Blender 2.92
blender.org
blender.org
Our department (art and design) moved from 3ds Max to Blender last year, and have not looked back. The rapid pace of development is testament to the fact that Blender now has more than twice as many full time devs working on it than Max.
I really hope that Autodesk starts feeling some pressure from FOSS and cools off on the blatant anti-consumer moves. The whole "now you have to pay us $XXXX/year in perpetuity or you lose access to all of your business files" has killed a lot of the goodwill they once had.
That's just not going to happen, it's in their DNA (and the DNA of many of the small companies they borgified over the years).
It's going to make them die, but it's going to take a loooong time (a generation of users who only know theri product and physically can't learn anything new will have to retire).
All of the drafters I work with are unhappy about the subscription licensing, know we're getting shafted from it, and wish there was a viable alternative. Maybe if Microsoft or some other tech giant decides to sponsor both opencascade and freecad we'd have something usable in a couple of years, but as it stands freecad is both the closest thing to and a _long_ way from even resembling a competitive product.
And freecad, ugh, sorry to be negative, but ... people in this thread are complaining about how bad the Blender UI was back in the days (not the case any more) ... they should try the freecad UI, it's truly, truly terrible ... if there ever was an app that deserves the label 'Kitchen Sink', freecad UI is it.
I have been tempted to buy it, honestly, but it's a little bit too expensive. I use it for 3D printing, not milling, so I don't feel like I should have to pay for those features. So instead, I am gritting my teeth for the day that I have to jump to some open-source CAD tool, which will be annoying to relearn, but ultimately probably not a big deal after I pick what to use.
The team from Nuke consisted of one former Nuke developer and his wife. It was clear that he was a fan of the software as well as a user. Guess who got the contract.
Looks like the nodes selection for this first iteration is a little bare, but that's okay. Gotta start somewhere
Agreed, the support is a little bare currently but from what I understand the hard work has already been done and we should see lots more coming down the pipeline soon.
decades even, but I wonder what edge Houdini will keep in the coming years.
It’s been a long road and for a long time it wasn’t obvious that blender would succeed. Just goes to show if you stick with an open source project over the long haul good things can happen.
It is a huge inspiration to me for how open source could operate.
the collection of features in blender are second to none, and I've enjoyed learning them a lot, but I really wish some of the overall consistency of maya really keeps me away (same hotkeys across the entire app for one thing, multiple overlapping well thought out methods of interaction with systems... and of course "everything is a node" )
another thing that stands in the way is how easy maya is to understand as a beginning coder, every action is echoed perfectly in the script editor, easing the creation of code that simply does a series of action and anything more complicated is fairly simple. admittedly that's currently a MEL only advantage but I'm really hoping bifrost and other future maya extensions help that continue.
all that being said I'm heavily rooting for blender to succeed, both for spurring autodesk into adding some well needed features, and just to having another wonderful 3d package.
"Substantially better" is a fairly large claim.
To be clear, I use Blender and regularly use Maya, Houdini etc and I think all of them, including Blender have so many pros and cons, that I'd hesitate to say any one is better than the other in general.
To your list of subjective metrics I'd also add scalability and performance. Also maybe pipelineibility. Each DCC has very different characteristics on those as well.
Also check out realthunder's freecad fork, they've made some good progress on topological naming (with the effect of keeping models far more stable through minor changes). It's still unusable for all but the lightest duty professional work but it's an improvement.
My criticism is mainly that the PCB layout tools were still buggy. I believe it is/was using some of KiCad's router code, but it took more work to get a layout done than doing the same thing in KiCad would. It also tended to get glitchy with increasing layout complexity. The schematic editor is a lot slower than KiCad's and not as feature-rich.
The GTK widgets look and work very nicely on gnome-based environments but have issues on kde / windows. To be fair that's a GTK issue and not a Horizon issue, but unless Horizon has changed a lot lately I'd have a hard time recommending it over KiCad right now for professional use.
> Library management quirks aside, KiCad is...
I use it very, very rarely so really don't remember more than that.
