By dsr
#371326
JD,

Hello. Haven't posted in a while.

We are starting to have some major problems. We do not know if we are identifying bugs, dealing with corrupted files/programs or both. This may also be a Rhino issue only, although the behavior is affecting Maxwell, too.

In a nutshell, we are seeing three things we need to fix:

1. Our materials are not rendering in full on certain sides of our models. We find black patches that are rather randomly distributing themselves along certain surfaces. The same material may show up just fine on one side of a model, but be problematic on other faces. We have tried to systematically identify the issue by: A. creating new geometry; B. deleting and attaching new materials; C. creating fresh new files, etc.. The problem is not going away. Here are some examples, including some "test files". They should be self-explanatory, but basically all of the black patches you are seeing should be white or silver (depending on the material).

Image

Image

Image

Image

Image

Image

2. We are finding now that the Boolean commands are getting confused -- especially the Difference command. The command will work for a while with a given set of objects and then it will create a Union (arbitrarily) instead of a Difference. This makes the workflow very frustrating as rebuilding the geometry does not always help. This occurs with both large and small files.

3. Maxwell Render is crashing Rhino when we open it.

Thanks for your help. The sooner we can address these issues, the better.

Regards,
dsr
By JDHill
#371327
Hi dsr, going point by point:
  • 1. My first inclination would be to say that this could be due to problems with using too long a lens, as may happen when using a parallel projection, but looking at your images, it seems as though this would not be the cause. Another possibility would have to do with the creation of poor meshes, which could result from these objects being located too far from the origin, and this would depend upon the units being used in Rhino. Either way (or if it would be due to something else), I cannot say much without looking at the model here, so please upload a copy of the 3DM to allow me to try to reproduce it here.
  • 2. I don't immediately see what this would have to do with the plugin, but it may dovetail with the above-mentioned possibility that geometry is located too far from the origin for Rhino to boolean and/or mesh it properly.
  • 3. So, what you are saying is that when you click the Render button in Rhino, the plugin successfully writes an MXS file, and starts Maxwell Render rendering it, but that Rhino then crashes, correct? On that assumption, does Maxwell Render also crash? And what happens if you only write the MXS, then manually start Maxwell Render and load the MXS manually? How much memory is on the machine in question, and how much is in use when the crash occurs?
Lastly, I find it interesting that you say you are "starting to have" problems -- has anything changed recently, such as updates to Rhino, the plugin, Maxwell, or the installed OS? If there are no such changes coinciding with the appearance of these issues, I would lean toward the geometry-too-far-from-the-origin possibility, which might likely be related to a particular model, or collection of models.

So please read through these points and let me know your thoughts.
By dsr
#371332
JD,

Thanks for the quick response.

1. We will look into the origin issue. We are working in millimeters and our objects are not located close to 0,0. We are going to change that and see what happens. I am guessing this mean that geometry needs to be rebuild? Hope not, but I am guessing my hope is for not.

2. I am sending you the file (w/o linked blocks). It is large.

3. I meant to write 'Maxwell Fire' is crashing Rhino previously and not Maxwell Render. This is still happening. Please advise on a solution. I am not sure this has anything to do with the other problems unless the program in general has become corrupted?

[url=http://farm8.staticflickr.com/7458/1001 ... e138_b.jpg]

I am hoping this is a solvable series of problems. We did update Rhino recently to the latest version update, but we believe the problems began before this update -- at least the Boolean issues did. If we have to move our geometry (and multiple referenced blocks) and rebuild geometry, that will suck, but if it clearly solves our problems, then it will be worth it. If we are dealing with program corruptions of Rhino and/or Maxwell I hope we will not find ourselves entering into a prolonged period of trouble-shooting. We are too busy for that!!! :-)

Best,
dsr
By JDHill
#371333
Thanks for sending the file. I cannot tell you the exact reason for the problem, but I do find that if I rotate the model a few degrees, it is able to be rendered. This appears to have to do with voxelization of the scene, and trouble with meshes that are perfectly axis-aligned. This is the best I can do at the moment, as I have to catch a plane, and will not be home until late tonight. So I hope that helps, and I will check back again later.
By dsr
#371334
JD,

We seemed to have knocked out the problems by moving the model closer to the 0,0 point. This includes the Maxwell Fire problem (?).
I can hardly believe this and am still in the process of testing, but for now we appear to be close to solutions.

Have a good flight.

Best,
dsr

We don't use Revit but you might like to try some […]

Multiple surfaces on object

I guess you are asking about chanching a material […]

New Develepments for 5.2?

Some areas that need addressing One why can't we[…]

..... help please ..... :mrgreen: