David,
That would be a valuable resource. Thanks.
Steve
bpmufx,
here is a website http://web.axelero.hu/karpo , check it out, it supplies a really low cost object converter, over 400 file types. you may be able to convert to a file which suits mastercam v8.
There is a very fine ballance between file size and development in this arena (At least with our modern PCs). When I am planning my way from point A to point Z every step needs to support later steps. If I were to put all of the detail in the leafs before rigging and posing them, I plug the pipeline and remove a number of options for manipulation. Manipulations that would have been reasonable if the surface was more light weight. There are a million ways to come to an end and each project is different. I try to build the surfaces in my head before I put my hand on the mouse. It takes a while using a tool or method before you can safely realy on it as a player in a preconcieved pipeline. Most methods and uses for tools come to me when Im not even using the computer.
Each leaf Started as a network of NURBS patches. I moved the leaves in and out of Zbrush in 4 seperate groups. ZB was toward the end of the pipe.
Will ZB be able to replace Maya in my pipe?? Not a chance… Just as Maya will never replace ZB. Once the new version comes out, Ill have a look at it but ZB is a long way from being a “Be All” modeling app… Dont get me wrong, I love ZB for what it does. Its got a lot of unique tools that I never expect Maya to have.
I think software in general is changing and we need to abandon the idea of broadly focused applications. You come from the CAD CAM arena so you should know… CAD CAM spells it out… What are those two things doing together? They are creating a lot of redundant work for programmers to deal with, thats what they are doing. I would rather toss my chips to the company that focuses on a specific function of the pipe rather than be the whole pipe… Ill make my own pipe Thank you very much… Thats why I think that ArtCam is cheezy… Teddy bear icons and wizzards. They are trying to be a pipe for a couple of industries and they will eventually pay for that. They have been boiling down thier product to the lowest common denominator of a user and they will eventually pay for that too. The next gen of kids coming through school are going to have been using computers since they could talk. Most of them are getting exposed to RP, CAD, animation and CNC in Jr High and High school. These kids will be wise to the likes of expensive industry specific software and they will be able to find thier way around. Marketing ease of use is no longer going to be a key in the door of manufacturing because intimidated baby boomers will no longer be writing the checks. Functionality and stability will rule the day. I would even go so far to say that eventually software will be compiled by the user. Look at Visual basic.NET. Make your own programs by dragging and dropping. Imagine a general 3D platform that everyone used from CAD to GIS. You would be able to download, buy or subscribe to specific functionality for your own personal needs. Eventually…
IC
IC,
Thanks again for another thought provoking reply. If there are a million ways to do things, there must be a million things to learn before getting to the goal of creating useful things on a router!!!
You’ve given me more than enough to start me on my way.
I’ll only say that it is a pleasure and inspiration to see your work. That way, now or in the future, if you think of any other pictures of your work or descriptions of your working methods that you wish to share, rest assured that they will be gratefully received.
Steve
Hapenby,
Im glad that I can help. You are one of a handfull of early inovators that are crossing the line from manufacturing specific software too the power found only in animation and CG. I have been here for a while and im starting to see the trickle. Its about time.
IC
Boy the passion on this topic is really nice to see. I didn’t realize there were this many of us into the postwork of 3d model data.
ullach- I’ve been rendering the X to NURBS function for 24 hours now, it’s almost done. At this rate not confident that this is the way to go. Keep you posted.
Also, sent an email to the link you gave earlier, thanks.
routb- Its a good thing we’re not neighbors 'cause we’d be talking about this all day. I totally agree with your statement about Rhino not having the ability to manage a parts history. In fact that is true of most hybrid modellers. Ashlar Vellum comes to mind. It really pisses me off because with this comes a higher risk of bloated files where everything can be at risk of slow performance or worse, corruption.
In CAD, Assembly modellers handle features in way that is nondistructable and editable through the life of the part.
I only bring this up because different apps handle the basic problem differently. But the hybrid apps seem to simply avoid this kind of issue.
In the end I think that if you’ve gor the $$ to create the dream toolbox, At one end I would have Maya as the main 3d app, with Zbrush next, then Rhino in the middle with Inventor at the other end for Parametric Solid/Assembly Modelling.
My problem lies with the fact that this pipeline only goes one direction at least to Rhino. But since Rhino is not an Assembly modeller, I cannot justify it’s use as such. However I also can’t get Mesh data into Inventor where I often need it. I only use Rhino as an awsome file translator and also to perform some specialty features not available in a Parametric modeller.
I guess I’m off point here. But I think this is a great forum for us to help iron out some generic compatibility issues with the two really different worlds of modelling. We are really the bastard children trying to force animation into manufacturing and vise versa.
Sorry for the wordy post.
Kinda funny… We found a little hide away in the troubleshooting forum… I wonder if the guys at Pixologic would give us our own area… What would it be called?? ZB Special Applications?? As much as I love crazy ugly monsters, would be nice to have a place for ZBs other applications… ++ I know several times I have recomended ZB to people in manufacturing and they were confused by the demographic in the main forum.
IC
Hey routb, I’m with you. I was thinking the same thing when I was writing the last post.
Where do you buy those types of end mills routb?
Tooling??
CET direct for me… Love those guys…
They do ultra long 2 flute spiral up carbide ball end mills. Good prices and nice tools. Fast delivery…
If your doing mini mill work with Wax, aluminum or butterboard, I use anteres tools. They are also very reliable and fun to deal with… Dont buy the ball ends from these guys, get them at CET. Antares has made me 7.5 deg. SA with a .003" flat tip… They also do prismatic tapered ends that work well for wax, bone and stone.
