I’ve managed to get it down to this with sub-d modelling. Not sure if it’s optimal but it seems easier to fiddle around with the imperfections. (I’m not modelling the exact version of the goggles)
Because most demos don’t have a long temple like this goggle:
I’m still struggling to find a way to smoothly create it. Do I just take the edges of the frame and start extruding bit by bit? Or do I create the long temple on the “right” perspective then bridge the two surfaces to create the curve?
i think you are doing a good job. SubD is in my opinion suggested too aggressive and too often for any kind of modelling of such sorts. with nurbs you can easily do this either just probably faster if you have some experience and way more precise.
I always recommend, model in box mode, evaluate and refine in smooth mode…
if your nurbs is a mess, it’s likely your Subd box mode (tab key) is really tortured.
I feel like you are on the right path and it could be as simple as some detangling of your box mode to get from good to great here.
the slide command may be useful to help clean things up.
you definitely have the idea on how this stuff is built in subd.
you can always build separate elements and then bridge them or stich them together, that is a totally valid way to do things, you just have to be very aware of your edge and point counts as bridge will only work for same=same edge counts.
Agreed for the final models, but for concept and the iteration needed for fit models, the ability to modify and squish a subd model around quickly is tough to beat.
Once all the prelim stuff is done in subd I would likely re-build the final model in nurbs.
(Unless the subd model was rockstar…then I’d just convert to nurbs and ship it. Especially since soft silicone models don’t need to be class-a.)
i really see the advantages in SubD only in stuff which is very difficult to model with conventional methods. maybe objects with a more biological context even though that may sound very much like a cliché but the more experienced one becomes with Nurbs i would say the smaller the gap.
also there are limitations which cause further complications or impossibilities. sure some geometry might not need and it of course also depends what you need the geometry for, but remodelling the SubD or salvaging it for a proper Model does not make much sense and sounds like wasting your time, then rather do the real thing 2-3-4-5 times which is with Rhinos abilities not that hard to accomplish and still faster than kicking SubD’s to and fro. but that is just my humble opinion.
The decision which modeling technique to use (nurbs or sub-D) should largely depend on what the intended use of the model is:
Do you want to illustrate a design idea, or do you want to create cad models dedicated for tooling/production? (to name two examples on different ends of the spectrum)
Generally, the closer you are to the physical product in the design workflow, the greater the need for nurbs surfaces.
I appreciate your comments, I’d offer these thoughts-
Subd is very very fast and very very flexible.
In experienced hands Waaaay faster than nurbs. for a product that needs to be fit tested like goggles, there are many iterations (sometimes dozens or more ) of fit test models.
The single surface structure of subd is perfect for iterative work where you need to scootch something a bit here and there to dial the fit in perfectly. Once you have it nailed, then Nurbs is the right tool for finalizing the model (or use the subd if its built well)
in the end, to each their own…
more tools, more options, more capabilities it all layers to make us all better at making the models we get paid to make.
if you are disregarding the precision and need the model for something else, then something what actually could help improving the speed would be if multi pipe would have the option to rotate the facets, right now the edges it creates are orthogonally facing the cplane which is in most cases not helpful. i am sure this was picked on at some point i believe i remember it coming up, no idea what the status is. but otherwise i would have to disagree, but as already said that is just my humble opinion.
would you be so kind as to provide an example (please start a new thread so we don’t hijack this one and @mention me.) I can use to write up a bug report? Happy to try and help tune multi pipe to be more useful.
So I’ve tried using NURBS method to do the modelling. There’s definitely less fiddling compared to SubD but I’m still struggling to get the perfect shape. Here’s the progress so far:
remember, you can use nurbs for what nurbs is good for and subd for what subd is good for… the gasket itself is a perfect subd model example.
The two systems work very well together.
also if you want specific fillets, crease your subd edges where you want fillets, then convert to nurbs and fillet them after you have made all your adjustments.
Pro tip-
Need to iterate? keep a copy of the subd parts on a hidden layer in case you need to change later-
also remember you can project subd verts to other objects-
this will be helpful for perfecting the fit of your gasket, assuming you have a scanned fit model for the head.
I tried the MoveUVN command on a circle. The corner points didn’t move as I expected. It looks like the corner points only moved in the y-direction and not in the x-direction.
The same problem with half a circle and an ellipse
Is this the expected result ?
I am sending this message from my phone so I can’t send a file to show what happened.
