Watchtower – Blender Studio
studio.blender.org
studio.blender.org
Blender users have baby duck syndrome and will argue that it's not the UI it is me and I always wonder why is it only blender's UI that has so much friction? Just manipulating the view port can be frustrating at times. Even the three.js 3d web editor feels much more intuitive to use than blenders UI.
The best analogy that I have come up with is someone who has learned to ride a bicycle without a seat, while uncomfortable is still 100% usable, but when you come and say you prefer the bicycle with a seat and it's a better experience they tell you, no you are wrong for liking the seat.
This gif represents what it feels like to use blender vs any other 3d software; https://media.giphy.com/media/l0HlHelCbB9fyXDOw/giphy.gif
Please Blender foundation hire some great UI and UX people, give them a budget and don't add any features until the UI is no longer a pain to use.
I want to use blander very much but life is too short to deal with awful UIs.
I will try Blender 3 but I doubt massive strides have been made to improve the UI/UX since 2.8
The one thing is that Blender is heavily keyboard oriented, either shortcut or search based and while you can do everything by clicking around like most people do in Maya or Max, this is definitely suboptimal. Also if you don't have a numpad and a mouse wheel you're screwed. Honestly though, if you're doing 3D professionally, use whatever your studio / industry uses. If you're starting out today Blender is great because it allows you to produce high quality professional work, has great add-ons that are often free or cheap compared to other packages and is free (as in beer).
There are other aspects of the UI that are annoying mind. Horizontal node graphs are irritating, its only now getting an asset browser, there's a distinct lack of proper progress bars, and from an add-on developer point of view it's a pain in the arse to draw on the viewport, the core codebase is basically not documented and the API is stupidly unstable and is broken All. The. Time. The UI / UX has improved for new users but I definitely agree they need to hire some UI / UX people and technical writers / tutorial makers to make the learning curve less cliff like.
You have a lot of inertia to overcome. That's 15 years of muscle memory and expectations built using different UIs, different paradigms, and different ideas to achieve the same result. You need to factor the fact you're going to be thinking "But I could do that much more easily in <thing I've used for more than a decade>!" a lot. It's possible that you won't ever overcome those thoughts.
Forgetting is a lot harder than learning, and while Blender is pretty amazing it definitely isn't so much farther ahead than the competition it's necessarily worth the effort. Blender's main benefits are that it's fully integrated (it does everything!), and that it's free. If you already have a successful workflow for 3D then maybe it just isn't the app for you.
The irony being that this situation basically is baby duck syndrome, GP is complaining that a "proper" interface is whatever they were initially taught in college. It reminds me a lot of people's reactions to Vim/Emacs. "I've used tons of text editing interfaces, and none of them had this weird modal crap." Nah, you were just taught text editing in Word, and every text editor you've used since then imitates Word.
Which is actually fine, I tell people who look at Vim that they don't need to use Vim. If GP doesn't like Blender, they can just not use it. "I'm used to a different set of UX paradigms" is a completely valid consideration for what tools you use. It only becomes baby duck syndrome when they start saying that every other paradigm is objectively wrong because they don't have >10 years of muscle memory using it.
With blender 2.7 and 2.8 they made lots of concessions to have more familiar interfaces.
The thing is, most every other 3D app uses the industry standard control scheme. Blender is different.
That means switching between Maya , Houdini, Max etc is fairly easy on your muscle memory. Pulling up Blender is a very rough context switch.
2.8 made it a lot better, but using the Maya keymap makes working in the Blender way kind of tricky, so you're still stuck in this weird hybrid.
Blenders great, but it definitely isn't as easy to switch between apps.
Perhaps you can afford a license to all these ‘different 3D programs’ you are jumping between but not everyone can or wants to.
Easy to use and intuitive should not always be the most important goal. Productivity and not breaking existing users work flow are also very underrated aspects.
I think the right comparison would be vim. Absolutely horrible UI, super difficult to use. The first few weeks will be pain but once you get over the hurdle, it is an absolute bliss.
For me blender feels similar. It took me multiple failed attempts before I got it but now any other 3d editing software feels really annoying and I will deeply miss my blender hotkeys.
Now of course there are indeed some actual UI problems with blender. It is a huge software so some parts have not seen much love for a long time and really need improvement. Some workflows are really annoying.
Generally though I would say, just take the bitter pill. The learning curve is steep at first but it is worth it. So yeah the bicycle might lack a seat but it is much more lightweight and reliable, sometimes it is the best tool for the job.
The other programs have lasted decades and have been used to make thousands of movies and games for good reason. Because they work very well with the minimal friction and have PROVEN reliability.
Blender is difficult to learn. Maybe it is not the right tool for you. As I said it took me multiple attempts to learn properly myself. Acting like the blender UI is objectively bad just because it does not suit you is silly though.
The part about other software is especially silly considering that blender is exceptionally loved by its users. Something that commercial 3d software can not claim at all.
I'm sure there is still room for improvement however.
I’ve used Rhino, Cinema4d, Maya, solid works and blender and while they all have their pros and cons blender doesn’t stand out as significantly worse in my opinion.
Well the left mouse button is now select rather than the right mouse button, for starters.
Blender still does have it as an option, in fact one of the few that are displayed front and center before you start editing when it first starts up.
IMHO, the thing that makes it hard is that everything, everything is context dependent. A lot of operations depend on where your mouse is, both what panel it is over and even how far it is from the objects. Which is weird and counterintuitive. Once you learn it, conciously or unconciously, it's very powerful -- dragging to scale something in blender "feels" a lot more accurate once you latch on to the idea that where you start the drag matters. I'm not sure that tradeoff is worth it, but they have been putting a lot of work into the UI and it is improving. At least they finally let go of right-click select by default.
