McNistor
The hallmark of a skeptic mind. Bravo!

Ha, don't think so From what I've read it almost sounds like it's worth investing in Softimage today until Houdini gets up to speed, it's that good

The reality is if you search the forums it is littered with requests for modelling fixes for years, and although the current gestures and listening is really great, until real movement on the ground- actually seeing new tools that we can use….. I am hopeful that changes can be made, as the recent changes through-out Houdini 11-13 show enormous momentum to improve things.

MartybNz
From what I've read it almost sounds like it's worth investing in Softimage today until Houdini gets up to speed, it's that good

Yes, it's that good, although I'm not sure you're OK with paying a price of a full 3d app only for its modeling features (not saying only the modeling is great in Softimage, quite the opposite). That combined with the fact that it just got whacked leads me to believe it won't happen.

MartybNz
The reality is if you search the forums it is littered with requests for modelling fixes for years, and although the current gestures and listening is really great, until real movement on the ground- actually seeing new tools that we can use….. I am hopeful that changes can be made, as the recent changes through-out Houdini 11-13 show enormous momentum to improve things.

Hmm. Perhaps I should stop investing my time and emotionally in the hope these will get implemented?

McNistor
Hmm. Perhaps I should stop investing my time and emotionally in the hope these will get implemented?

This is your call; all software engineering projects work with a promise but no guarantee. It affects everything from large Telecom firms to software companies to everything in between - sorry if this is a shock but that is the computing industry.

It's not a shock, I'm aware how this industry works it's just that I got a taste of skepticism from your part, based on your past experience regarding modeling requests not being addressed.

Well, I'll probably get into scene interaction and viewport improvements after which I'll stop.
These have to be implemented even before start addressing specific modeling tools as these are the foundation upon which the others rest.
I mean having a bevel tool that works well is one thing, how you access that bevel tool and how you interact with your objects and elements upon which you'll apply that bevel is another - that is an workflow, not a bevel tool.

Modelling has been discussed ad nauseam - the difference and hope here is:

1. Fresh blood.
Softimage's axing has created new users with great experience, expectations and ideas.

2. Momentum.
Sesi has indicated that they were already looking into this properly

This coupled with recent improvements to H12 to H13 brings high hopes.

Yeah, don't stop. I for one would like to stay in Houdini for modelling rather than go to Modo (which is getting expensive for just pure modelling), formerly I was using Silo. I wish Nvil had a Linux port.

It's exactly as MartybNz has said on both accounts. Lots of scattered suggestions have been made in the past, though none have really been clear, consolidated and endorsed by SideFX as they are now because of recent events and actions.

McNistor
Speaking of which, I posted an inquiry on the previous page regarding brush picking tool - is more than a selection tool?

We have brush-based tools (sculpt, paint), but these don't use the same 2D brush as the brush selection tool. The brush for sculpt/paint orients itself to the surface and doesn't leave a trail, much the way you've suggested the brush select should not.

The reason the selection tool currently leaves a trail is that we haven't implemented interactive selection display yet, and this gives you a better idea of which elements will be picked. Interactive selection display has been planned since 12.0, but keeps getting bumped due to other priorities.

Once primitives (& points, verts, edges) highlight as you brush, it won't be necessary to leave a trail anymore, and we can go back to a simple outline of the current brush position. I'd like to have that (finally) implemented for the next Houdini version.

MartybNz
Modelling has been discussed ad nauseam - the difference and hope here is:

1. Fresh blood.
Softimage's axing has created new users with great experience, expectations and ideas.

2. Momentum.
Sesi has indicated that they were already looking into this properly

This coupled with recent improvements to H12 to H13 brings high hopes.

Alright, I hear you. I probably should keep my expectations low and be pleasantly surprised rather than vice versa.

Gyroscope
Lots of scattered suggestions have been made in the past, though none have really been clear, consolidated and endorsed by SideFX as they are now because of recent events and actions.

Exactly the reason for which I created this thread - in the hope to avoid info being all over the place.

As far as modeling goes, I want to make something very clear for everyone. This is of course my opinion, but I can (and dare I say already did with regard to the few things I've addressed so far) substantiate any suggestion or critique I have:

modeling in Houdini, and forgive me for putting it so bluntly, sucks not because its modeling tools (well, not mainly because of that). I know this sounds crazy but hear me out. Modeling tools such as extrude, polybevel, polysplit are among the best as far as results they produce goes if not THE best of all 3d apps I've tested. Others need a bit of improvements here and there and others are completely lacking, but we'll get there soon enough.

