stabilization cause its always crushing have been using it for two years by now great thanks 🙏 again

  55. caustics! Please implement caustics for cycles! thank you X 1000

  56. Bit confused by this:
    ‘Blender 3.x will keep using the old physics systems such as for Bullet, and modifiers for Mantaflow, Softbody and Cloth. An OpenVDB modifier will be added. For the rest, new development will go to Node based Solvers. See previous point.’

    Does that mean you are developing node based solvers for rigid body and softbody etc… or totally different solver types for other physics…if so what kind of physics?….Apart from particles, I can’t think of many physics sims that aren’t fluid, rigid or softbodies!

  57. Absolutely stunned by the outlook of rigging tools: everything nodes and muscle system would be the game changing features

  58. This is all very exciting !

    The “interactive mode” sounds like it could very well merge with the node-based solvers. After all, isn’t a solver a handful of rules that use the previous state to compute the next ?

    I’m also hoping for more work on gizmos, and hopefully gizmos tied to assets and node groups : linking a handle to a property to control asset variations from within the viewport, etc.

    Cheers !

  59. Forget everything else (not literarily) and just focus on these:

    Animation UI Overhaul
    Hair Grooming
    Cloth Modelling
    Multilayer Texture Painting

    You will have become the biggest 3D Software

  60. Wow, this is absolutely incredible!

    Still one more wish: Proper Lightlinking without the need of renderlayers.

    But everything else is everything I would ever want to use, thank you.

  61. I hope to prioritize the solution to blender has been the shortcomings of the broken (split)

  62. I can’t tell you how excited I am to follow this great progress being made! Blender has become essential to many of my various creative workflows over the past few years, and I’m blown away by every new update.

    Also, in all honesty, whenever I see anything to do with one or other of the commercial 3D applications, I’m always struck by about a dozen features that it lacks but are fundamental to Blender’s workflow. Usually the commercial softwares seem to specialize in one area but fall short in five others, whereas Blender is better overall—and indeed, Blender looks like it may soon catch up with those applications in almost all areas.

    Keep on protecting the freedom to create. I salute you!

  63. Blender professionalism is growing fast it’s very commendable. Keep it up and give.

  64. 100% amazing !!! I Love You Blender!

  65. It’s great to see Blender evolving, Tom! I believe in the progress ahead. I’m specially happy with Apple news, since I’m a 20+ Apple user!

    But I would love to see more precision modeling love from the developers. It was promised, it was shown a roadmap… and almost nothing happened. I understand Blender is huge, and there are important areas to develop. But precision is basic and, I’m sure, much simpler to develop…

    Anyway, I want to congratulate every developer helping us do our art better!

  66. One thing: Just please add the very usefull Fracture modifier !

  67. …tools, cloud generation, etc, like Houdini, is what I would love to see. :-(

  68. The best present for my birthday, thanks alot! :D
    But some things are missing: clear ACES support fro example. I’m aware of option to replace OCIO files in blender folder or declaring global environment variable, but abyway it’s little cumbersome to use %)

  69. – Well before sarting with some of this tasks, better make sure the long time not fixed Keymaps-editor will be right up front. Makes no sense to come up with Application Templates and this is still an issue.
    – Meta is good for online cloud rendering. Not on farms but like BlenderStudio over the network machines in the office etc.
    – Metal backends at the end of 2022 is way to long. Some told it’s coming to 3.1 – 3.2.
    (It can not be to get blender BEAUTY changes, but the whole apple blender community left out of GPU rendering for years)
    -Same for other fields like Caustics, light linking etc. Those things supposed to be basics for 3D, and only get progress in very small portions.

    Anyways, blender is a great idea, software and pice of hard work.
    But every year we learn and make something new, so why not fix the basics that a lying around for a longer while.

    Thanks a lot Ton Roosendaal

  70. Nice road map Ton and team.
    Can’t wait to see what the next 12 months bring the last 12 have been a fun ride with some solid improvements in most areas.
    Thanks guys!

  71. Epic. Just Epic.

  72. sounds very promising. hopefully enough developers are on board !
    can’t wait!

  73. Artistic control and tools are the most important for hair, well said. I’m glad this was acknowledged, I was getting worried that fancy hair generation with nodes was getting more attention than basic things like grooming tools, performance, and most importantly, the data structure and handling, ie hair objects.

  74. Re: everything nodes. Perhaps it would be worthwile to consider integrating (parts of) Sverchok into mainline Blender? A lot of work on modeling using nodes has already been done in that project.

    • 1000000% this!!!!!!!!!!!!!!!

      Also integrate animation nodes!

      We have 2 POWERFULL systems right here why not integrate them?!

  75. Really like the roadmap, but would be careful about taking on too much and end up not doing anything really well.

    I think blender needs to overhaul its internal management of staff to ensure pushing in the same direction on a consistent basis. Really should make it clear which developers will lead each of these individual 2022 objectives and then the weekly reports should measure progress against only these with an ‘other’ section for any work not covered by the goals

    Ps. Would also be great to have a ‘contributors day’ once a quarter were all developers review and give feedback on open patches submitted by the community

  76. About the “Meta” module, how about instead integrating some features for the web, why not do a full-port for the web?