Its also completely scriptable in python. I used blender on google cloud platform for my startup (https://lulimjewelry.com) where I digitally engrave my customers fingerprints on a piece of jewelry and then 3d print it for them. I build the pipeline in blender on the backend using NYT's blender docker containers. Everything can run headless (although headless rendering with Eevee was a pain to set up).
I have an un-related hobby where I often have to edit video. In the past I've used iMovie just because it comes on my machine, but I've been meaning to try out Blender. I've heard some good things about its editing capabilities.
I am just wondering, is such a consistent plugging allowed on HN? To me it lowers the quality of the discussion and their posts are often voted to the top on all Blender related posts.
I think it's cool what doctoboggan has done with Blender, I feel it's okay to mention it in every Blender thread, but he should actually post a fresh comment.
And then it's up to us users if we feel it's become a bit stale. Downvoting exists.
Smart play, as it feels organic until you pull back the veneer.
Blender as a professional tool till has a steep learning curve, as all professional tools have. This is not a negative at all, as some has been led to believe.
There is a difference in "Hard to master" VS "Hard to get started with". GIMP is both at the time, and Blender used to be the same, but now Blender is easier to get started with than before, but as with any professional tool, it's still hard to master as the area is difficult (3D modelling, animation, simulations and more), not because the tool itself. Luckily, mastering 3DS Max, Maya, Blender, Cinema4D or any other 3D suite makes it easy to master the others, as it's mostly the UI/UX that is different between them, not the concepts themselves.
> I have an un-related hobby where I often have to edit video. In the past I've used iMovie just because it comes on my machine, but I've been meaning to try out Blender. I've heard some good things about its editing capabilities.
You're probably better off using something else unless you're using Blender heavily in your process already. It's nice that the UX of general Blender translates to the video editor so when you're deep into modelling and just need some quick video edit it's fast to do it without too much context switching. But for larger video jobs, you're better off using a dedicated video editing tool like Resolve, as it's still miles ahead of the competition.
I distinctly remember how incredible it was when Big Buck Bunny was released with all assets and project files freely available. I learnt so much from that, and wouldn’t be surprised if it was a catalyst that led to many, many people gaining the skills and drive to produce new creative works. I hope Blender continues to inspire and enable many future generations of creators.
This year, I’m giving it another try with the infamous Doughnut tutorial and I’m very impressed. Once I got a general idea on the thing, I’m able to navigate the interface and figure out how to do stuff without too much effort. It’s incredibly powerful piece of software(it has scripting and visual programming capabilities) with a very healthy community.
Oh an my Macbook Air M1 is more than enough to get it going. Disney doesn’t need to get scared just yet but having access to such a powerful software on such a portable device makes you feel powerful.
You can also create in blender and then import into unreal.
Both creating ML-enhanced content creation and editing tools of course, but also synthesizing images, videos, 2D and 3D content for ML training.
For example, you can use using GPU-accelerated ray tracing and procedural character animation in photorealistic environments to train body tracking and pose estimation networks, since you get the precise "ground truth" and pixel-perfect semantic segmentation for free without enlisting Mechanical Turk to tediously and unreliably annotate millions of video frames.
And you can use randomization and retexturing to plug in different body types, skin tones, hair styles, clothing, etc, to train it to work with diverse sets of people, in a wide range of environments, camera angles, and lighting conditions.
BlenderAI : Artificial Intelligence in Blender:
https://download.blender.org/documentation/bc2008/BlenderAI....
Blender for AI Devs, pt. 1: Overview:
https://www.youtube.com/watch?v=UtRxC9HPGd0
Blender pipeline to generate images for deep learning (BlenderProc) - Maximilian Denninger:
https://www.youtube.com/watch?v=1AvY_iS6xQA
>Using blender to generate realistic images for ML training:
>This is a two parts talk. It is based on the papers "3D Scene Reconstruction from a Single Viewport" presented at ECCV 2020 and the "BlenderProc" paper. The speaker is the main author of both papers. This is the recording of part 2.
>Recording of part 1:
>References to everything covered in the talk:
https://www.reddit.com/r/2D3DAI/comments/jbkjdy/references_f...
>Lecture abstract: We present BlenderProc, which is a modular procedural pipeline, helping in generating real looking images for the training of convolutional neural networks. These can be used in a variety of use cases including segmentation, depth, normal and pose estimation and many others. A key feature of our extension of blender is the simple to use modular pipeline, which was designed to be easily extendable. By offering standard modules, which cover a variety of scenarios, we provide a starting point on which new modules can be created.
>arxiv:
https://arxiv.org/abs/1911.01911
>git:
For me that's always been the issue. They change the UI all the time, which means 90% of the tutorials out in the wild are invalid in various different ways.
After that, Blender makes it easy to manage concurrent versions, so it's fine to just download and run additional versions if you're feature-chasing from a new tutorial or looking for a bug fix.
Blender is a lot of people's first modal, process oriented software, and it does require a different mindset than working in other tools where menus and windows are all available from one interface. I'd still argue the amount of conceptual mastery and memorization needed for each is basically the same.
The repo doesn't have much context: https://github.com/nytimes/rd-blender-docker
Edit: They have writeups here: https://rd.nytimes.com/research/3d-web-technology
I've attempted to start with something basic - some code to pare down an audio stream into basic hash identifiers a la Shazam - so that media can be identified regardless of quality etc (even encoded to 64kbps mp3, for example). Which is surprisingly difficult...