The Houdini Modo combo

   37048   32   4
User Avatar
Member
8 posts
Joined: Jan. 2013
Offline
Hey all. I thought I´d post a few things about my experiences. A year ago I was hired by a company to make the transition from SI to Maya. However that was not that easy. First SI is just the better program in all fields than Maya. Second SI user don´t like Maya to put it mildly, just the the thought of transitioning to Maya makes them shiver.

I´m a Maya user but I always used Modo to get the grunt work done, modeling and lookdev, then going over to Maya mostly for Arnold or Vray. So Modo gets in to the pipeline and it works great for certain things. Used together with Mari, Nuke and Zbrush it´s really effective. You can get a lot of basic to mid complicated work done. The great benefit is that you can get your Nuke compers to learn Modo in a very short time.
When it comes to “real” 3D it falls short. I´m talking about advanced particle effects, rigging and character animation. At the moment those tools aren´t enough. It´s not that easy to compete with SI in rigging and animation, they had decades of improvements and all kinds of specialized tools. There are nice features in the Modo animation system that is very good indeed. But setting up a complex rig with facial controls etc. is probably not doable.

Next thing I did was to dig in to Houdini. I´m an art kind of guy that always shied away from TD stuff in Maya. But a year later I´m slowly turning into a TD. With Houdini it´s so much fun and actually easy compared to Maya.
The rigging system in Houdini is the best I´ve seen anywhere. I´ve never gotten so good results so fast anywhere else. I have not gotten deep into the animation but if you set up the rig with the proxy rig like you are supposed to the feedback is good. The viewport is a bit clunky, a little tricky to select handles.

To sum it up a bit.

Modo pros:
Amazing modeling (including mesh fusion)
Great previz tools, rendering, painting, shading, .
Nice GUI
Easy to learn and get great results

Modo cons:
Hard to make really advanced 3D
The alembic exporter
Animation tools promising but not there.

Houdini pros:
Amazing TD tools
Great rigging
Amazing rendering and shading with Mantra
Fluids, pyro and destruction.

Houdini cons:
Classic modeling not possible.
No texture painting and sculpting.
Animation tools promising but not there.

Conclusion:
Houdini and Modo complement each other really good but animation is still missing. So sidefx and the foundry, just create an animation system that is up to SoftImage standard and we are good to go (and get around Maya completely).
User Avatar
Member
245 posts
Joined: Sept. 2011
Offline
Hi there,

This all sounds pretty familiar, and as a big Modo and Houdini fan I, tought I might as well share my experience as a Maya, Modo and Houdini user from the past year.

I have been keeping my eye on Modo for a few years and decided to try it out when version 701 got released. And I absolutely loved it! But when I started working a bit longer with Modo I eventually also found out the weaker parts of it.

My main focus was looking for alternative software to replace certain areas that I would not like to use in Houdini which are: modeling, uv's and texturing, fur and rendering (mantra is nice but Arnold is better), rigging and animations (houdini's rigging and animation tools are not too bad could be alot better, I still use houdini for some rigs and animation types)

First I used Maya <-> Houdini and then I dropped Maya and tought I might give Modo a shot, after six months I actually decided to put my Modo journey on a pause because it still lacks many many features and switched back to Maya for now.

Modeling:
Modo has GREAT modeling tools (the best one in my opinion) Maya is a close second in my opinion, well designed hot box tools really speed up the modeling process, but in my opinion Modo's viewport controls + modeling is still hands down the best on the market. Modo and Mari are the only two packages that dont feel awkward navigating on a laptop by the way, I absolutely love the controls.

UV's and Texturing:
Modo actually has some really good UV and texture painting tools, however I decided not to use Modo for these areas. I already use Unfold3d + Mari for this, they are highly specialized in these areas and Modo just happens to be able to do both of them quite ok. If you are looking for a all in one program for modeling, uv's and texture painting I can recommend you to stick with Modo because it can do this very well.

Rigging and animation:
Modo's rigging and animation tools are also pretty neat. I've create three identical rigs in Modo, Houdini and Maya and I the only thing I found out missing were the muscle systems. Maya and Houdini muscles are really nice, but Modo did not have any which was a big bummer. I was working on a python script for importing Modo riggs in Houdini and calculate muscle sim after animating in Modo but the modo FBX exporter and alembic exporter are also missing some key features. I still use Maya as my main rigging / animation software because of the deep integration with Pyhton and rigs with many control objects feel much faster when moving around.