Footprints and schematic symbols should belong to a parent component class with provisions for inventory tracking, etc. Their library manager works fine for hobby level work but if you're doing production you'll have to have a separate component management process that you integrate yourself.
What they have is fine, it works, but it's a far cry from what you get with siemens eda (mentor graphics) or altium.
That said, anything using an HDI stackup or a reasonable number of differential or matched traces, KiCAD becomes a struggle. Even simple things like a single pair can be pretty tricky if you have space constraints. These features are pretty fundamental for modern PCB design.
It seems like KiCAD v6 will be making a lot of progress to manage more complex stackups and impedance control, so I'm very excited to try it.
I'm sure it has some level of feature parity with commercial CAD buried somewhere, but it's so hard to get to even on the new 0.19 pre-release that I genuinely wouldn't try to put it in the same conversation as Blender or KiCad as open alternatives to commercial software. It's just not there.
FYI if you use the draft workbench you can snap to midpoints and then convert the autocad-equivalent draft workbench sketch to an inventor-equivalent part design / sketcher workbench sketch. You can also use the symmetry constraint in the sketcher workbench to constrain a point to the center of a line.
If you get familiar with it you can technically model almost anything you want[0], but if you want to go back and change a feature (which is kinda the entire point of parametric cad) it'll likely destroy your part if you're not using the topological naming branch. Freecad realistically needs 5-10 more years to get basic feature parity with professional CAD from the early 2000's and then another release or two focused entirely on stability.
[0]: Their opencascade geometry kernel has some moderately serious deficiencies for complex operations/objects, the assembly workbenches are just now approaching the point where they're stable enough that you can use them without having to rebuild your assembly every week, and while the new TechDraw workbench theoretically gives you the features you need to make manufacturing drawings (which is really the end goal for any professional CAD package, those are what you need to get your parts made) in practice it's too buggy to use with even simple assemblies.
It's exactly what I want of a CAD tool: a parametric solver with a 3D UI and 2D sketches. I haven't really have an opportunity to use it a lot since I discovered it, though.
Kicad is quite nice, though.
It's a fantastic platform though and I encourage all engineering/CAD nerds to try it out. It bridges the gap between writing OpenSCAD and using a full fledged CAD software perfectly.
That behavior is controlled by the Mouse Model [1] which can be changed with a selection menu in the bottom right of the window, or by right clicking anywhere in the viewport. I generally use Touchpad. While the default behavior is not desirable for you or me, it is based on a sort of standard interface type, so it's desirable for some. It could be possible to start a discussion about changing the default, but I just wanted to mention that this behavior is not set in stone.
> civil feedback on the forum about it got a dev response of "Those are just our quirks, go back to Solidworks if you don't like it."
Yes, this can be very frustrating to me sometimes too seeing people dismissed this way. There are several issues at play in the forums. Besides the most obvious one of expert blindness from people who have been using FreeCAD as it evolved, you have the problem of people who have the time and ability to help newcomers but not necessarily to fix tough underlying problems. At some point a certain jadedness creeps in and their desire to help can actually be harmful.
Hopefully after some time you'll be willing to give FreeCAD another shot as it is improving rapidly, although the problem space is so big it can seem to happen in spurts.
I'm certainly open to trying it again in the future, and honestly do hope that it becomes a feasible alternative at some point. In fact, I think think that for certain tasks, it's not that far away, but I think that some of the UI decisions, whether from expert blindness as you've pointed out or other reasons, will keep coming up as a point of friction for even moderately experienced CAD users from other programs unless there's a deliberate decision made to try to make the defaults and jump-right-in process more familiar to those coming from the existing options.
Photoshop's UI is nothing great.
Good luck trying something similar for the desktop market starting from scratch as FOSS.
Studios en-mass deciding to fund a shared, Libre resource that anyone can use so that every studio can all benefit and get better features at a faster pace is basically the dream for most Open Source efforts. There's nothing for anyone to be ashamed about from that model, it's good for Open Source to be funded by the players who are most invested in having a really good industry standard available to everyone.