Modeling in Houdini sucks because how you select and interact with objects (mentioned before how it selects points after you go into edges, or how it reverts to ‘Y’ tool even after an undo - I'll continue to add problems on my page) and their sub-object elements (points, edges, polys), how you access those modeling tools and how fluently you can go from one to another, how gizmos are working, etc.

Should the things I've tackled so far plus those to come be solved, you'd instantly have a much better experience modeling in Houdini w/o even touching “real” modeling tools.

twod
We have brush-based tools (sculpt, paint), but these don't use the same 2D brush as the brush selection tool. The brush for sculpt/paint orients itself to the surface and doesn't leave a trail, much the way you've suggested the brush select should not.

The reason the selection tool currently leaves a trail is that we haven't implemented interactive selection display yet, and this gives you a better idea of which elements will be picked. Interactive selection display has been planned since 12.0, but keeps getting bumped due to other priorities.

Once primitives (& points, verts, edges) highlight as you brush, it won't be necessary to leave a trail anymore, and we can go back to a simple outline of the current brush position. I'd like to have that (finally) implemented for the next Houdini version.

I understand. In light of lacking an interactive selection display, it makes sense to have a trail.

Alright, great to hear you'll tackle this.

Have a great weekend everybody!

Softimage is an amazingly intuitive, fast modeller. And it is very stable now, which is a critical thing. They upgraded one year and the main purpose of the upgrade was to address stability issues. I think that was the 6.5 release. 3dsmax and Maya are not exactly unusable modelling apps, either.

Softimage is fully mature at model + texture + render, and irreplaceable. I would consider it impossible to remake. I'm not sure anyone should even try. Should Houdini attempt it? I don't know. It will be a big uphill battle to piecemeal features from Softimage into an effects app like Houdini. I'm not sure that's a good use of their resources, given the competition. I can find 5-7 apps to model in, including Softimage, since it can be installed and left for years… the modelling aspects are complete.

Another main benefit to Softimage is scene and instance management, and how instances interact with the material editor. There is a certain simplicity to Softimage that is unmatched. It is extremely easy to build unbelievably complex scenes based on instanced assets, and manage and render the scene with no issues.

Autodesk is killing one of the best 3D applications ever created. It is unreal to see this happen.

tfddi
Autodesk is killing one of the best 3D applications ever created. It is unreal to see this happen.

It's more common than you think. Pinnacle systems killed off Commotion, Avid killed Eladtic Reailty, Apple killed Shake. All leading apps.

Edit added here so as to be inline with standards for this thread : and now AD has killed off a really good viewport and modeller which we are trying to address here

tfddi
Softimage is fully mature at model + texture + render, and irreplaceable. I would consider it impossible to remake. I'm not sure anyone should even try. Should Houdini attempt it? I don't know. It will be a big uphill battle to piecemeal features from Softimage into an effects app like Houdini. I'm not sure that's a good use of their resources, given the competition. I can find 5-7 apps to model in, including Softimage, since it can be installed and left for years… the modelling aspects are complete.

Another main benefit to Softimage is scene and instance management, and how instances interact with the material editor. There is a certain simplicity to Softimage that is unmatched. It is extremely easy to build unbelievably complex scenes based on instanced assets, and manage and render the scene with no issues.

I understand this is the SI part of the forum and as such somedegree of fanboyism should be expected.
But I would also expect a couple of considerations from our ex-SI mates, which are more than reasonable.
For starters, I agree that many other apps (actually most) have a better “destructive” experience regarding modelling than Houdini, even some that haven´t been mentioned at all (such as the apparently “doomed” Lightwave, which provides a fantastic modelling experience, sometimes better than Modo, if a bit rusty at some points).

But mainly, I frown when I see people with almost not experience on Houdini speaking about scene management, complex scenes, or instanced assets or anything similar, simply because it´s obvious to me (and not to them) that when some of those users get confident with Houdini they will realize it´s simply a much better software for dealing with that kind of scenes.
Simply put, the more complex the scene gets, the better the comparison will be for Houdini against any other software, because in Houdini by phylosophy you still deal with many small elements as opposed to one complex scene (please read Jordi´s guide).

