So....I have this thing, I always want the maximum out of everything...So Houdini being Houdini, I wonder to myself "Why not for the most amazing arch viz possible?"

So before I did anything, I went to the Reddit for Houdini and floated the idea of exploring the platform for archviz, and eventually, modelling some things beyond my regular Revit workflow. Then I basically got laughed and chided out of my own thread.

Background: 3d Modelling in Revit since RTC in 1999 Beta, 3d modelling in autocad before that. Started a rez design firm in 03, got into rendering, and hated 3ds max/Vray, so I mostly used Paranesi and did NPR stills for ten years (in PS after Para went belly up).

Then along come real time. I dive head first (D5/TwinMotion/Enscape). Many years later, I want back into the quality game, I want cinematic renders and ani's for my visualizations. Options: FBX (IFC?) out of Revit to whatever works best, so I explore, ready to sink time, money and resources into learning from a deep comit persp.

So I've spent the past three months exploring Ytubes and every media/prod_lit I could get my hands on, and I see:

Blender, I like, but it's so, so good, but @ ArchViz? I'm not feeling it.
MAX V/ray, beautiful combo, the rock of gibralter, and exensive along with Revit, but I still hate both interfaces (max/viz).
Maxon: I love the feel of C4d. But once I dug in, I got discouraged due to what I perceive as a destructive workflow, and then there is the price: 1200 year for the core, plus 5 things I don't need.
Vantage with anything: not ready for me, does not keep up with revit's release schedule, while completely loving on SketchUp.
Skecthup. Babes in Toyland.
UE5. Just wow. I'm still not convinced I shouldn't go down this rabbit hole, but most of what I see in the wild (ArchViz), still looks Uncanny Valley to me. So that's holding me back.
Omniverse: Looks incredible, I'm not sure I am ready for it.

Then my biggest, badest want of all, Houdini.
90% I've talked to try and tell me all the reasons why I shouldn't touch it.
But the cost of the indie licence, for this level, a no brainer. I dare anyone to do a cost/beni analisys, then go tell everyone why they shouldn't be JUMPING on this monster. Of course, I'll probably never use more than 10% of it's horsepower, but then, the same could be said for so many complex wares, including my own brain.

So I rationalize: Why in the name of god's green earth would I not want to combine VFX into my client presentations, such that they are superb, and my clients tiny little brains are completely blown away? Why is this a bad thing?

So I am hoping there is someone in here doing Archviz, (I don't want avante guard spaghetti-techture, I want to take this to the actual client/buyers, not explore the limits of academia), hopefully...one or two who might be able to point me in a direction? All options are on the table. Learning platforms, private tutoring, business partnerships, purchases, whatever (!)if anything like "this" is even available. Or, I suppose I simply consume as much as I can, and adapt myself and come up with my own workflow. I just want to learn it and do things unique and novel in my field.

And....I'm not a real big "all the reasons why I shouldn't" do a thing kind of guy. I see folks like MellowMesher and his line of architectural products, and one or two firms using this tool, doing amazing things for client visualizations, but otherwise, it is a true desert for Houdini users in my speciality. So given I'm paying over 6k yearly for the priv of renting all the other stuff, what the hell? $260 for the whole suite of the most progressive and powerful software (DCC OS?) on the planet? Or am I just being stooopid?

Please excuse my rant, but most of all, I hope some of you guys and gals can understand why I feel like a very thirsty man, standing at the side of a well, not sure if the water is life giving, or poison.

Any ruminations would be greatly appreciated.

Thank You,

Billy

Billy, Here's my 2 pence worth : )
First of all I'm not an archviz guy, so hold that in mind, but I am a 3D generalist that has moved to Houdini from C4D.

Do I think you can do archviz in Houdini? Absolutely. Is it the best option for you? No one can say, there are just too many variables - your mindset and approach, the specifics of the work, and so much more. You've really just got to try it and see.

