Open discusson = what is missing in rhino?

Hi there rhino users. first of all, pardon for my bad english, maybe this is not a welcoming topic, I have addressed this issue on other forum associated with rhino (such as Grasshopper3D) and would like to get some opinion from you guys in here…

I hope this does not bring any hatred or any bad feeling as I just want to share, what I think is missing in Rhino.
the reason why I am opening this topic is probably because,as rhino users I want to get the best out of rhino. I want Rhino to be the peak of “visualization” in the next few years, as I have seen rhino community is very very supportive and the development team is open for critics and they do care to make rhino much better in every release. so I will just cover some few points that I think is Vital to every visualization platform =

1. Rendering (high quality Image)
to be honest, in terms of rendering, rhino is always falling behind other 3d platform (name it 3dsMax ,Maya or C4D).

Ive been using Rhino, 3dsMax and Zbrush for 7 years. the first 3d platform I knew was 3dsmax and one day I was introduced to rhino by my friend and was amazed by the flexibility to work with any types of geometry inside rhino (Rhino+Tspline+Grasshopper)
ever since then most of the modelling I do is always inside rhino, and then for rendering purpose I export it to 3dsmax and use Vray inside 3dsmax to do the rest of the job. depending on what kind of geometry I am dealing with, I had to use Zbrush since both Rhino and 3dsMax does not support Pixol.

after doing hundreds of rendering I realized that working in 2 - 3 different platform really disturb my efficiency, especially when we have to deal with the client who demands changes every weeks, exporting can become a big mess when the deadline is very tight, it is so hard to keep the model “tidy” when u have to keep transfering data between rhino and 3dsmax.

I did try to use Vray for Rhino, I have to say, Chaosgroup(vray) does not pay too much attention to rhino users, there are so many features in Vray 3.0 for 3dsmax that u will never find in Vray 2.0 for rhino.

and then Maxwell is my only choice to get highquality image inside rhino. have been quite satisfied with the result however, since maxwell works as standalone renderer, there are so many bugs and error during rendering process. (not to mention the lack of IES preview in rhino)

currently, 3dsmax and Maya always attract so many giant rendering company and Rhino however, I think is more focus on modelling only.

I am not saying there isn’t any good rendering image produced in Rhino, there are some of course, but if you compare it to rendering produced in 3ds or Maya, there is a huge difference between them.

I really hope that I am wrong. but, working alongside various 3d artist, and visualization company, they themself never like the idea of “rendering inside rhino”

other friend recommends me other renderer, however, I am more interested in big rendering platform only, since they can give more support and their online libraries are more reliable compared to the small one. (just look at how easy to get Vray material, Vray proxy, and other bunches of Vray tutorials released by chaosgroup.)

so currently I will stick with Maxwell if I need to render inside rhino, (actually I was hoping that pixar renderman will have interest to join Mcneel, although it is very unlikely,do you guys know if there is any workflow to connect rhino to renderman.? )

2. Subdivision Modelling
okay, this is actually not an issue if you have purchased T-Spline or Clayoo or installing Weaverbird. but the thing is, having to use plug-in to do subdivision modelling is a little bit awkward. not to mention the price for T-Spline oh my god, is almost the same as buying entire rhino package. would there be an implementation of advanced subdivision modelling (T-spline, catmull, etc) directly in rhino system?
having all of those inside rhino system would decrease the amount of bugs as well as keeping rhino installation simple. (currently there are so many bugs in T-spline) hopefully rhino 6 will have advanced subdiv as part of its main tools
(I am super happy to know that GH will be inside Rhino 6, and not as plug in anymore, therefor, mcneel should also pinch T-spline and merge it with rhino :smile:)

3. Pixol in rhino
I have addressed this issue several times, might as well bring it here again. Sculpting highly-detailed mesh is the best in Pixol since it can save memory and the ability to use Wacom as a replacement for standard computer mouse.
it is good to have this feature in any 3d modelling, because it can really bring modelling closer to your “hand”. use wacom stroke to cut, brush, bevel, and modify object is really a thing nowadays… my Wacom currently cannot do much in rhino and it is a pity that a complex 3d software like rhino doesn’t have pixol tools.