Donations and patreon like schemes hardly allow to pay for a mortgage, supermarket, healthcare, transport every single month.
I don't understand what you're getting at.
You're arguing that user donations and patreon schemes don't pay enough, but there's something wrong with Blender being funded by the companies that use it?
The companies paying for Blender are a substantial portion of the community. They're funding a common resource that all of them share, and that common resource is being made available to everyone, decreasing the barriers of entry for new talent to get into 3D modeling, increasing the quality of potential hires they can bring on because of its wide availability, providing a common target for addons and tutorials, and providing those same studios with a Libre base that they can build on top of if they need to. This is exactly what is supposed to happen in Open Source, people pay for it because it's a common resource that they care about. That's why companies are investing into it.
What is the funding model for Open Source that you consider legitimate? Pulling in enough money to full-time hire an entire development team doesn't count because you don't like the primary user base? Pretty much everything that Blender is doing here is a success and more Open Source projects should be looking at their governance model and learning from it.
I don't understand a viewpoint that says that Patreon and individual-funded projects are uncompetitive and unsustainable, but also that corporate-funded projects with large donors don't "count" as Open Source success stories. It feels to me like you're defining Open Source as a project with zero funding, and then complaining that it has zero funding. Is Matrix a snowflake project? Is Krita? Is Linux itself?
And anyway, all of this ignores the facts that:
A) Even outside of corporate donations that make the majority of its funding, Blender has also been comparatively way more successful than other OS projects at getting individual user donations (I even donate to Blender just as an individual), because they approach PR and community in a really smart, engaging way that drives enthusiasm. And
B) Blender's funding model has nothing to do with the original thing this thread was talking about, which was just that FLOSS software can be of comparable or better quality than its commercial alternatives. You're kind of jumping around, from first arguing that Blender doesn't count because it started as commercial software, and now arguing that Blender doesn't count because it has funding and that's not typical for other projects.
If FLOSS software overall can pull off commercial competitiveness in many cases underfunded, that's almost more impressive. Imagine how good the Open Source ecosystem would be if we as individual and corporate communities started funding more stuff to the same level as we fund Blender.
If the companies sponsoring Blender, many of which already customers from the commercial days, withdraw their support tomorrow, in a couple of years Blender will turn into another Gimp, as everyone switch to something else to pay their bills.
What I learned from my zealot FOSS days, is that trying to sell FOSS desktop or developer tools is hardly any different from trying to make a living as street musician.
It is cool for a while, eventually it turns into survival, as bitter Winter arrives.
And if the companies buying Slack suddenly stop buying Slack, it will also die. The observation that all software development requires an investment of resources is not a FLOSS-specific observation, and it's not particularly relevant to the conversation about whether FLOSS software is regularly competitive with proprietary alternatives.
It's not really an insight to say that if you starve developers they'll stop developing. We all know that already. And it's also true of every commercial product. When a community switches to another product, the original product withers because it lacks support. That is a correct observation, but what does that have to do with FLOSS? That's just a general principle that applies to everyone.
I don't see you arguing that Slack doesn't count as a competitive product because most of its individual users and small communities use the free version, and some day the big corporations it relies on for the bulk of its revenue might switch to Matrix/Element instead.
> many of which already customers from the commercial days
This is also just not true. Blender's current funding exists because they put in the work to get completely new grants and investments from major game studios like Epic, Ubisoft, Steam, and Microsoft, as well as from hardware manufacturers like NVidia and AMD. Blender's funding has solidly increased from its early days.
Blender is not just sailing on its previous commercial support from its previous commercial days. It didn't get grants in those days, companies generally don't give massive no-strings-attached grants to commercial products. Modern-day Blender has greatly expanded the support it has because the people managing its funding/organization are very smart and have put a metric ton of work into understanding and innovating on existing Open Source PR and community management models.
This is part of why Open Source maintainers and organizers should be paying so much attention to what Blender does, because it is genuinely doing an excellent job selling the idea of a community-supported Open Source standard to large companies.