As you're well aware, Houdini has a very strong VFX background, that's it's bread and butter. Sometimes that can result in a bit of a 'if it's not VFX, Houdini is not for you' attitude. However, over the last few years I think it's very clear that H's user base is widening to accommodate a greater range of users and work. The good news is that Houdini is an incredibly powerful and flexible toolset - and it's open for any use. If it fits, use it : )

Here are a few 'bullet points' I'll send your way:

1. Indie is cheap, but the real cost and investment is your time. Budget for a year's learning to start with.
2. Houdini is unlike any other DCC, be prepared to learn like a beginner.
3. Be prepared to be 'kissed' and 'slapped' by the software : ) There will likely be times you'll want to wipe it off your hard drive forever, and dopamine hits when you create something you couldn't with anything else.

Here are a couple of things you might enjoy watching:
Fraser Shiers: Houdini for the generalist: https://www.youtube.com/watch?v=MH6sl0gKiyc [www.youtube.com]
You might also enjoy Adrian Lambert's channel on YT. He does environment work in H and has some archviz modelling type videos - check out his NeoGothic Cathedral stuff: https://www.youtube.com/playlist?list=PLCWTAH7pGMWQYjgN46TCLDpT5v7ICW8EC [www.youtube.com]

Hi Mike,

Thank you so much for your thoughtful and useful reply.

As for your bullet points, I appreciate your experiential advice, and so I'm OK with all, and esp., as my business deals with never ending crits and handling "I don't love it" like marriage proposals, I think a little "kiss and slap" is just fine. Everyone just better be ready if they slap too hard, I might just kiss back. (And even more, rejection can be a very useful tool to understand how to get where we really need to be, when handled well)...

But all that aside, I really am completely stoked to jump in and see if I can make this work, from a working studio perspective, on real world projects. Archviz is taking a pretty interesting turn toward less experienced pros polishing what would otherwise be marginal renders, and running that stuff through the generative scrapers, which is fine, but there is also some sameness, and a look emerging which makes some of it mundane, at least to me e.g., golden hour stills with misty shadowy backgrounds, which works well until every firm in the country is an instant AI viz "artist": (and truth be know I'm completely jealous and jaded because I'm also a traditional artist, and I worked in graphite/ink and water media for years to develop what modest tools I now possess, only to see so much of that artistic/intrinsic humanity reduced to a prompt by folks who may not even own a proper pencil sharpener). So all prompt envy put to the rear, perhaps within Houdini, if I combine what I know, and fold that into it's immense tool-set, who knows? I might find something unique, or even better, even develop my own signature style which is not the run of the mill today.

Anyway, now I'm rambling. So much to consider, or, just put down the mouse and get to work. I think I'll do that, and perhaps I may even post some stuff soon. Perhaps I'll ping you if I do, if you wish to crit, kiss and slap, or just kibitz about the state of the art world.

My very best to you Mike, and Thank You, again.

Billy Coop

(Oh! Almost forgot, thanks for the links, and I've peeped a little bit of Maestro Lamberts creations, so absolutely I'll be following him closer after I've run through some of the rudimentary aspects of H. )

Gotta say though... your enthusiasm is contagious.

Hey BillyCoopDraws,

I would say go for it. Learn Houdini.
It is rough around the edges. But it provides unparalleled capabilities to
develop tools and workflows catered to your needs.
And once you get the hang of it every other DCC feels limited and constraining.
Just prepare for a rough start.

Cheers
CYTE

Hi,

Houdini can definitely handle architectural visualization and so much more. I think the reason it's not as popular in ArchViz is because of its steep learning curve and the possible lack of ready-made tools for that market, which are common in other DCC software where you can just drag and drop without needing to build things from scratch.

Back in the day I did some level 5 autonomous vehicles sensor analysis and real time visualization setup for Lyft using Houdini. I can't post the actual images but it looks roughly something like this:







Here are a few other examples of what you can do:



Just a shameless plug from my VEX course



This is just the tip of the iceberg. The sky is the limit.

