Make2D gives me strange results!


#1

I love the make2d command and usually it works just fine. Now though its totally flipping out. It creates these strange things outside of the viewport and skips so much stuff, ill explain better with images.


(Pascal Golay) #2

Hi Ameda - can you send the file, with the view that you are using, to tech@mcneel.com - if the file is large, please send via rhino3d.com/upload. Also, you might test the same file using the V6/WIP build (https://discourse.mcneel.com/c/serengeti) and see if it works any better - Make2d has had a lot of revisions.

-Pascal


#3

Not saying that this is the case, but https://wiki.mcneel.com/rhino/make2dtips says the following:

Make2D doesn’t work with objects behind the camera – so make sure there are none selected for your scene. For views where the camera is inside the model, you may need to draw a plane that represents the camera location, split the model there, and delete/hide everything behind the camera.

Could that be causing issues?


(Pascal Golay) #4

That is my guess of the moment as well…

-Pascal


#5

Sorry for not replying but I attached the camera-look in the image and in that one it is quite obvious that the camera is showing the stairs etc but still it is screwed up?

Anyways, I am having the same problem now, it adds hidden lines and shit even though I have not enabled it. Here is what I get and the look (might be a slight rotation but not much more) I marked some of the hidden lines in red just to show that it is showing stuff that is not visible.


#6

Btw, the pen view now gives strange results while the shaded model is just fine. Here are some images


(Willem Derks) #7

Hi Ameda9,

Could it be your model is far from the origin, or in other words, could it be the coordinates for these objects are very large numbers, somewhere in the 100-thousands or millions. If so this is likely an issue with floating point rounding errors.
Can you state the units your working and file tolerance as well as the position of for instance the stairs?

-Willem


#8

Yes and no. The scale is like 200m x 1000 m and while copying it to another file, it went from meters to mm. However, I thought of this and scaled it 0.001 and yet it looks crap. Also, for me it does not really explain the smaller scale thing I had, I have a model of a smaller area of 150mx20m or so and it also have the same problem. Also, posted some images of a difference between penmode and shaded mode. It would be amazing if i could fix this before the 29th, I have been using make2d for architecture for years, it always worked flawlessly.

But what you say about coordinates though, its very very far from 0,0,0 but sounds crazy that it would matter? I will try it tomorrow though and I really hope the fix is as easy as changing it closer to 0,0,0. That would make me very happy!


(Pascal Golay) #9

Hi ameda - if you can send us the file, we can take a look. Either here or to tech@mcneel.com.

-Pascal


#10

Ill send it now! Thanks!


#11

Sent both files.


(Pascal Golay) #12

Hi ameda - the objects are very far from the origin as the file opens. Still testing…

Yeah, it looks like you need to move things to the world origin. The way the files are, the extents are gigantic, even if some parts are on the origin - for the crane file, the objects are at

7998337326.7244663

units away…

-Pascal


#13

I had the same problem last time. Is it possible to embed a message box when file opens that tells the user if some of the objects are too far from the origin? What is the tolerance?


(Pascal Golay) #14

Hi onrender - that might be feasible - I don’t know what the magic number might be. For now, here’s a python you can run if you are suspicious - it will tell you the distance from the world origin to the farthest corner of the scene bounding box.

SceneExtents.py (486 Bytes)


#15

Wanted to say that I moved it to the 0,0,0 coordinate and everything went perfect and I delivered first part of my thesis just earlier today! So thanks support and thanks for this thread, didnt know that the coordinates mattered but obviously this was the problem.


(Pascal Golay) #16

OK, good, thanks for the update.

-Pascal