----
Stepping outside of the conversation, just to reiterate one more time: the fact that FLOSS tools are regularly competitive despite receiving on average much less support than their proprietary alternatives is a sign that there is something very cost-efficient about the FLOSS development model, and that pumping more funding into it might be a better use of resources than buying commercial products.
You can argue all you want that Open Source is not sustainable, but there are still a lot of Open Source projects that are of competitive levels of quality compared to their proprietary alternatives.
Could the funding situation be better? Yes. Does the funding situation mean that those projects somehow don't exist, or that they're not currently of comparable quality to their commercial alternatives? No.
>Good luck trying something similar for the desktop market starting from scratch as FOSS.
Krita.
Now admittedly it's a little weird since it's basically the underlying software that powers a commercial paid variant (MixBus), but it always had been FOSS from the beginning.
Most FOSS desktop software that doesn't have such schemes, or is able to have them, hardly makes enough for anyone to make a living out of it.
Once you use ZBrush, you'll never want to go back to traditional CAD-style creation.
I won't claim that Blender can't solve it, but, the workflows are completely different. You end up thinking of models like clay that you can shape and paint, not triangles with textures.
Please, someone, somewhere, spend five years making that an open source project. (Realistically, it's a bit melancholy that such a project probably won't ever get made; there's no incentive to. ZBrush exists, and it has a fine price point. I just love thinking about all the algorithms under the hood that power the whole thing... It would be so cool to see the code.)
That's a little too broad of a statement IMO.
If you're building organic shapes / characters, then yes.
For mechanical parts, using ZBrush is torture.
Also: as others have pointed out, there's already quite a lot of ZBrush-type features in Blender.
It's a common misconception that ZBrush is for organics. You can create some badass space marine type mechs.
Eat3D has a fantastic series of videos that goes into detail on this precise topic: https://www.youtube.com/watch?v=_xOKVGbZmEc&ab_channel=Eat3D
I respect Blender and like it a lot. I just don't think it can do anything close to what you see in that video. But that's just, like, my opinion, man.
Artists love Blender, and that's good enough of a reason to stick with it. I was just hoping to see more of this style of modeling, since it was the only style that made intuitive sense to me personally.
EDIT: (Sadly, it looks like Eat3D is now offline. It was such a fantastic series. Perhaps https://www.youtube.com/watch?v=EdEVj39ur-U&ab_channel=Chofn... "Zbrush Hardsurface Techniques" is a sufficient replacement though. Scrub through that, to random places, and you'll see exactly the sort of flexibility I'm talking about with ZBrush's flow.)
Blender is actually good enough to handle this. Not as well as things like Fusion 360, obviously, but for simple parts, it works.
https://www.youtube.com/watch?v=L3XtAFUWNuk
I've never used ZBrush personally, bu from what I can understand they're a pretty similar workflow...
- Zbrush has modes that are extremely precise. For example, you can mask off a specific section of the mesh, similar to masking off a portion of an image in Photoshop. Except, in Zbrush, you can do it spatially -- it's not a "mask per triangle", but rather "a mask per pixel". You can literally slice a mesh in half, vertically, screen-wise, then hide one half, do some work on the other, then un-hide. It's an effortless flow that people use all the time.
- Zbrush is most natural to use with a tablet in one hand and a keyboard in the other. That video isn't using a tablet at all. Having a mouse interface just isn't appealing for artists. The reason I bring up this point, is because I can't overstate how natural it feels to use zbrush with a tablet –– you can bring up a "context menu" just by tapping a button on your pen, select a new tool by flicking your wrist slightly, and immediately begin using that tool. It's far more configurable than I'm describing here.
Ultimately, Blender has to serve too many masters. If you have to design a UI that can do anything, it can't do everything well. It must make tradeoffs. And these tradeoffs are reflected as reduced artist productivity, which is why many artists use Zbrush for their professional work – it's a tailored experience specifically for artists.