    The web now has support for webAssembly and with tools like https://emscripten.org/ we can port OpenGL to WebGL.
    And with the move to the web, it would make blender way more accessible and would make features like real-time collaboration much easier.

    Some desktop applications have moved to the web, like Sketchup, AutoCAD, and recently PhotoShop (partnering with the chromium team)!

    Here is a link with more details https://web.dev/ps-on-the-web/

    Pushing Blender to the web would make it live for decades to come.
    There are already 3D design tools that are coming up that are based on the web https://spline.design/

    Blender should be a part of that space.

    • That’s an absolute waste of energy and resources. Blender is thriving and you want to knee cap it with a WebAssembly interface and kludge? Just download the app for your operating system of choice. It’s on every platform. Creative minds don’t need a multi-level abstracted browser platform to work under. Their work will be interoperable across all major OS platforms. Done.

      • Hmmmm, I guess you’re right Marc that it might halt development 🤔

        I was more talking about a long-term direction though, not 3.x specific 😅

        Is it on every OS? I don’t remember blender supporting ChromeOS, and won’t porting blender to the web make it OS-agnostic therefore reducing development time maintaining multiple OS distributions? And in theory Blender could also run on Android and ios 👀 (not sure if it’s a good idea though 😅)

        I guess creative don’t necessarily need their applications to run on the web.
        We could make the same case though for Google docs, Figma, etc.
        They had desktop counterparts that worked on “every platform” and their work was interepole between devices, So why bother right 🤷‍♂️?
        It’s weird though that the web version seem to have had more adoption 🤔

        So, while creatives don’t really need thier applications on the web, they might need to have thier work easily sharable through a link, invite another creative for collaboration, have thier files saved across devices safely on the cloud (maybe even auto versioned?), be able to continue thier work from another computer, have thier application auto-update without installers, or have shared preferences and configuration across devices.

        So I guess it’s not that much of an “absolute waste of time” XD

        Anyways, I wish all the best for Blender and the team, I know they have the community’s best interest at heart and they know what’s best for Blender more than I’ll ever know XD

    • this is impossible to do actually… webgl and webassembly cant handle it.
      Its not easy porting an app to the web, let alone blender, where its impossible…

  77. Fantastic plan, as someone working with Blender 6 days a week for the past 8 years, I’d say it’s very ambitious too, but you have all of 2022 to work on it, so lets see how much can happen!

    There’s too much here to respond to all of it, so here’s some key stuff:

    > Complete lack of mention of baking.

    Please don’t tell me baking is going to be forgotten *again.*
    Baking has been in desperate need of an overhaul for years and it’s repeatedly been mentioned as a target for an update and somehow been forgotten entirely.
    After Cycles X is done, I think baking should be a priority.
    A perfectly good overhaul for baking (D3203) was left to rot without a good conclusion all the way back in 2018 that would have been a major improvement for most users. The current baking system is almost unbearable to use.

    > Texturing
    > Blender’s procedural texturing system is in urgent need to be upgraded.

    Couldn’t agree more. There are only two things I have desperately wanted for years from any kind of overhaul to texturing, and they are:
    – The ability to “paint” with multiple textures onto multiple texture targets at the same time. (For example, painting the base/metallic/roughness/specular image data of a bolt onto a surface with base/metallic/roughness/spec image textures with a single click).
    – The ability to easily layer and mask materials on top of each other. Ideally using existing materials, not needing some kind of complex setup.

    If any overhaul to texturing could achieve that, I’d be extremely happy.

    > Cycles

    Job well done, you’re all magnificent. Cycles X is a major achievement, you should be all very proud.

    > Eevee

    Screenspace global illumination is going to be *game changing.* SSGI combined with baked probes will allow for very fast and very impressive real time GI. This is huge.

    However, I ask that the current patch being tossed around for real time eevee displacement mapping (or relief mapping perhaps it should be called) be forgotten about. That would be an incredibly useful thing to have for Eevee for creating stunning real time materials, without the cost of insanely high polycounts.

    > Real-time mode

    Speaking of real time modes.

    Could we please have some improved quality options for Eevee for 1 sample based renders? Such as options to control the number of per pixel samples for effects such as SSAO, SSR, SSGI. And post processing AA effects like TAA, FXAA. PCSS shadows would also be a blessing. Perhaps we could really get Eevee running at 60fps with enough fine tuning.

    As always, thankyou for your incredible work, 2022 is going to be a great year for Blender.

    • Nice feedback, and I’m totally agree with you about the baking part.

    • Also bring back baking to vertext colors. Baking AO to vertext colors is an extremely valuable tool.

  78. Holy mother of normals! This is absolutely fantastic!! The compositor, texturing, Eevee and node based solvers sound are probably the best things coming to blender and I REALLY cannot wait at all. So excited to see where blender is going. HUGE HUGE props to the developers and contributors that make blender, blender! Exciting times are ahead!

In order to prevent spam, comments are closed 15 days after the post is published.
Feel free to continue the conversation on the forums.