Fur and rendering:
Maya + Arnold is kinda unbeatable when it comes down to fur and rendering but again I have to say that I really like Modo's fur tools and the renderer is also kinda nice they feel intuitive and I can get nice results really fast, the fur styling tools sometimes feel kinda odd. The downside really was the amount of controls over the animation of the fur for example when I created some grass it was kinda hard to create some good looking wind result without spending to much time. Rendering fur also feels quite fast, I think its faster than Houdini but to create very sharp noise free renders you add alot to the render time. The renderer overall is really nice, I also love the preset library that comes with Modo. But again I have to say its not the best one out there, Maya in combbination with Shave and Haircut and Arnold Renderer is just unbeatable.

My conclusion:
Modo is very capable of doing modeling, uv's, texturing, fur, rendering, rigging, and animation but for me it didnt feel complete, I wanted to have a little more in each of these areas and Modo just couldnt offer me this. I think Modo has some great potential, I still keep track on the development and I think somewhere this month Modo will release some news about version 801. Really looking forward to that.

To be clear I'm not saying Modo is bad, like I said before I'm actually a big fan.
Modo is great for some small stuff if you want to stay inside one program.

If you have some questions about Maya <-> Modo <-> Houdini please dont hesitate to ask, I kinda have been button bashing these three and know my way around them.

I hope I did not offend anyone with this in any way, I just wanted to share my experience..
User Avatar
Member
4189 posts
Joined: June 2012
Offline
Reading through the Foundry forums of Modo shows some good stuff but there seems to be a lot of bugs and performance issue on quite a few setups. Do you find Modo up to the quality level of Houdini?

An an aside it was quite entertaining to see ‘T4D’ thoughts on what a node is, contrasted with a developer

http://community.thefoundry.co.uk/discussion/topic.aspx?f=4&t=85519&page=6 [community.thefoundry.co.uk]
User Avatar
Member
1573 posts
Joined: March 2020
Offline
These Lightwave guys never really grasped what node networks are good for (even after they implemented one, it seems) DDD
Imre Tuske
FX Supervisor | Senior FXTD @ Weta Digital

qLib -- Houdini asset library
http://qlab.github.io/qLib/ [qlab.github.io]
https://www.facebook.com/qLibHoudini [www.facebook.com]
User Avatar
Member
8 posts
Joined: Jan. 2013
Offline
I agree completely with Johnnycore. Modo and Houdini are really polar opposites and I think they complete each other pretty well. Modo has bugs but not more than any other software, they are just very open and honest about it on the forums. At the moment we are struggling a bit with the Mavericks viewport, but there are workarounds. Houdini used to be unusable in OSX until H13, but now it works pretty good.
What I want to stress is the importance of getting the animation up to the highest standard. Once soft is gone there is only Maya that can deliver a complete animation package. Competition is good. So dig into the rigging and animation in Houdini and give some great feedback. If not I can´t see why any studio wouldn´t swap in their XSI licenses for Maya.
User Avatar
Member
1573 posts
Joined: March 2020
Offline
Let me elaborate on my previous semi-sarcastic remark…

I'm thinking about looking into modo more deeply for some time now, because as much as I know about it, I too think it would be a great companion to Houdini (modeling/UVs especially).

Last time I checked, it was a great interactive modeler but without any serious modeling history or node-network features (and functions that pointed toward that direction could crash it).

Although there could be an entire separate thread of philosophical discussion about which approach is better in the long run (node networks or not), I rather not go there right now. But.

Interesting news is, a few days ago we had a presentation about the current state of Modo, and a few important remarks I can make (apart from it's still an excellent modeling app)

1) it has certain procedural aspects to it now (including a node network, heh heh), it can do things now like Houdini's Scatter/Copy SOP

2) more interestingly, the guy said (I'll have to look up the details) that it is (will be?) able to do things like exporting instanced geometry properly to alembic (as instances), exporting custom geo attribs, too, and there is emphasis on the alembic (and proper interconnectivity in general) side of things.

Which is very good news. For people who look for integrate Modo in their pipelines, that is.
Imre Tuske
FX Supervisor | Senior FXTD @ Weta Digital

qLib -- Houdini asset library
http://qlab.github.io/qLib/ [qlab.github.io]
https://www.facebook.com/qLibHoudini [www.facebook.com]
User Avatar
Member
4189 posts
Joined: June 2012
Offline
It's interesting though, there have been a few posts over on their forum showing how Modo could do with SI snapping improvements, and some ex-SI users saying that the modelling in SI is still the very best.
User Avatar
Member
1573 posts
Joined: March 2020
Offline
MartybNz
It's interesting though, there have been a few posts over on their forum showing how Modo could do with SI snapping improvements, and some ex-SI users saying that the modelling in SI is still the very best.