Now, all that said, I am impressed with what Blender can do. I like that it's open source. But, if you want to go toe-to-toe with the pros, the pros are going to point out exactly why the interfaces just don't cut it. If it takes 20% longer to accomplish X, those small costs add up. But there's also a "step change" that you get from the Zbrush flow that I don't think you'll get with Blender's interface. It's subtle, hard to describe, and honestly something you'd have to invest time in to understand.
Caveat: I've been out of the art world for around five years at this point, so feel free to ignore all of my points. But, I approach this from a gamedev studio mentality: from experience in the industry, I can say that Blender never once came up as a realistic option. It was 3ds max for a long time, then mudbox pioneered the sculpting concept (I remember Aron Zoellner from my first job going on and on about why Mudbox was so cool and a big deal), and then Zbrush totally dominated the scene. There are reasons for this, despite my inability to perfectly articulate them.
Scrub through https://youtu.be/EdEVj39ur-U?t=1545 and watch how effortless everything is. That artist is doing so many operations per second relative to your Blender video that it's hard to overstate just how much more productive they are. (I realize your video was an introductory example for Blender, but it's hard to imagine someone doing in Blender what that person is doing in Zbrush.)
https://wiki.blender.org/wiki/Reference/Release_Notes/2.91/S...
I don't quite understand why to wish for something without having tried out what's there?
> That video isn't using a tablet at all.
What leads you to say that?
In the description it literally says: "I use a graphics tablet to paint with..." Did you read that as meaning that he doesn't use a graphics tablet to, say, sculpt?
If so, I think that's a misreading. I own a graphics tablet that I use with Blender in part due to Grant Abbit's (the presenter) advice that it's really a necessity when it comes to sculpting in Blender. :)
> – you can bring up a "context menu" just by tapping a button on your pen, select a new tool by flicking your wrist slightly
That's a pie/radial context menu and one is shown within the first twenty seconds of the video: https://youtu.be/L3XtAFUWNuk?t=17
> If you have to design a UI that can do anything, it can't do everything well.
While that may be true, Blender's UI is more of a UI toolkit (in the not bad sense of the concept :) ) and the tabbed "workspaces" and/or "application templates" of Blender enable the UI to customized to the specific task at hand.
> Caveat: I've been out of the art world for around five years at this point, [...] I can say that Blender never once came up as a realistic option
Ah, okay, that's useful context--yeah, Blender today is not Blender from five years ago. :) The 2.80 release pretty much changed everything and there's been huge development effort in that time.
There's probably better examples of non-beginner focused sculpting in Blender but here's another teaching video from the same guy: https://www.youtube.com/watch?v=KOSdRdeS5oM
And here's a video of the features that got merged into the most recent release: https://www.youtube.com/watch?v=HuSL1F9Sd78
Anyway, I think it'd be worth trying out Blender of 2021, if only to be able to be able to update your critique to aspects not currently available. :)
You could organize some crowd funding to expand it to the feature set you need.
In that regard I'm really looking forward to messing with the geometry node stuff. Based on videos done with beta version it seems like there is a gigantic amount of potential there.
I know this is going to aid me when I start my projects.
Thoughts? My goal is of course to use blender to build stuff, but also render them out performantly.
Nvidia does have some caveats for linux machines in general, it is a slight pain in the ass to deal with drivers. But at least the support is there.
I think with all the cross-platform APIs maturing it's going to matter less and less, but that is me being optimistic.
AMD seems to either be more hit or miss, OR it used to be, and a lot of that is just a poor history.
Either way the GPU buying landscape is difficult for me.
And with 2.92 it's even faster and it now supports CPU+GPU rendering with Optix. Unfortunately, AMD doesn't have such good Blender support. You're stuck with OpenCL and slower rendering time. You could try using their Prorender renderer, but to be fair, it's a closed source and non-native to Blender. I'd rather stick to Cycles+Eevee. I wish one day AMD will work with Blender to better support it, as I much prefer them as a company and I like their opensource drivers. But really, Nvidia RTX cards with Blender at the moment are fantastic.
https://www.phoronix.com/scan.php?page=article&item=blender-... and Ampere is _fast_.
