groups?

Have you tried with the Labs hda, delete small parts?

That is such a confusing question to ask, but for the sake of others who are reading this forum I will offer my thoughts. A pipeline for what? Rendering, set design, character animation, sims, texturing? Pre-vis? Compositing? For the VFX studio, the pipeline (if you get an offer from the producer to take a job on the VFX) is checking the brief(script/story-reel)->R&D->prototypes->tests->approval->fine tuning->delivery. And it is most likely every single time tailored to the specific project/client by a pipeline designer professional and under a fat NDA disclosure. Why do you imagine that on the public forum would someone share such details? Or you don't, since you already said it is probably not a best place to ask. (like, how is that related to Houdini, other than being a dcc as one of the components of the pipeline)
I suggest to check from studio to studio with your portfolio and try to work for them, if you want to learn a real deal of business (and if they let you). If you are curious enough and from now on become clever in asking information without exposing someone to breach the nda, you might end up designing pipelines yourself, in order to optimize the workload on artists/coders. And that makes the difference between a freelancer and fully developed studio with a big catalogue. There are no tailored solutions, only clever people who are motivated to deliver the product as fast as possible and that is the competition - who is faster, because, from the quality point, we are all using, more or less, the same tools. And in the future maybe the AI will design a pipelines for us, but I am a bit worried of too efficient optimizing because my experience tells me it usually breaks over the labor's back.

Difficult to answer where to point you, but there are many python in houdini tutorials out there, free and commercial ones. The new python state, after you check the Paul's stream, Jeff Wagner is covering a lot about in one of his masterclasses (sorry, no time to dig links) when they released the states, etc. Just now in the last masterclass Wagner is covering ten new stuff in 18.5, he demonstrates some user interface, as I understood, for the handles authoring/adding, but I am too busy to start to dig into it. You will benefit most if you do your own research, then you can go trough a lot of scenarios and find exactly what you need.

Python, yes, but not as a fast trick. You will have to dig into it and learn scripting, python states, etc and possible pack your own transform hda (or are you thinking of the edit sop? The same). To me, this video was an eye opener https://youtu.be/nS4FSMEas-I [youtu.be] but since you think that pressing M takes too much of your time for the fast editing, maybe your time will be well spent in learning python, since python is the place to go in optimizing ones workflows and user interfaces.

Hi, do you understand the concept of attributes, different contexts of the geometry, where do attributes live in geometry and what is their purpose? Also, it would be easier to help you if you would specify more precisely what are you trying to do instead of asking general questions like this, because Houdini is a whole operating system which I advise you to take a good look before diving into more concrete problems. So from your question it is not clear, how did you import your fbx, how exactly are you getting rid of extra meshes and where are the materials written, are they attached with separate nodes, do they all live in a material network? How much can materials be in the material network, 10, hundreds, thousands? Why do you need even to delete anything? If you already removed or keeping the geometry you don't need, why is that extra material even bothering you? Is it maybe bothering that it clutters your network view, or are you worried that the bunch of unused materials are just taking your memory space? If you are worried about the memory, you should remove all the attributes which you don't need for the final output and there is not going to be any memory problem or unused material.

When you are talking about the modeling as your message suggests, the first method is the only method in Houdini, because the modeling means that you are operating on some geometry/surface using Surface OPerators (SOP-s). Putting objects in the hierarchy is not modeling, that is, I don't know, kitbashing or rigging? Modeling is manipulating geometry, not rearranging already modeled geometry containers in hierarchy. I prefer that method since it is more logical to have all the elements of one object inside of one object. Jumping up and down from node to node is sometimes tedious task and my operations are hidden. Working in sops is more intuitive, like a circuit board, you can see all your operations laid in front of you. That is why we love Houdini, no? The problems you list under the first method are the rigging problems, not modeling. If you have problems in figuring the orientation or you are worried about making a lot of groups, that is only when you start to think how the thing is going to move, and with that we are going into the rigging territory. Btw, using python tools instead of vex is also confusing statement. Vex is vector operating expression used to manipulate geometry, shaders, attributes in general and the biggest advantage is that most of the operations in houdini are already vex based so it is easy to adopt new algorithms or ideas in manipulating data on a fly. On the other hand, python is a scripting language which you can use to optimize certain user interface scenarios and automate lot of your tasks, but the python script is not the same as performing a multithreaded expression which operates in parallel on every point of your geometry, for example. So even comparing python and vex is misunderstanding, since it is not the same context.

I guess I have to have my phone charged so I can watch it after my swimming pool on a way home!!

harryabreu
Hi N2e when I tried some character animation all character reaction became so much slow and all the time I need to use flipbook to check the animation…
Any deformation in a character face is very slow too.
I Know 3 packages one is Blender other is Maya and now a little about Houdini… no software is perfect simple modeling in Houdini is not good but of course in direct modeling Blender can help but if Houdini i so powerful in a lot of aspects for game and movie I missing some help in character animation.
But I start in Houdini 15 and step by step Houdini have been improved.
I complain about it only to remember SideFX that us character animation still love Houdini we only wait for a opportunity to jump in.
To make things in Maya and finishing in Houdini is a “mantra” but do more in Houdini will be better.
I and a lot of users complained about render some time ago and now we have Solaris and Karma… Ok now I think only one step more…(laughing).

