Ace, what would you recommend for rigid body physics simulation for thousands of objects? Houdini? THX
Houdini just has no competition in the dynamics space (it pretty much has tools and settings for nearly every conceivable physics effect known to man).
But Blender does have a pretty competent and easy-to-use rigid body solver using Bullet (which can handle a pretty large amount of objects). The main limitation though is that it doesn’t come with dynamic fracturing unless you pre-fracture the object yourself (which can be done automatically via an addon that comes bundled with Blender).
If all you need are objects falling, rolling, piling up ect… (even moderately advanced effects like chains) Then Blender can easily do it.
MAX and Maya are entrenched and established in the 3d world and have been industry standard for an intimidating length of time.
That’s the main reason really, much more than technical advantages or user friendliness.
It’s possible though that stuff such as pipeline proficiency, commercial plugin viability, Tech support, customer service and development model may be minor influencing factors among others.
…but still why there are no Top studios using Blender?
Even if Blender somehow suddenly became the world’s best 3d program it would still be an extremely uphill task to replace the profession’s default software. Companies have invested years in MAX and Maya workflows and to an extent their livelihood relies on them.
So the industry has mainly MAX/Maya seats and asks for mainly MAX/Maya operators and Uni courses naturally train students in MAX/Maya. It’s a self completing circle.
These sort of threads always become a haven for specific gripes about Blender or the BF but the whole “but Blender can’t do XXXXX/Autodesk does XXXXX better” thing is a complete red herring.
Blender seems to be having some success with indie studios and lone artists but top-tier industry adoption seems a bit optimistic in the short term at least and I’m saying that as a committed Blender fan. Maybe 2.8 can push Blender above the radar a bit more, we’ll have to see.
Blender seems to be really good I tried 3Ds max and I was disgusted.
It’s great that you like the program but you’ve got to get your head around the subjectivity thing and realise that for someone else the reverse may be true and it doesn’t mean you’re right and they’re wrong or vice versa. Horses for courses eh?
These kind of questions seem to come up now and then. Is it possible to have some kind of sticky thread that provides new users with a bit of perspective on this kind of thing? I suppose it’s easy to ask for but what would go in it?
That actually sounds great. I wouldn’t have known otherwise. I hope some details in particular that I find troubling are being re-done as well. I doubt though, BF is going to actively fix any issues with all the addon contrib scripts. There are some things very wrong with some officially included addons. Such as object creation addons, their objects, when created, are at an absolute size regardless of what scene unit I and grid size I use.
The Ivy Gen in particular, only works on 1.0 scale. The code were written with no relative sizing in mind, I tried to fiddle around with it to make it work in different scene scales with no result so far. It’s such an amazingly complex algorithm that I wasn’t able to find those places where I can multiply by a scaler and make it work with any scene scale. Headache.
I believe the difference between the official addons and the contrib addons is that the official ones are proven to work well without major issues (which is why the contrib ones are never bundled with the release builds and are filed under a ‘testing’ section). If the issue is with an official addon though and it seems like a genuine bug, you can actually report it to the tracker.
Thanks; I made some quick and not too deep tests with Blender, but the simulation seemed to be crazy slow (I’m new in this area and of course I wanted to keep everything in Blender if it was possible). I would like to animate the objects through a dense environment (collision) attracted by a force, guided by some guidelines).
For speeding things up, you can simply make use of a simpler collision bounds type like sphere and box. You can also choose to use the convex type which can give a good fit depending on the object.
If you use a triangle mesh bounds type, then it’s going to be really slow if it’s on a high-polygon object and the object is a rigid body. Usually that type would be applied to static meshes.
Thanks; I would need mesh colliders unfortunately. Anyway, I will make some new tests in the future.
Several of these things like dedicated tech. support is actually a major factor among companies that choose commercial software (because it allows for a faster and easier way to solve problems correctly than forums).
Also, you might want to walk back the first line quoted a bit, Maya 2018 came out and people are actually praising Autodesk again (they brought out a huge amount of bugfixes along with many improvements. Development is also speeding up greatly relative to previous years with their new approach to updates). For some studios, this might deep-six any idea they might have had regarding a serious look into Blender because Maya is showing up to perhaps actually be worth the high subscription costs.