I have another sarcastic-funny remark, but this time I'm keeping it to myself
Imre Tuske
FX Supervisor | Senior FXTD @ Weta Digital

qLib -- Houdini asset library
http://qlab.github.io/qLib/ [qlab.github.io]
https://www.facebook.com/qLibHoudini [www.facebook.com]
User Avatar
Member
1755 posts
Joined: March 2014
Offline
MartybNz
It's interesting though, there have been a few posts over on their forum showing how Modo could do with SI snapping improvements, and some ex-SI users saying that the modelling in SI is still the very best.

Reading stuff on internet forums would be unwise to not be taken with a grain of salt - that's the general rule.
However, there are as you well know, people that do research and/or testing before opening their mouths and then it's even more unwise to not at least consider and perhaps follow in their footsteps with a few tests (production tests, not orbiting around a cube).

Modo's modeling is very good. Where it lacks is stability (last time I checked) as it crashes randomly without the possibility to reproduce them. And, also a construction history as it was already mentioned.

Since modeling tools between XSI and Modo are on par as far as quality (the result a bevel tool outputs on the same geom for example) and easiness of use, coupled with the fact that XSI has some sort of history as well as being rock solid (as far as modeling goes at least), then I can conclude that Softimage's poly modeling is indeed better.

I'm obviously leaving out MeshFusion and as time passes I'm sure Modo will get better while Softimage will remain still as far as modeling goes.
User Avatar
Member
1573 posts
Joined: March 2020
Offline
McNistor
MartybNz
It's interesting though, there have been a few posts over on their forum showing how Modo could do with SI snapping improvements, and some ex-SI users saying that the modelling in SI is still the very best.

Reading stuff on internet forums would be unwise to not be taken with a grain of salt - that's the general rule.
However, there are as you well know, people that do research and/or testing before opening their mouths and then it's even more unwise to not at least consider and perhaps follow in their footsteps with a few tests (production tests, not orbiting around a cube).

I was thinking about people who acquired certain personal preferences over the years and now getting goose bumps that things work slightly differently in another application.

(I for one am a perfect example, as I'm getting goose bumps from an app that doesn't have a modeling history – but maybe you don't need that for modeling. I certainly need to get over that quickly if I want to use it. And I do hope that Modo saves the scene on crashes.)
Imre Tuske
FX Supervisor | Senior FXTD @ Weta Digital

qLib -- Houdini asset library
http://qlab.github.io/qLib/ [qlab.github.io]
https://www.facebook.com/qLibHoudini [www.facebook.com]
User Avatar
Member
1755 posts
Joined: March 2014
Offline
riviera
I was thinking about people who acquired certain personal preferences over the years and now getting goose bumps that things work slightly differently in another application.

(I for one am a perfect example, as I'm getting goose bumps from an app that doesn't have a modeling history – but maybe you don't need that for modeling. I certainly need to get over that quickly if I want to use it. And I do hope that Modo saves the scene on crashes.)

So what's the sensible thing to do when in this situation, adopt it as it is or the highway?

Because if that's what you're saying, it could very well be interpreted as a hint to the users (among which I belong) that come here to propose improvements regarding Houdini's current tools/workflow.

“get used to it” argument again?

Might've misunderstood your “hint” though, case in which disregard what I said.
User Avatar
Member
4189 posts
Joined: June 2012
Offline
One of the most important issues in modelling in Houdini compared to others is to identify what it's strengths and weaknesses are compared to other implementations. A good first bet is on selections, snapping and handles improvements.
User Avatar
Member
1755 posts
Joined: March 2014
Offline
MartybNz
One of the most important issues in modelling in Houdini compared to others is to identify what it's strengths and weaknesses are compared to other implementations. A good first bet is on selections, snapping and handles improvements.

Indeed.
+s: procedural, good implementation of the tools (as far as I tested, which is not foolproof since I didn't get my hands very dirty) such as bevel and extrude which gave the best result when I tested the same geom in Maya, Softimage and Houdini

-s: what you mentioned and also viewport and interaction quirks. Preferences are also rather scarce and the hotkey system could benefit from an improvement or even overhaul.
User Avatar
Member
1573 posts
Joined: March 2020
Offline
McNistor
So what's the sensible thing to do when in this situation, adopt it as it or the highway?

Because if that's what you're saying, it could very well be interpreted as a hint to the users (among which I belong) that come here to propose improvements regarding Houdini current tools/workflow.

“get used to it” argument again?

Might've misunderstood your “hint” though, case in which disregard what I said.