When it comes to the char animation, maybe we indie artists rely too much on the technology here, and not enough on the skill and imagination. I produce also in the stop-motion and 2D char animation and in the stop-motion there is no much choice other than working frame by frame and then playback that animation to check your movement. If you want to change something, most of the time you have to reshoot the whole shot. In the 2D, unless you use some 2D puppet style animation, you also have to draw, frame by frame, and then you can test it, after that, you can fix every drawing. I have no problem doing flipbooks to check my movements, and that is the case even for the FX shots, particles, etc. The realtime is still not fully there yet. There are workflows (some are covered in fantastic Goldfarb's rigging tutorials on this site)where you use the proxy geometry which represents all the movements and speed things out quite a lot. In the hands of experienced animator, this is not a problem. If you think how Lasseter did his first CG movie, which didn't have any of the fancy deformers and skinning, but they all had to make tools as they go, they were still able to make pleasing char animation. Saying all that, I would greet performance improvements in working with characters in H, since we have packages like Maya which can handle that much better, so obviously, the technology exists, but it has to be brought in Houdini, as well. When I look at the demos where Pixar artists are able to have responsive characters, fully textured and even with the fur, just let's keep in mind that they are multi billion company with enormous resources for R&D. One day this will also be available to indie artists, but so far, these tech is there to optimize industrial production of the CGI, therefore, not indie. Once when you go and play in the bigger league (for example, they commission your show for Disney+) then it is another story, of course.

https://www.sidefx.com/docs/houdini/nodes/sop/pointsfromvolume.html [www.sidefx.com]
Try points from volume sop? Maybe for each primitive, make a small extrude (a bit bigger than point separation) in the loop and add points from volume on each primitive would do the trick. I haven't tried. Good luck

p.s. another option crosses my mind, use divide with the brick separation with a small brick value, until you get the density of points you want, select the edge points and keep only the ones inside with the add sop later. This should be more or less what you are looking for according to your illustration.

p.p.s. I see you are using the quad, but each quad is divided into two triangles, and then apply the bricker polygon on each triangle separately, with the rule that one of them has an angle of brickered polygons rotated 45 degrees. Then you will have exactly the effect you need. In some crazy maths, you could even read orient of each poly and then adjust the divide node according to that information and have points mathematically precise, but I think you can get away with the approximate method.

If you want to keep your primitives colored and kept the colours not bleeding between the primitives then you should promote Cd to vertex attributes.

I am not going to be able to answer that but I am just curious, why do you need a point function to fetch the P attribute from the first input (your geo number is zero, which is the first input) when you can simply use @P, I mean, the pos variable and @P should have the same value, no?

Ahh, it makes sense, so he is struggling with a homework.

Tony-Wu
I'm learning the vex expression recently. In chapter 14, I met a big problem. How to use vex expression to do this effect as the picture describe?

What exactly, under 3? because I find it difficult to understand what is written under 4. and 5. Also, what chapters are you are referring to? Chapters of some book? Documentation? I have no idea what are you talking about, mate.
Cheers

atonvision
Slide mode

https://www.sidefx.com/docs/houdini/nodes/sop/topobuild.html#build [www.sidefx.com]

where i can find this part in houdini UI?

The toolbar shows up when you press Enter in the viewport, left up, like in the video.

It must have been the wind!

Hi Midphase, what I did, and it is probably not the only way, to use the name attribute on the primitives. If you are buying your files from the services like Kitbash3d they usually come with properly named primitives. I would then import the fbx with the file node and use this attributes to separate objects. My setup was also to group objects into separate buildings, because my fbx came with models which are separated by materials rather than use. So for it to work I had to use some string operations in the attribute wrangler, create a new attribute with the new string value and then use for-each loop to run the process for each of the primitives, in my case, pack primitives which belong into the same type of assets into a packed primitive, which in return gives you a point. The point number can be used in the group parameter of the split node (or delete, or blast) and then you can use that to make your layout in Houdini, or export this into separate fbx files if you want to use it in the game engine. It is strange at first, but once you set up the first one, you can use the same setup (or even make an HDA for that) for all of your collections.

illusionistics
Whew! Managed to get it working by deleting the content of the geo1 node inside the subnet created by the Skinning Converter then pasting in the original unmodified grid I was using for the example cloth sim and recapturing that to the bones (which are all properly keyframed by the S.C.). Had to toggle off “Destroy Existing Weights” in the bonecapture biharmonic node and it was a bit of a fuss to get all nodes to recook, but I´m fine with that
If there´s a better way though, please let me know …
Have Fun & Cheers!

Hi, there is rumor that this is a bug, I will update you because I am struggling with the similar problem right now. Your system works. I don't know about Destroy Existing Weight, but what I did I just used the first node in the geo1 and wired it to the timeshift to take just a first frame and then I was able to click capture geometry in the shelf which then lets me choose the root bone. Then in the UE the animation works and it is captured, however, the import prompts the warning listing all the bones in the capture saying it is not able to something, I don't understand what, and I am not sure if this warning will effect the Unreal Asset or not. (I am very fresh in the UE4)

OmidRG
I guess I didn't deliver my question correctly. I'm aware that there are tutorials for rigging in Houdini. I just wanna know, is it worth the time? Or should I use Maya which everyone says is the beast of character animation ?

it is worthy to learn rigging in Houdini because it will give you a solid foundation for understanding how H works. In the case you need to learn for the job, then I guess you have to find out what is the pipeline you are going into. But these days it become worthy to go for the Blender too.