maybe some would say that we shouldn’t add too much to rhino, currently it stands out as the best NURBS modelling for window. but u know, not all modelling can be convenient in NURBS. sometimes you might want to start with NURBS, sometimes mesh, sometimes you have to work with subdiv, or pixol sculpting. I don’t mind if rhino installation file will have to be 3 GB as long as it can have all the stuff needed to create different styles of geometry. after all the main point for 3d platform is to produce geometry with various method and visualize it to the highest possible quality.

I think those 3 points are the only thing I have trouble with. I am having a big holiday for a week that is why I have time to write this topic. hehehe. any suggestion or thoughts please share. :wink:

These kinds of discussions have been had since Rhino 0.xxx What should be in Rhino? What features do you need, how much are you willing to pay for them?

The classic McNeel solution towards trying to fulfill everyone’s wishes has been to open the software up as much as possible so that 3rd party developers can add high-powered tools specific to one area (rendering, machining, etc). It has been very successful, and allows the Rhino platform to be tailored to user’s needs in a huge variety of fields without adding cost or features not needed by that field or user.

There are lots of opinions out there as to which renderer is “best” and whether rendering should be done inside Rhino or some other software package. There is a huge choice of render packages that do run inside of Rhino, both from McNeel and others - and new ones come online pretty often. So a large number of people must be happy with the results somehow, enough to convince developers of the economic viability of porting their render engine to Rhino. I know the McNeel developers work hard to support the features that the third party renderers need.

V-Ray 3.0 for Rhino is promised for this year, but Rhino does seem to be last on ChaosGroup’s development list.

Rhino 6 will have some form of SubD modeling as part of its tools. Whether or not you will consider it advanced remains to be seen.

–Mitch

Well, maybe. Probably. We’re working on it and it will be available in Rhino WIPs soon, but whether it makes it into Rhino6 remains to be seen. It could be the feature is delayed so much due to unforeseen circumstances that it misses the Rhino6 cut-off date for features.

Thanks David, I was just about to write something similar.

We are changing our development process a bit so that we can place longer term projects like SubD into the WIP. When we get close to releasing the next version of Rhino (V6 in this case), we will need to decide which features are complete enough for the commercial product and which ones need to stay in the WIP. If SubD isn’t ready when we feel it is time to ship V6, then the feature will not be in V6, but will continue to be available in the WIP for V7.

I stand corrected, sirs… :blush:

Mitch is right, this is a very very long debate.
But do we really need every known tool just inside Rhino?
For better rendering I’m using photoshop and illustrator, so, why not to have a complete photo manipulation ?
Or why not to have tools to write text or read e-mail?
80% of my time was spent in writing to clients.

I don’t see any lack in rendering quality compared to 3Dmax or Maya, maybe the workflow isn’t so smooth and efficient for pro users.
I don’t see any way to integrate pixel sculpting in a cad software, better work on Zbrush or 3DCoat. Of course, again, a better integration would smooth the workflow.

The openness is the solution, third-party developers can integrate their product with rhino:
The example could be Octane render, for example.

Personally I prefer a light, fast, precise, efficient, solid CAD software.

in the demise/buyout of vsr and other I would really like to see some tool to help a low cv/high surface quality workflow.
in my opinion most needed are:

-interactive controllpoint modeling with blending factors for u/v cvs and locking for rows of cv (blending = to distibute the change over a larger number of cvs)
-interactive crv/srf blend with control over degree and spans
-interactive crv/srf matching with blending for the rest of cvs
-create symmetry/ (constrain a srf/curve to be symmetrical over a given axis)
-generally a way to influence input srf/crv while using a tool (e.g. dynamically trim a srf when blending)
-more ways to check srf quality ( like AD lightlines or Alias curvature combplots)
-approximate edge: rebuild an edge to a low span value
-a tool to pull/project srf-cvs with blending