Blender System Preferences for CUDA, OpenCL, and OptiX all report "No compatible GPUs found for path tracing. Cycles will render on CPU." I'm using nVidia's proprietary driver and NVIDIA X Server Settings reports 240 CUDA Cores. No Joy w/ nvidia's cuda packages, either. Blender's been running well on this workstation since ~2.7x for video editing work, so further system tweaks have been low-prio. Will revisit to look at OptiX.
That means that apps using newer versions of CUDA will not work, and your hardware is also far too old to use OptiX.
Put in another way, your current GPU is slower than the one in a Nintendo Switch...
> Fermi ... (last drivers in 2018, you're on a legacy branch).
IYO, is there any value to using legacy nvidia-driver-390 over X.Org?
> apps using newer versions of CUDA will not work, and your hardware is also far too old to use OptiX.
Will look to see if apps I use avail of my CUDA cores, else I'd prefer to remove this sole proprietary driver from my old-and-busted workstation.
Thanks again for saving me time. I’ll re-de-prioritize further GPU config hacking, though if you have suggestions for a Nintendo Switch-to-FireWire 800 interface, please share.
nouveau doesn't support compute at all on those... (as in OpenCL).
Also, as far as I remember, rechecking work on Fermi didn't seem to be implemented the last time that I played with one (it is implemented for the later generation, Kepler, though)
> Will look to see if apps I use avail of my CUDA cores
If your app is compiled against CUDA 8.0 or earlier, targeting compute capability 2.1, it'll work on Fermi. That excludes any kind of modern binaries.
(the newest compiler that CUDA 8 supported was GCC 5.3...)
More recent apps can either be built with an old compiler + C++ library... or use OpenCL instead.
>> "No compatible GPUs found for path tracing. Cycles will render on CPU."
I'll research Fermi for OpenCL and Cycles Render Devices. Blender's working well, so I'll resist the urge tweak until that's no longer the case, and hope to spend more time exploring 2.92.
tl;dr: (sadly) nVidia provides (at times _significantly_) better rendering performance than AMD due to OptiX denoising, which requires less samples per frame. CUDA vs OpenCL is - as far as I know - on par performance-wise.
I did use Blender roughly 10 years ago for school projects, and I would like to get back, anybody has a good tutorial lying around? :)
Highly recommend Ian Hubert! His tutorials are short but if your a bit technical and don't mind pausing to Google you can get pretty far pretty fast.
Default cube, ducky 3d? There are a bunch of good ones on YouTube.
https://www.youtube.com/watch?v=j14b25SnYRY&list=PL3UWN2F2M2...
follow along..
I've started with
Learn basic interface
Vertex, Edge, and Face Basic editing. (Make the following without looking up tutorials) Brick, Cup, Rubiks Cube, Wooden Furniture (Stool, Wardrobe, Bed, etc)
Learn Basic Material
Get Familiar with Curves
Low Poly Create a simple Torch
Animate lit Torch
Before that I went through https://www.youtube.com/watch?v=EmqEP2-UvAY which was also quite good.
Never hurts to dream...
Q: When the new-version splash screen asks to select new or use previous config, am I missing out on optimizations or new features by selecting previous?
- The render cache indicators in VSE are gone. The orange bar segments that show strip cache building is comforting while Preview FPS stutters. Confirm that caches are being built ... but in default /tmp/ vs what's in my previous config. Will start w/ a 2.92 config and update.
Tangent: 2.92 was in the Snap Store earlier this week and updated without hassle. Understood there are issues w Snap, but this was a good experience.
https://developer.blender.org/rBf448ff2afe7a
This page is well-done for other-than-major release highlights:
https://wiki.blender.org/wiki/Reference/Release_Notes/2.92/M...
I must say I'd never imagine to be able to do stuff I do. Blender is not only extremely stable, but also it is so easy to use. It is the finest software I have seen.
