Houdini Engine for Unreal and Unity Now Free | Hacker News Reader

Houdini Engine for Unreal and Unity Now Free

sidefx.com

85 pointsdoppp38 comments

Houdini is an exceptional tool! It took me a few tries to get through this press release... this only seems to change things for companies using the engine plugins that make >$100k.

Yes, though a major annoyance is that it's only the official SideFX plugins that are free, not Houdini Engine itself (the API). So you can't write your own specialized tools.

I ran into this recently when we needed an interactive rendering technique that Houdini doesn't support. The Houdini Engine API would have been perfect, but we use Indie and it doesn't support anything but the official plugins.

Instead, I'm writing HDK nodes in C++ that accomplish the same thing by copying to shared memory in a subprocess running Optix 7. It works, but it's ugly as hell. Oh well.

Ugh. Maybe this was a move to simplify licensing, I hate having to dig around to see what features are gated especially if there are caveats like API access vs Official Plugins only.

Clearly a response to the new node editor in Blender 2.92 which threatens Houdini.

I doubt it. This is only for the engine integrations, not the core software.

It's fun to watch from the sidelines how Blender, as an OSS product, is just plain winning. It'll drown out everything but specialized tools and/or products that are essentially just supply chain to particular large operations very soon (i.e. customers who need staff to interact with to do complex integration).

It seems to be down to economies of scale - the business model of something like Maya, moving a small amount of "units" at high margins essentially, has been relatively easy to surpass in volume. Get the governance right, make sound technical decisions and ... you can't really compete with it without emulating it.

I don't say this with any particular glee. This niche genre of proprietary software has always appealed to me as a confluence of some amazing complex problem-solving and UX innovation - I'm still so impressed by Modo, for example.

Does anyone care to comment why this is happening with Blender vs. Maya/Cinema 4D/Houdini, but it hasn't happened with The GIMP vs. Photoshop, or Ardour vs. Logic/Pro Tools/Ableton, or Godot vs. Unity/Unreal?

It seems like overall the pattern is for the commercial options to be significantly more popular (and arguably more featureful) than the open source options. Why has Blender been able to reverse this pattern?

I think it's a lot of little things.

- Blender came out of commercial, and the community was never really anti-commercial. Lots of classic FOSS communities have trouble identifying with paid development for example, worrying it will lead to a form of classicism in the community, etc. Blender, OTOH, tried to make sure there's core staff doing it fulltime basically from the get-go, I think. This makes a lot of stuff happen - taking care of not-fun things, having stable contact points, professionalism, doing good fundraising, etc.

- Blender took dogfooding seriously, with their open movie projects

- Blender took user research seriously, running events and inviting artists over and watching them use the software

- Blender did good community management, with Blender Artists and other initiatives

- In Blender's area integration, automation, tooling are all very complex, and they made some good key decisions, e.g. adopting Python for scripting just when Python was becoming the default programming language for non-programmers

- Because the application domain Blender is in is so complex, training is important, and the competition probably underestimated making their products affordable for teaching institutions (call me out if wrong, I am not as confident on this point). Blender supported creation of training materials pretty well, too

They simply got a lot of things right. Projects that paid attention (e.g. Krita) are also blossoming.

These are good points. Going one step further: Do you think the other projects I mentioned (Godot/The Gimp/Ardour) could replicate Blender's success in their own industries by similarly making better decisions? (I don't have an opinion on this, just curious if you think Blender was also successful for reasons that would be outside of the control of these other apps to replicate.)

Not the person you replied to but here's my take for the two that I can comment on:

1) Godot: No. Godot is doing pretty well with the resources they have. Making a 2D/3D game engine is an incredibly complex task and they have only a handful of people writing code. To compete with unreal and unity they'd have to have a ton of funding, a layer of FOSS-competent management (a somewhat rare thing), and many more developers.

2) Gimp: Yes but that will never happen. Gimp has a handful of issues. They're chained to a difficult framework and a large amount of the value they provide as a tool comes from their plugin library so they can't just start ripping things out. They also have rather questionable branding (both in the name and the splashscreens, especially the ones in development versions) and a hearty resistance towards throwing on even a veneer of professionalism. Whether or not they should change to fit what the rest of the world considers appropriate is a philosophical question I'm not touching, but the effect of their not doing so is pretty evident.

Krita (a very well managed project in comparison) has basically eaten Gimp's lunch for a lot of workflows and will continue to do so while gimp withers away (which has basically already happened, gimp's GTK3 builds are only just now about to release, 10 years after the first GTK3 release and right after GTK4 dropped -- bear in mind that GTK was originally developed as a custom widget toolkit specifically for gimp).

Curious how "The GIMP" is so behind the "GIMP Toolkit".

Price point may be another important factor. 2D Design apps like Photoshop are not as pricy as the 3D counterparts, last time I checked. Hence there's more motivation to find an alternative to Maya, Max for the hobbyist then there is to replace photoshop.

These days Maya and Max are the same price as Photoshop - $250/y vs $20/mo. Incidentally Houdini is also about the same, $279/y.