Even then, Blender’s position is still pretty safe due to the 2.8 work and the zero cost of entry, but Autodesk is aggressively ratcheting up pressure on Maxon (through MASH) and Modo is now being squeezed on both ends (due to it being too pricey for hobbyists yet being inferior to Maya in many cases).
It’s funny you say that. Certainly Blender development has issues. You say it’s resource based and that’s definitely true. However, I would argue that it’s also project based. The open movies push development into overdrive and implement more features in a short period than most other software. That’s really the key to why Blender makes any kind of waves in the industry. I remember back before I started using Blender, hearing about how they had LCSM mapping or heat map weighting or dual quaternion skin before anyone else. It was the open movies that made that possible. And it’s those features that catch eyes and make people take Blender more seriously. Certainly, we will probably never see large, established studios switching to Blender. But what we will probably start seeing is more and more small studios start up using Blender.
I work in Max just about every day and I have to say, I hate it. I’ve been working in it for some 10 years now and it’s really behind in features. We have Maya here as well and I’ve been trying to get them to use it more but there is a lot of resistance from the core team.
One issue is that Max doesn’t have a built in render pass system. You have to use a plugin (RPManager) and it’s pretty clunky. The other big issue is that their new file referencing system is pretty sucky.
But by far the worst issue is that there are bugs that have been in there for years. You think Blender has bugs or abandoned features? It’s a sht stack in Max. We complain about them all the time and there’s no way to even know if they are even working on them. However, since Max is market driven, they don’t really give a sht about anyone not doing games or Arch/Vis. Also, if there is a bug that gets fixed in the mean time, we have to wait till the service pack come out. Don’t even get me started on how totally stagnant development has been either.
I know you all have your reasons for being frustrated with Blender’s development strategy. But to tell you the truth, It’s transparency is actually one of it’s greatest assets. We all know exactly what’s being developed and roughly how long it’s gonna take (as long as there is a roadmap). And if something isn’t going along as you wanted, you can always make a loud noise about it on the forums. You never know, it might actually get fixed if enough people joint in on the cry. You can bet your ass no one is on the other end when you complain to Autodesk other than some poor tech support guys.
Just a thought: I probably have more chance to effect Blender’s development than Autodesk: If my studio wanted some feature that we just can’t live without, we could just take the money that we spend on Autodesk and Foundry licensing per year and fund that feature’s development in Blender. No way, no how, is that going to happen with Autodesk.
I think most who’ve worked in a commercial environment will recognise the importance of tech support and if you need it you’ll probably not even be looking towards Blender but it’s not a ‘major factor’ in Autodesk’s industry dominance (which is what the OP was referring to).
Also, you might want to walk back the first line quoted a bit, Maya 2018 came out and people are actually praising Autodesk again.
So great it’s flavour of the month again. It’s not like the bigger studios were chewing on anything else much. Even if on occasion, Maya tasted like worn out trainers they didn’t exactly vote with their wallets. Yeah they might curse and gripe about Autodesk for whatever reason is in vogue that week but it simply isn’t viable for them to change to Maxon/Modo/Blender/whatever.
Dug in establishment is the main reason, never said the only reason.
The big players seem to use Autodesk (or at least commercial) software pretty much exclusively and always have, I’ve seen the job postings. I dunno, maybe mid-level and smaller outfits are more receptive to other solutions if the shoe fits. The OP was mainly referring to “top studios”.
I’d like to see greater infiltration from 2.8. Success with the pros is good for the rest of us, just not expecting miracles.
People always point to features, but for established studios the immediate cost of switching would still weigh stronger than the distant benefit of switching, even if all those features were implemented.
Change tends to come from the bottom. It’s proven that Blender can do feature animation, so there should be a market opportunity here: Lots of cheap Blenderartists that can’t find a job because they can’t do Max/Maya plus savings on licenses means an economic advantage. The market doesn’t only need the very high end.
Different methodologies have been explored over the years. Long ago the waterfall approach was state of the art, then there were many others like extreme programming and nowadays agile programming, especially Scrum and Kanban are widely used.
There are plenty of companies that spend a lot of money to switch their methodology. I don’t see this kind of change as playing safe at all. It is a huge risk and both companies and programmers are often still open for those changes.
Exactly and there is even a middle ground between high-end and indie.