I really hope rhino gets some (more) tools to enable a workflow where surface quality is key :smiley:

1 Like

it would be a long pain if I have to wait till vers 7 to get subD into rhino system.
as subdiv modelling is a vital point in my opinion, NURBS modelling works only if you already know the final geometry that you are trying to achieve. as a designer, the needs of tweaking and adjusting things or just “concepting” the geometry is really crucial that sometimes we can end up with completely different geometry than the one we first started with.

the key point here is that during the “tweaking”, we need more tool than just NURBS tool. be able to convert NURBS to Polys then go back to NURBS for adding fine detail is a very strong selling point for rhino. RhinoNURBS + Tspline so far is the best combination in 3d Modelling, I cannot even get that same flexibility in 3dsMax. this is the Future of RHino I think, where Concepting and Precision modelling meet in a single software

the SubD development you guys are working, is it like 3dsmax Polys or more like T-spline?( well Autodesk managed to get the patent for TS so I guess Rhino team needs to come up with another name or at least write an entirely new algoritm for that :sunny: )

I really hope rhino development team consider to make Rhino more robust and really provide tools that is flexible for DESIGNER, not just 3d modeller, after all Rhino is different from CATIA, Rhino3D is a DESIGNER’s tool.

another thing about Pixol (or Voxel what ever it is called), Rhinouser normally thing it is unnecessary because they would say "oh that is only for character or creature modelling, rhino doesn’t need that stuff"
the thing is nowadays design is much more complex and even a building can look so organic that it resembles a human torso, or a silhouette of a creature, or any other similar stuff. even architect from ancient rome used human body, nature’s shape as the base concept for their building.

a little story here from Zbrush, before they tend to focus on Pixol only, but now, the latest version 4R7 they have implemented some Advanced SubD that can go well together with pixol ( check it out on Youtube, it is awesome ) they wanted to extend their flexibility of dealing with geometry, and I think Rhino also need to do this.

so, keep on developing more robust tools. I cant wait to see Rhino V6, :smile:

The built-in Rhino subdivision object will offer robust Catmull-Clark and Loop subdivision surface modeling tools. T-splines uses a different subdivision algorithm and is an AutoDesk product. Users can choose the modeling solution that best fits their needs.

Also, Rhino SubD work started with Giulio’s Weaver Bird project and will have support for Catmull-Clark quad subdivision surfaces and for Loop-Warren triangle subdivision surfaces.

1 Like

when you say “robust” catmull-clark and Loop subdivision, I cant help but to get excited about what’s coming in rhino6. Can you explain a little bit further the key tool that you will offer in Rhino6(SubD)? I am glad that rhino now starts to take subD more seriously :)) so we are not just limited to NURBS anymore. I think it would be a good idea to create a suggestion forum for SubD, so that we, users can give some inputs about what kind of feature needs to be on SubD.

I totally agree that rhino should develop their own subD algoritm based on Weaverbird. can you also point out what kind of mesh modifiers we can do in rhino6? a little hint?

I don’t want to sound like dictating you guys but I just want to share what I think a good SubD tool should be. here I will just list some stuff that is crucial in SubD and take a little bit of T-spline and mudbox paradigm into it. = (I wouldnt mention some common stuff that is obviously known to subD modeller)

1) Jumping from lower to higher SubD level — ( I guess you guys have a very clear idea of what I mean, I just wanna know how far rhino system can handle Polys? is it possible for rhino6 subD to handle, let say 50 mil Polys for a descent computer set up? I know it sounds crazy for mesh, and I know weaverbird is not aiming for pixol. however, SubD modeller like -MudBox- use real polys (not pixol) and it’s somehow able to handle 10 mil polys
so can u clarify how good rhino6 will handle polys? (mesh)

2)Creasing, and softcreasing. — this is a pretty common situation we deal all the time when doing SubD. Tspline solve this pretty easily because their “meshfaces” acts like Nurbs. adding Crease by just selecting edges or points that we want and then boom u got creases. even though it is not advised to have G0 crease for rendering, it is quite handy for some cases. softcreasing is the ultimate goal for creasing topology, adding extra edge loops to get a really tight G2 patches so it looks like creases. I know mesh works in different way, and there is an example for Mesh creasing shown in --CLAYOO-- I think they have done a pretty good job for that. :smile: so, for Rhino6 mesh, would they have direct connection to NURBS in terms of how mesh faces welded together?