Personally I feel like Archvis has been heavily dependent on fast GI renderer, rather than being about modelling, especially for indoor lighting when considering realistic archvis

And Houdini has been lacking in that area and one can argue still does out of the box as both Mantra and Karma are both "unbiased" path tracers with some exceptions of basic pointcloud caching support in Mantra and some path guiding in Karma which can help a bit but definitely not on par with much faster renderers usually used for archviz

However nowadays there is a variety of renderers available for Houdini as well as USD or one can potentially also render in UE or Omniverse, so rendering part shouldn't be a limiting factor

And as for creating the architectural models themselves and then larger assemblies etc. you may find that Houdini's procedural tools may actually be better suited for such tasks than having to do everything manually, at the expense of some learning curve and potential mindset change, Houdini is definitely more choose your own adventure type of software rather than rigidly constraining you into a specific workflow

Thanks Tamte,

I appreciate your take on viz/GI/speed over modelling. Coming out of a early 3d modelling background, while all around me were still clinging to the flat earth society (God loves Autocad), at least in my case, I don't feel confident with my own work unless my geometry is faithful and pretty meticulous, Ans i have relied on that, at least in my renders. How that translates here, I guess I'll find out.

As far as M&K being unbiased path tracers and their ability to compete with some of the unicorn render engines, I'm a little disappointed to understand they are lagging. I was actually hoping the opposite was true as I was looking forward to diving in head first. However, I hold a license for Vray, (and others) and as it's integrated tightly with Houdini, and because it brings so many assets to the party, in retrospect, I'd probably be a fool to abandon everything that it has to offer. That said, I still love travelling down new roads, and so I hope Houdini, and especially the upcoming release, has some nice surprises on the rendering side.

Then to your last comment, the more I study how Houdini works, procedurally, at least with some of the geometry I've studied so far, the more I begin to realize it is actually much closer to how Revit sees the world, from the perspective of parametric modelling/object creation, and so on, than any other DCC I've tried. In fact, I see allot of similarities between procedural modelling, and parametric modelling. The difference at the rudimentary user level, with Revit, a tech can get in and drive the car, push the pedal, apply the brakes, and drive pretty well. While in Houdini, one seems to have to know how to assemble the car, then drive the car as one sees fit, and as creatively too. Under the hood, at Revit's family creation level, parametrically, I see some very encouraging similarities, so other than learning which node does what, etc., I feel very much at home, at least so far.

Therefore, I am extremely encouraged. Architects are known as great copiers, and the same is true with the tools we use. I've seen many free form DCC's come and go (nurbs mostly), some have stuck, so when we see others creating beautiful forms and works, which have the potential to enhance our traditional work flows (some of us grow so tired of straight lines), I guess we can't help ourselves. So I hope Houdini has its moment in the sun, in Arch/ArchViz, and I hope what little work I can explore and contribute helps shine some small light on the possibilities. THANKs!

tamte
Personally I feel like Archvis has been heavily dependent on fast GI renderer, rather than being about modelling, especially for indoor lighting when considering realistic archvis

And Houdini has been lacking in that area and one can argue still does out of the box as both Mantra and Karma are both "unbiased" path tracers with some exceptions of basic pointcloud caching support in Mantra and some path guiding in Karma which can help a bit but definitely not on par with much faster renderers usually used for archviz

However nowadays there is a variety of renderers available for Houdini as well as USD or one can potentially also render in UE or Omniverse, so rendering part shouldn't be a limiting factor

And as for creating the architectural models themselves and then larger assemblies etc. you may find that Houdini's procedural tools may actually be better suited for such tasks than having to do everything manually, at the expense of some learning curve and potential mindset change, Houdini is definitely more choose your own adventure type of software rather than rigidly constraining you into a specific workflow

animatrix_
Hi,

This is just the tip of the iceberg. The sky is the limit.

Hey, Thank you for all of that, Anamatrix. Very interesting work that looks like you were pushing some boundaries.

Best,

B.