> These days Maya and Max are the same price as Photoshop - $250/y vs $20/mo. Incidentally Houdini is also about the same, $279/y.

As I recall, pricing of 3D modeling/animation/rendering suites was much steeper when Blender started gathering momentum.

[Checks pricing]

Umm, it looks like the price is still quite high (Maya and Max are each $1.6k/y). Where are you getting that figure from?

And now that Photoshop/CC have been creeping up in price you have other contenders like Affinity which are decent at much lower price points.

Gimp is (I think) a struggle and ui/ux nightmare to most people still.

Software like Clip Studio Paint, SAI, and Krita have definitely taken a bite out of the digital painting audience for photoshop.

I'm not sure how good the alternatives are for photo editing/manipulation.

Affinity Photo has some traction. They have several good products, and momentum started when Adobe went subscription and they stayed perm license (with a reasonable cost).

Affinity Photo (and their other products) are great. And for £25 with COVID discount is a no-brainer. An actual alternative to Photoshop (and InDesign/Illustrator).

GIMP is still an unusable mess. I've tried so many times but it's just too different to the industry standard.

The best change they made was add a single-window mode but it's still nowhere near Photoshop.

In my opinion, a lot of this stems from the UI/UX focused improvements they made with Blender 2.80. Before that, Blender had some weird UX patterns (like right-click to select) that were different enough from other packages that it was hard to get into. It was always frustrating that it on its surface it looked like the other 3D packages, but it didn't ever work the way I wanted it to. Also, they coupled those UI/UX updates with a large focus on polishing their realtime viewport renderer. With better UX patterns, and a fancy new viewport, it made demos and onboarding much more intriguing.

That's the same way I had always felt about GIMP or Audacity - the windowing system, icons, layout, and other tools never quite felt natural or intuitive and it never made me want to get into it more.

I honestly believe that Godot will overtake Unity one day. Unity has a huge amount of momentum, but it kind of feels to me like the difference between JQuery and React right when React was a fledgling library. While Unity has a lot of entrenched developers that have been using it for years, my (unfounded) opinion is that new developers entering the gamedev scene will pick Godot over Unity more frequently, and that will bear out in the long run (years).

I know this is kind of unrelated to your point, though :) Additionally, I use Ableton all the time and I don't see any FOSS alternatives that hold a candle to it. Then again, I could be one of those entrenched devs just like those I said exist working with Unity!

Are there any fundamental differences between unity and godot? React and jQuery are philosophically complete different - are there any salient aspects of godot that makes it better than unity/potentially more popular?

That’s a good question. It’s not a total paradigm shift, no - it’s just better in so many subtle ways. Perhaps a better comparison would be lodash vs underscore! (Of course that doesn’t work either because one isn’t a subset of the other, but I think you see what I’m driving at)

Blender is simply as good or better than the alternatives.

While the open source image, audio and video editors are not better than the alternatives.

Sure there are other factors to. There are always other factors. But given enough time the best product will win.

How is it actually winning, I work in the industry I don't see Blender used but actually the same old "proprietary" tools ( Max ect ... )

I've seen a surprising number of resumes that are listing it come across my desk recently. I don't think it will occupy the same place in the tool stack that say, Linux does for IT, but adoption is definitely on the uptick. I suspect that this has a lot to do with rendering tech moving more towards the realtime side of things for many applications these days. Blender has a very good realtime viewport, so if you are delivering content into that sort of an environment, it's a natural fit for some things.

It may seem strange but when discussing things like this ("This new thing is going to win") I'm always reminded of the response people here on HN had to Adobe going to subscription-only. There were pages and pages of comments about how it was disrespectful, obviously dumb, CS2 was so good they didn't even need a subscription, how they were just opening themselves up to their enemies who would now win, and how because of all that, Adobe would clearly be regretting such a disastrous decision, they'll see, just wait for them to rue the day. Surprise! Adobe is over 10x more valuable than it was 10 years ago, and is still the 900lb gorilla in the room.

I suspect Blender is "winning" in this area in the same way that Adobe was failing: in the minds of people on this website, and nowhere else.

Google "studios that use blender".

The number of studios not using blender is much larger.

Blender needs to spend at least 10 more years to threaten Houdini, Houdini handles far more stuff than geometry nodes.

I mean don’t get me wrong, I would love blender to be a direct competitor to Houdini but there is a long long way to there.

I use both professionally, and Blender isn't even close. But I do love the direction it's going, for sure.

I wonder if this is in any way related to Blender's new node-focused workflow tools.

Epic put a bunch of money into the company behind Houdini I believe. I'd wager that's the bigger reason for this happening.

I don't feel I fully understand what Houdini can do for me.

Is the Houdini Engine useful without having purchased a copy of Houdini?

What does this engine do?

Houdini is a package for authoring procedural content generation tools. You can construct tools in Houdini like a building generator or level generator and no longer need to use up licenses to include them in your software at runtime.

As in beer?

As in candy.

I guess this is a "game changer"...

Reply on news.ycombinator.com