You can see my results here if you like: https://twitter.com/SummonTheJson/status/1365363920417005570...
> GPU rendering is only supported on Windows and Linux; macOS is currently not supported.
https://docs.blender.org/manual/en/latest/render/cycles/gpu_...
It'll prolly be in the 2.93 alpha soon.
Blender's UI has evolved enormously in the last five years.
It's come a very long way towards being more:
- A) User-friendly
- B) More "conventional" for people used to the (by decreasing order of user-friendliness) Houdinis, Mayas, 3DSMax, etc ...Long story short, I've been professionally out of VFX/Animation for two or so years now, taking a break doing something else.. and recently I had to do a small type of work in 3D. Naturally, I lost access to most of my paid licenses during the period so I said, fuck it, let's try Blender again (2.8something). I kid you not, I managed to move my way around it on my own, with a bit of help of reading docs, and finished the actual work and be satisfied with it. I still did a little bit of work in Houdini, since I knew the tool better, but 90% of it was in Blender and it was great (Cycles is great too! We'll see about Eevee).
So, something happened, probably recently, where Blender morphed (UI-wise) into this love child of Cinema4D meets Modo kind of thing and one can get around it with no to little help, if you have prior experience with similar tools. I also picked those industry-standard navigation key/mouse bindings and everything's almost there where you expect them to be.
There are also concepts in Blender (like the whole node system) that might inspire some creative solutions in your day-to-day programming. Also, if you're at all inclined to try make video games, you can make your own 3d models. It pairs nicely with Unity or Unreal for a relatively low-barrier-to-entry to build something small and fun.
I feel like this question is like going to an art forum and asking if you should learn Kubernetes. You could, but it's not going to help with your art.
I like that the UI shows me Python functions that I can use to invoke the menu item, and I have written some Python before (read an STL, do some transformations, write an OBJ; Blender was easier than cobbling together other tools to do that). My main takeaway that I apply to my own software is that if I know how to do something in the app, the apps should tell me what API method to call to do the same thing. (Google's Cloud Console does this!) And, that if I know the name of the command I want to run, I shouldn't have to poke through the menus to find it. (Blender fixed this with F3 global search. It annoys me to death in other apps like Fusion 360, where things move ever release and I have to search through every menu because the YouTube tutorial I'm watching is a week out of date.)
The possibilities with CG are sky high but the barrier to entry has been equally so. Lots of money and man hours. But I truly think the next ten years will be exciting and that something like on the level of Star Wars will be made by a small team collaborating remotely.
I think Blender is really user friendly.
Seriously, though, Blender is incredible, and I really really enjoy using it frequently on my own film projects. It’s a joy to work with.
I picked it up as a hobbyist/artist and I really enjoyed it. I haven't used the modelling capabilities in my day-job, but I have used the video-editing capabilities to make instructional videos that were useful in my day-job. For that alone I'm glad that I picked it up.
So for me it's been both fun and useful to know Blender as a software developer, though not strictly required.
Is it helpful as a software dev? Who knows, but it's been an enjoyable hobby for now. I particularly have enjoyed the motion tracking and adding some special effects to videos of my young daughter.
I myself somehow keep gravitating towards Blender for casual use, because every now and then, I want to do something in 3D. The last time was for VR (I also re-learned Unreal Engine for it), this time is for architecture visualization (in particular, I want to design the home office I'll be having once we move from the apartment we currently live in, and I'm half-hoping that beyond usual interior visualization, I'll be able to abuse Blender to do some lighting evaluation and maybe even airflow modelling).
In my mind, Blender represents one of those capabilities that computers have that could be useful for me personally, so I try to get at least passable proficiency in it. Same thing got me to learn the basics of Audacity, even though I have zero hobby or professional need for sound work (I mostly trim out sponsorship messages from youtube-dl'ed mp3s these days, so that my 22mo doesn't get trained to "like and subscribe" while listening to her nursery rhymes). And same thing dragged me into software development in the first place - I wanted to do stuff with computers (mostly making videogames).