Netvudu
But mainly, I frown when I see people with almost not experience on Houdini speaking about scene management, complex scenes, or instanced assets or anything similar, simply because it´s obvious to me (and not to them) that when some of those users get confident with Houdini they will realize it´s simply a much better software for dealing with that kind of scenes.
Simply put, the more complex the scene gets, the better the comparison will be for Houdini against any other software, because in Houdini by phylosophy you still deal with many small elements as opposed to one complex scene (please read Jordi´s guide).

Nah - disagree - it's good to shout out if ‘the emperor has no clothes’.

The best way past the ‘philosophy stage’ is to give concrete examples to compare with. When someone says - it has a great viewport and modelling tools, great - give us some examples and we can compare notes. Who knows, maybe Houdini is best, maybe SI is better.

Edited to adhere to thread topic

MartybNz

tfddi
Autodesk is killing one of the best 3D applications ever created. It is unreal to see this happen.

It's more common than you think. Pinnacle systems killed off Commotion, Avid killed Eladtic Reailty, Apple killed Shake. All leading apps.

SideEffects killed Maya

This thread is dedicated for the discussion regarding making Houdini better for modeling and everything involved in that.

Please keep it on topic and refrain from making s/w comparisons for the sake of it, especially if they're not related to modeling, viewport interaction and all that jazz.

McNistor
Should the things I've tackled so far plus those to come be solved, you'd instantly have a much better experience modeling in Houdini w/o even touching “real” modeling tools.

This sounds good, and, thinking about it- the selection tools updates will go a long way.


At a guess, the timeline to implement all these ideas will be; viewport snapping tweaks will happen soon, next version hopefully , then the selection tools will be updated, then the modelling tools will be updated. Going by past releases, a point and full version per year - I would guestimate 1.5 years to implement a thorough update from where we are today.

1.5 yrs would be an acceptable time frame. Plenty of time for Softimage users to accustom with Houdini as far as FX and all the non-modeling related things go.

I was going to address something else other than Snapping, but since you say they'll work on it soon I will tackle snapping next. I only hope I'm not too late as I have a few ideas about snapping to share with you.

What are the main complaints people have about snapping so far? (other than that it seems very buggy)

In general the main issue everyone seems to have, snapping is inconsistent and does not work. I've personally found it to snap to the pointer not the point/vert you are trying to snap!

Please look at the Korny Klowns post in the link, and maybe the page before too, as good issues are raised.

http://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&t=30366&start=50 [sidefx.com]

twod

I'm glad these talks are on and I can't wait to see what's SideFX has in store for Houdini. It's a little sad that it took the demise of one software though

Coincidentally, there was already a lot of internal discussion at SESI about modelling, animation and general viewport workflow prior to Autodesk's sharpening of the axe. Thus we're pretty open to making changes at this time, both subtle and drastic. Funny how things work out sometimes.

Please keep up the discussion, we're listening. I just don't want to inject my 2c because I'd much rather see how this discussion evolves (except to correct an incorrect assertion about the current state of Houdini).

Wow an actual answer from a developer, this is actually a great improvement over atdsk. Not to bash any company, but this is how the things are. Anyway, can i come up with few suggestions over the animation tools for you guys? I have a huge experience in doing MOCAP manipulation and cleaning, and working with animation curves and animation layers. I just finished 3d for a huge Hollywood producer(the killing of JFK to be launched) and i am very proud of this achievement of mine. So far i don't see much functionality in Houdini animation curves(fcurves). Partially to the fact that i don't know the program yet, but i still find some missing parts and some oddities. I would like to come up with just a few small suggestions with screenshots and idea if you are open to this. Will take me awhile since i am in the middle of a project right now, but if i receive a positive input on this i will do it.
Thank you,
Constantin

For those that miss Temporal Pivot (under “alt” key) from XSI:

There is something similar under ‘ key. Hit it once (the solid shaded Gizmo handles became wireframe) then you can offset and rotate pivot as you wish. Than hit the ’ key again to get back solid shaded look of transform gizmo. Then you can transform your geometry around this temporal pivot without changing actual pivot.

So only thing to make it work as fast and user friendly as in XSI is to be able to map it to sticky (transient) key. And I think and hope that this transient key functionality comes soon.