 | | | | | |
I’m having a strange issue with LDView. See the
picture
for the display behavior.
I have tracked this down to happening during the following conditions:
* There is a rubber belt in the model created with MLCAD’s rubber belt
generator. * The belt is in a submodel of an MPD file.
If you look at the
submodel,
the belts look fine. But in the top level they are all garbled. The really odd
bit is that I cannot reproduce this behavior. If I make a model with ONLY a
rubber belt in a submodel, it works fine. The belts themselves are made of
4-4cyli primitives, so I don’t see anything odd going on in the files.
The problem is not limited to only one file. I have seen it in at least a dozen
files, all of them large models.
Ideas?
| | | | | | | | | | | | | | |
In lugnet.cad, Eric Albrecht wrote:
| |
I’m having a strange issue with LDView. See the
picture for the display behavior.
I have tracked this down to happening during the following conditions:
* There is a rubber belt in the model created with MLCAD’s rubber belt
generator. * The belt is in a submodel of an MPD file.
If you look at the
submodel, the belts look fine. But in the top level they are all garbled.
The really odd bit is that I cannot reproduce this behavior. If I make a
model with ONLY a rubber belt in a submodel, it works fine. The belts
themselves are made of 4-4cyli primitives, so I don’t see anything odd going
on in the files.
The problem is not limited to only one file. I have seen it in at least a
dozen files, all of them large models.
|
Looks like the transformation matrix for the cylinder primitive maybe
got a row zeroed out. Could be a memory issue, or a bad graphics driver.
Does it work when you disable primitive substitution? How about on a
different PC?
| | | | | | | | | | | | | | | | |
In lugnet.cad, Don Heyse wrote:
| |
Looks like the transformation matrix for the cylinder primitive maybe
got a row zeroed out.
|
If it was a discrepant transformation matrix, then wouldn’t you only expect it
on ONE of the belt cylinders, not all of them?
| |
Could be a memory issue, or a bad graphics driver.
|
This is my work PC which is a high end CAD workstation. You couldn’t really ask
for more memory or a better graphics card, so I don’t think that’s it.
| |
Does it work when you disable primitive substitution?
|
Yes! If I shut off primitive substitution, it works. Of course, the rest of
the display does not look nearly as good.
| |
How about on a different PC?
|
Same behavior.
Any ideas on how I can fix the bad transformation from within the file?
| | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.cad, Eric Albrecht wrote:
> In lugnet.cad, Don Heyse wrote:
> If it was a discrepant transformation matrix, then wouldn't you only
> expect it on ONE of the belt cylinders, not all of them?
No, it could be something wrong with the substitute 4-4cyli, Either
in LDView, or in the display list created on the graphics card. But
I'd expect this to affect the studs as well since they also use 4-4cyli.
Maybe that's a different substitution to include the top of the stud?
Anyhow, I'd say it's a problem with the cylinder because in the picture
it looks like its happening to the belts of multiple subfiles included in
the main file. Maybe the primitive substitution code has a bug that
depends on how deeply nested the file is.
> > Does it work when you disable primitive substitution?
>
> Yes! If I shut off primitive substitution, it works. Of course, the
> rest of the display does not look nearly as good.
>
> > How about on a different PC?
>
> Same behavior.
If the other PC has a different graphics card then that rules out a
driver/memory issue.
> Any ideas on how I can fix the bad transformation from within the file?
Not really without seeing the file. I suppose you could substitute the
hires 48\4-4cyli for 4-4cyli and see if that changes anything. However
the LDView primitive substitution may override this anyhow.
Another idea is to dump a POV output file and maybe that will reveal
what's wrong with the matrix. I'm not sure if that'll tell you why,
but it still might be useful for bug fixing to know what happened
to the numbers.
Good luck,
Don
| | | | | | | | | | | | | | | | | |
In lugnet.cad, Eric Albrecht wrote:
| |
In lugnet.cad, Don Heyse wrote:
| |
Does it work when you disable primitive substitution?
|
Yes! If I shut off primitive substitution, it works. Of course, the rest of
the display does not look nearly as good.
|
Please send me the file, and I’ll try to figure out what’s going on.
--Travis Cobbs
| | | | | | |