Studios come and go. While it might be difficult to change everything for an studio working with Max/Maya for decades, if a new one gets built up from scratch Blender might have a better chance. Especially in today’s age where outsourcing is much more rampant than decades ago.
If the studio works primarily in one of the lower/easier ends of production like pure modelling and texturing, Blender could get a huge piece of the pie. A Zbrush/Blender/Substance/Mari workflow/pipeline is possible today and not less efficient than the same with Max/Maya.
The Animation and VFX area is the most prestigious and this is the place where Autodesk is heavenly entrenched (but gets strong competition from Houdini).
I personally think Blender alone (as a tool for everything) will never be as successful in an AAA+ Animation/VFX studio environment, but in combination with highly specialized tools Blender might throw out Maya as the 3D Hub, at least for medium small studios it could work.
I mentioned earlier that Blenders weakness can be sidestepped by using Houdini and a dedicated lighting/rendering/look def tool like Katana or Clarisse. You do the VFX in Houdini, all the models and animation in Blender, textures in Substance/Mari, export to Clarisse/Katana where you built the final scene, shaders and light and render it.
This is much more effective than rendering in different programs and comp it together in post.
True it is a big company but each product has product managers and also people who direct and lead them too.
True is also that AD in the past did a lot of research also funded a quite a lot of project - I was recently invited to pier 9 but cannot go. However from a user points of view we also see AD buy up a lot of technology and then not really using it well or not using it the way we would. But that can be attributed to their size too.
Hey look at Adobe - Photoshop still has the same silly path tools while Affinity offers a workflow where their photo and vector tools offer the same tools and even save in the same file! Instead Adobe tacks on silly 3D effects and such …
That’s I think what Carl mentioned. 3D was more fun to do then working on boring tools that would however greatly improve the way we work - well since Adobe does not we switched. Best ever invested 100$ for Affinity.
In short you need to know Maya, even if basics, to have employment in the field. This is if only to fit into collaborative environment. Maya is the final part of asset assembly where everything’s given life and rendered. Until that step though it does not matter what program you use especially for modeling, texturing. For animation and rigging you must use it in a studio with maya pipeline, very little exceptions.
Where maya shines though is extendability. You have no glass ceiling, no limits with plugins with such as Fabric, Houdini engine. You cannot extend Blender in same manner (e.g no .dll’s to stack functionality). We do have Alembic now and it is a game changer! Hope to see OpenVDB one day aswell.
On the other hand these plugins are very expensive as are the render licenses. To be fair most adsk users own their soft illegaly and have no guilt parading a work that might have taken them up to 5k worth of soft and plugins to generate. Blender is remarkable for being able to match it for free. A wild guess would be, should ADSK enforce copyright tomorrow, we’d have a lot different usage statistics.
Lastly by knowing Blender and Maya you will gain knowledge and new perspective that helps you in either software. You might find that Blender is more efficient for certain parts, such as modeling and this might give you performance advantage in a collective. This is precisely what I am doing having utilized it on a few AAA games by now. I am also plesently surprised to hear that a lot of asset creation studios are adopting Blender. Given what’s happening to Eevee and it’s modeling capability, I would say that future’s bright for Blender.
As for max. Used it about 18 years. Cringe every time I open it now. If you’re into architecture though, you should probably learn it.
That’s just my subjective take
Or, to put in other words: People who pirate Maya instead of using Blender aren’t hurting Autodesk, but are hurting Blender. The Blender user community could be much larger and active if it wasn’t for piracy.
Your assumption is wrong, skw.
If they are working in an environment which forces them to use Autodesks software, they have no option to choose. They just choose not to pay Autodesk.
If somebody who is free to choose is pirating Maya or Max, he doesn’t want to use Blender. Simple as that.
A bunch of people forced to use Blender even if they hate it isn’t a healthy addition to the community.
People choose a sotfware several reasons.
- Because their school teach it.
- Because jobs are on Maya or Max
- Because of the hype at this moment
- Because they like it for X resons
If they crack Maya or Max it’s because they need it for one reasons or another or just to try.
In france we have more and more jobs for Blender, the 2.8 will be a game changer and people will use it more.
Companies use blender more and more and if they find people who know it, Blender will be used at the same level as maya or max.
After, depending of what you use blender, it can do everything, but it’s better for some points, modeling etc.