There definitely must be a misunderstanding somewhere as you seem to be hearing things I didn't say. I especially not “hinted” at anything. I also wasn't aware of “get used to it” was ever a valid argument (I definitely wouldn't call it one.)

This topic was about the Houdini/Modo combo anyway, and my comments were related to that (well, mostly). Obviously I'm planning to look at Modo because Houdini lacks certain things that I might need (mostly interactive and/or modeling related).

Btw, let me not just hint but state clearly: if you want to propose improvements, you're in the right place – the sidefx guys do listen to their users and very open to suggestions, I can only say good things about them. I'm looking forwards to improvements to certain areas, too.
Imre Tuske
FX Supervisor | Senior FXTD @ Weta Digital

qLib -- Houdini asset library
http://qlab.github.io/qLib/ [qlab.github.io]
https://www.facebook.com/qLibHoudini [www.facebook.com]
User Avatar
Member
4189 posts
Joined: June 2012
Offline
Modo looks good, but I'd like a proper evaluation of where Houdini modelling stands as all you usually here is ‘it’s no good compared to other apps'.
User Avatar
Member
1573 posts
Joined: March 2020
Offline
MartybNz
Modo looks good, but I'd like a proper evaluation of where Houdini modelling stands as all you usually here is ‘it’s no good compared to other apps'.

Okay, I'll give you a few examples, just off the top of my head:

- lots of poly-append tools (e.g. PolyKnit, PolyBridge) rely on the order of your component selection to determine the created poly face normals (if your selection is in the “wrong” order, face normals will point the opposite direction). they should be “aware” of the neighboring face normals.

- many of these OPs are just “awkward” to use: you'll eventually figure out what selection to pass to them and which order, but often it's not very intuitive.

- practical example: the PolyExtrude SOP's global translation parameters have no pivot (e.g. it uses the world origin for scaling), this is positively a pain even (especially!) for procedural modeling

- however: some modeling OPs are okay if you're modeling with a “procedural idea” (e.g. a reusable parametric house model for a procedural city). My personal favourite is the PolyStitch SOP, great stuff

IMHO the reason for these things is that most of these operators are 10+ years old and practically never got any overhaul from an user-experience point of view (even if they did, they also had to be backwards-compatible which can be a problem – although now with operator namespaces, versioning can be done.)

So, my bet is once things start happening, new versions of many modeling operators will appear.
Imre Tuske
FX Supervisor | Senior FXTD @ Weta Digital

qLib -- Houdini asset library
http://qlab.github.io/qLib/ [qlab.github.io]
https://www.facebook.com/qLibHoudini [www.facebook.com]
User Avatar
Member
1755 posts
Joined: March 2014
Offline
riviera
There definitely must be a misunderstanding somewhere as you seem to be hearing things I didn't say.

Good to hear that then.

MartybNz
Modo looks good, but I'd like a proper evaluation of where Houdini modelling stands as all you usually here is ‘it’s no good compared to other apps'.

Paying for Modo (or any other modeling s/w) doesn't appeal many people, especially when Houdini could become just as good if not better.
And as for the particularities, we have the sticky thread for modeling issues.

And then there's another matter, one more related to the general state of affairs.
To me this community looks way too similar to Softimage's, wallowing in its software's greatness. And then disaster happens and people start looking back wondering what went wrong.
In these times a product has to appeal to a larger audience for a more secure future. It doesn't have to sacrifice any of the current paradigms which made Houdini the powerhouse for TDs that it is, just add and improve in certain areas.
Keeping it in this esoteric state does not bode well with popularity, which, whether we like it or not, does have a lot to say in this economy as far as a product's survival goes.
User Avatar
Member
4189 posts
Joined: June 2012
Offline
State of affairs is very positive here
User Avatar
Member
1755 posts
Joined: March 2014
Offline
MartybNz
State of affairs is very positive here

Thanks the universe for that.
BTW, I didn't intend to bring my bitterness regarding recent events around here, but that's one reality you have to keep in the peripheral vision, otherwise it could sneak up behind you and bite you in the ass.
Anyway, enough of that, let's get back to how to make Houdini even better.
User Avatar
Member
1755 posts
Joined: March 2014
Offline
lots of poly-append tools (e.g. PolyKnit, PolyBridge) rely on the order of your component selection to determine the created poly face normals (if your selection is in the “wrong” order, face normals will point the opposite direction). they should be “aware” of the neighboring face normals.

In Softimage the polyknit equivalent has arrows which means that if they're pointing trigonometric way then the normal will be “good”

  • Quick Links