Hi Michael,

I've been continuing to test MoI over the last few days and have found that a large number scenes/models we work with cause a problem because of the limitations of 32bit memory allocation. To put things into perspective, I work at IDEO and a large part of what we do could best be described as design/architectural visualisation. A pipeline we are exploring is to use FormZ for modelling (we keep everything parametric although not necessarily all Nurbs based); exporting the resulting models as IGES or SAT for import to MoI/Rhino for mesh conversion to OBJ. The OBJ models are then taken to either Max or C4D for texturing, staging & rendering (via V-Ray). It sounds complicated but it enables us to use each tool to their strengths and when it works as it should it saves us a huge amount of modelling time. Doing architectural stuff through a pure nurbs or polygonal workflow is far too inflexible hence our use of FormZ (but FormZ's tessellation engine is horrendous hence our need for MoI/Rhino for conversion). We obviously see MoI as far more than a meshing engine but it performs the task very well.

The problem we're having in our tests with MoL (I can't send specific models due to client NDA's and such like) is that we're encountering what appear to be out of memory glitches that should easily be resolved by some form of instancing/components system. I've read in various places here on the forum that it's on the development roadmap but there's no indication as to whether it's a V3 or V4 thing.

Are you able to give any indication as to when 'instancing' is likely to be implemented? The team at the studio have fallen in love with the MoI user experience but we're worried it may not yet be capable of the sorts of task we'd ideally like use it for.

Cheers,

jm

Hi Jon, instancing is probably going to be a V4 thing. V3 has already been in development for quite a long time and I'm pretty much trying to wrap it up. There are still some things that I want to add in for V3 but instancing is not a good fit for a "tail end of the beta" type feature because it will involve a lot of UI, and changes to low level things like the display engine.

MoI was originally conceived as a CAD modeler that would put a priority on ease of use and being friendly for artists to use rather than only targeting mechanical engineers. It's still got that focus, and although it has a very good mesh export and I'm glad that the mesh export is useful for people, it's not really a primary design goal for MoI to sit at the later stages of a pipeline like you want to use it, it's more intended to be used for drawing concepts and creating early stage things from scratch. Hopefully that may help explain why the things you are specifically focused on have not been implemented yet.

Eventually I do think that MoI will get the whole set of features that you're looking for, but it will take some time. V3 will still not hit these particular areas that you have has your own priorities, sorry.

Usually you can still get the export job done with MoI despite a lack of instancing by being more strategic about exporting just some portion of your overall set of objects at any one time rather than just going for a one ginormous export of everything all at once. There are also some settings you can adjust in MoI to make a rougher display mesh which can also conserve quite a lot of memory as well. See here for some tips:
http://moi3d.com/forum/index.php?webtag=MOI&msg=5322.2

If you apply the advice there it would probably reduce the out-of-memory problems you're currently seeing.


- Michael

Thanks again for your rapid reply.

I fully understand where MoI fits in the greater scheme of things and think that it's great you've developed such an accessible Nurbs modeller in first place. The only reason I ask about instancing in particular is that I wasn't sure where it sits on the development plan.

To be clear IDEO isn't an engineering design house, we're a multi-disciplinary design consultancy - http://www.ideo.com/about/ .The part that I run is mainly staffed by people from VFX & gaming. We mainly use the tools of those industries and apply them to design consultancy projects so my team are far more at home with Max, Maya, Modo & C4D than they are with CAD toolsets. As you're probably well aware MoI is very appealing to this type of individual as it provides accurate CAD modelling without the complexities of traditional CAD software and is a great complementary tool to polygon modellers (in a large part because it exports such usable geometry for texturing and rendering pipelines).

I myself have used a variety of modelling technologies over the last 20 years, so my knowledge of CAD tools is a little deeper than many in my team but I'm still far more at home in C4D & Modo. We've changed our way of working recently because of our adaption of FormZ which enables us model with far greater flexibility and adaptability. Traditional polygon (of Nurbs) tools don't allow us to rapidly adapt our models the way we can with FormZ. However as I mentioned in my first message, FormZ exports unusable geometry and it's native rendering/animation capabilities are very weak. It's for this reason I've been looking at integration tools for our pipeline that can take IGES or SAT files and make them ready for Max, Maya, C4D etc. But being a creative studio I ideally wanted a tool that creatives warmed to which isn't necessarily Rhino!

I still believe that we're likely to purchase a few MoI licenses for the studio, especially for individual modelling projects that can be completed more efficiently with a Nurbs modeller but it is likely at this stage that Rhino will be a better fit for our pipeline tasks.

I really look forward to seeing how MoI develops in the coming years.

Cheers,

jm