3)Edge Matching —many Polys modeller are struggling with this, the ability to match Mesh Edges (whether they are naked edges or interior edges) to any given curves with given tolerance.(since rhino can calculate mesh continuity with curves there will be an option for G0 to G2 continuity) any possibility this can be achieved in Rhino6?

4)Retopologizer, Remesher, MeshRelaxation algoritm — everybody knows those are “must-have tools”. what I am afraid is that since SubD Will be a new stuff in Rhino, how far Rhino6 can “fix” meshes?

5)Compatibiity with NURBS — how advance is the compatibility between mesh and Nurbs in Rhino6? are they just linked to “export-import” function? my suggestion is that Mesh Points, Edges or Mesh “polylines” can be linked to NURBS Curve tools to the greatest extend. I am pretty sure NURBS algoritm can handle that.

6)support external devices such as WACOM — have you guys made any big changes in rhino6 so that you can utilize WACOM pen pressure and tilting angle input to modify geometry in RHino6?

7)New tool that is never been done before in SubD— I believe u guys have developed some awesome tool for Rhino6. and I don’t want u guys to spoil that. Im just curious would there be any chances we see a new types of tool you will be introduced in Rhino6? (subD)

that’s all, thanks for reading this long post. :smile:

I do agree that hybrid modeling is the future.

Here are 2 reasons why i never liked T-splines, 1 its slow, 2 i dont like the mesh quality, if youre after high quality surfacing it just wasnt good enough, ive seen far better poly results in other modelers and im sure they have their reasons (algorithm?)…

As a clarification, I just posted something that @dalelear had said. SubD in Rhino is still very early stages and we will have to wait until we have something we can play with - at that point you will be expected to give feedback on what you like and don’t like. That said, I’m sure McNeel is reading your wish list.

Ciao Riccardo !

I agree … except for light maybe
(assuming ‘solid’ means strong and not solid modeler only)
What do you mean by ‘light’ here ?

Grazie

I guess the main reason is because Tspline control points behave like NURBS. which means these control points control the flow of the whole model (star point can add creases to your entire model if the topology is not arranged correctly). imagine having a cage box 6 x 6 x 6 and each edges are connected by elastic string. you delete one surface or one edge and the entire string will pull your box geometry and it doesn’t look like a box anymore. the behaviour of T-spline is a little unpredictable at first and it is not surprising since T-spline is considerably a new type of algoritm
the good thing about T-spline is that because it behaves like Nurbs I guess it has a better integration to NURBS tool.
in T-Spline you can control tangency for each points, and adding weight to it making it very easy for patching or matching edge works. T-spline is aiming for NURBS control in Mesh topology. however it is still under development and Autodesk guy keeps adding more tools to make T-spline better.

one disadvanntage about T-spline is that it is slow and often crash if you are accidentally make a bad geometry. this is due to the fact that TS is just a plug in for rhino.
however in the new version, T-spline 4, they said they have specifically written the plug in for RHino 5 64 bit. which means less crash and more compatible with Rhino. and the new creasing ability is I have to say immensely good.
check out this youtube video for their new creasing feature on V4 =new creasing algoritm in V4

currently they are trying to develop some tools to integrate trimming and boolean in T-spline, (without having to convert it to NURBS) if really they are able to deliver that, I must say that will change the whole concept of subD.

speaking about other poly modeller, SUbD like catmull is more predictable. compared to working with control points, catmull SubD works just like sculpting in real clay. you don’t have to worry about deleting some faces or few edges it will not wreck your entire model. also by not having complex algoritm, catmull subdiv is relatively lighter than T-spline.

T-spline and subD actually is totally different, however, they can give the same result, and the same quality.