IC
ullach-
1- My file took somewhere around 36 hours to convert using x>nurbs. Pretty good to know about it, but with files of a high poly count, too much overhead. Granted I could lower the poly count, but on the other hand V9 can do the existing file inside of 30 seconds as an STL, using routb’s method.
2- Even after x>nurbs was done, I still didn’t have any luck with generating a toolpath. Maybe you can shed some light here. My thinking is: great, I just converted polygons to nurbs but that only created a zillion tiny nurb curves instead of poly edges. Point is: still no actual surface geometry. I don’t see what part of the surface toolpath options provide for reading the nurb curves, instead of surface geometry.
What am I missing here?
bpmufx,
i’am still not convinced that stl is the way to go(using mastercam( although the only way at the moment)) they are to cumbersome and of high poly count, speedwise v8.1 and my system…
amd fx57, twin nvidia gtx7800 is fast, using certain machining params., too much and it slows way down and no flexibility on machining one area of the file( mybe I’am missing something), thats why i looked at 3d object converter.
I imported a large stl file into solidworks 2006 and edgecam v7.75 they handled the files with ease, but solidworks could not convert to parasolid as it said they it was the wrong type to convert(parasolids are light on mastercam and allow individual faces and edges to be machined). mastercam makes heavy work of large stl files, mybe zoltan at 3d object can come up with a solution.
As was mentioned in earlier posts, horses for courses, tailored apps to achieve what you want, i’am new to cg and it was clear from the start it is forging ahead in innovations which can be used in real world solutions, and at an affordable price, eg carrara5 pro($400) imported a ARCHICAD obj file of 75,000 polys in under 4 secs with full detail and textures, 3d max could not handle it, closed the app down with a submit a complaint popup, maya could only bring it in as one solid, both 10 times the price.
What is needed is a low cost cam to only generate code and leave the modeling to cg packages.
bobcad does a app called bobart artistic to manufacture cadcam system
http://www.bobcad.com/
you can download a demo version.
ullach,
I’ve got a seat of bobCAD also. Already been down that path. What it does is turns graphics data into vector data within levels of grey scale that you set.
Not real great results compared to using ZBrush to create a hight field from alpha skinning.
Can you elaborate on your earlier post: what entities would MasterCAM be using as a surface when converting edges to curves?
Cool posts! I did lost wax casting a few times and loved it. But I don’t have the time for the carving any more. So… What would be a nice 3D Router with reasonable precision to do fine work in Wax and Wood? I don’t see Metal coming up for me any time soon…
Lemo
lemanodo- check out Techno-Isel
They make off the shelf systems. There’s plenty of folks providing components to build one yourself though.
Check
Hey guys,
I noticed that there are some folks trying to convert Polys to NURBS… Let me start by saying that there is no automatic way to do this… If you absolutely must, I would suggest Raindrop’s Geomagic, Paraform, Rapid Form or a plugin for Rhino called SMURF… All of these are used by engineers to convert high res 3D point cloud and mesh data from laser scans into Aero space quality NURBS surfaces… They are pretty slick and most are very expensive… SMURF is the cheapest but still costs about $700. There are more manual ways to build NURBS from mesh but it takes a long time and the results are not great.
That being said, STL STL STL… There is no reason to convert unless you are going 5 axis. If you dont like your CAM for doing STL, there are several CAM apps available for much less than the likes of Geomagic. DeskProto is about the cheapest CAM I can think of. I would toss in a few hundred extra dollars and get the trimmed down version of Visual Mill. Andrew Werby at www.computersculpture.com has about the best deals for CAM and minimills that I have ever seen. He runs an underground railroad to get tech to us artists at nice prices… Check him out.
IC
routb- I’m guilty as charged in my need to find a solution for converting polygons to nurbs.
With the type of work I do, it’s not a matter of modelling then reproducing it, end of story.
In most cases, I have a model that needs to be used further downstream in the pipeline, further modified to work there, then it can go to reproduction.
The actual method of reproduction is irrelavent to me until I get the engineering aspect worked out.
An example is that I did a fusilage that was to broken in 1/2 down the center, lengthwise. This was shelled out for a .040 thick wall. Then it needed features added inside the shell surface to accomadate OEM components.
This example represents two major issues:
1- being able to create match molds for pressing the .040 carbon fiber layup. Granted this may be done in Maya, but it’s really better suited for a parametric solid modeller.
2- being able to import CAD models directly into the assembly from the vendor is huge! This cannot be done without translation work as the file type is always a solid.
So, for me, I’ve been able to get by, but I know it could be a much better Pipeline if there was an affordable translator from OBJ > surfaces, and why its important to me to be able have my pipeline work “bidirectional”.
Currently I am stalling on a small project. And this thread might actauuly be in the right place fir the right question!
I imported a .stl into bodypaint and exported it into .obj. So far so good. Now I loaded it into ZBrush to detail the mesh with PM… BUT it is the typical triangulated mess… It doesn’t get muchbetter even if I subdivide the hell out of it… Is there a method to create a new polygon mesh come back to quads? Not with ZB probably… I tried Unified Skin. That is pretty much what I would like to achieve… but… cough cough… The result is to ‘organic’…
Cheers
Lemo