Jump to content

Viewport Rendering Problems


Recommended Posts

I'm trying to render 8 viewports on a sheet layer. I'd love to use hidden line rendering, but I can't get any type to work at all.

The method I've tried is:

-Highlight a viewport

-Choose a background render option in "object info"

-Click update in "object info"

Each time I've done this, with a vast array of settings/tweaks, I get one of 3 errors:

-"Out of Memory"

-"One or more operations were aborted due to lack of memory."

-"One or more viewports could not be rendered. Try reducing the sheet layer dpi or the size of the viewport with respect to the page."

I'm attempting the rendering on the following rig:

-Intel Core2Duo E8400 3.0Ghz

-NVIDIA Quadro NVS 290 (256MB)

-4GB DDR800 RAM

-Windows XP Pro (yeah, I know 2GB process limit)

-VectorWorks 2008 Fundamentals + Rendorworks

I've tried Final Quality, Hidden Line, OpenGL, Unshaded Poly, etc. I've tried turning off Ambient Info & Environment Lighting in "Viewport Lighting Options". The viewports have been set to 72dpi for most of the attempts, though I did reduce them to 10dpi once.

If any of you have ideas or suggestions, they would be most welcomed.

Thank you all!

Edited by dovermyer
Link to comment

After installing SP3, I tried to do a hidden line rendering of just 1 viewport. It shows (2) pieces of extruded aluminum tube with a custom cross-section & a solid end-cap. Fairly basic geometry. After 20 minutes of being pegged at 1,675MB RAM usage, it threw up the error "One or more operations were aborted due to lack of memory."

So, any other ideas are greatly welcomed. I can't believe that this system isn't good enough. Perhaps there's a geometry feature (fillet?) which is causing the problem? I'm fairly new to VectorWorks, so I'm not sure what pitfalls might be common-knowledge.

Thanks!

Link to comment

The problem being mentioned is a memory leak in VW 2008 SP3. I have 10 GB of memory and I can barely run two section viewports (OK they are huge and go through the entire building, but so what) in a row (FQRW w/ Hidden line) without a crash, memory out message or some strange rendering.

I have reported this as a bug, but there has been no response.

What happens is that not all of the memory is released after an update. Do this enough times and VW gets very mad. I can monitor the system as as VW gets closer to 2 GB things go bad.

Here is part of a console log to show you what I mean:

6/21/08 3:29:07 AM VectorWorks[541] VectorWorks(541,0xf42ce000) malloc: *** mmap(size=27840512) failed (error code=12)

*** error: can't allocate region

*** set a breakpoint in malloc_error_break to debug

6/21/08 3:29:07 AM VectorWorks[541] VectorWorks(541,0xf4350000) malloc: *** mmap(size=27840512) failed (error code=12)

*** error: can't allocate region

*** set a breakpoint in malloc_error_break to debug

6/21/08 3:29:07 AM VectorWorks[541] VectorWorks(541,0xf43d2000) malloc: *** mmap(size=27840512) failed (error code=12)

*** error: can't allocate region

*** set a breakpoint in malloc_error_break to debug

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] VectorWorks(541,0xf42ce000) malloc: *** mmap(size=27840512) failed (error code=12)

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] *** error: can't allocate region

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] *** set a breakpoint in malloc_error_break to debug

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] VectorWorks(541,0xf4350000) malloc: *** mmap(size=27840512) failed (error code=12)

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] *** error: can't allocate region

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] *** set a breakpoint in malloc_error_break to debug

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] VectorWorks(541,0xf43d2000) malloc: *** mmap(size=27840512) failed (error code=12)

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] *** error: can't allocate region

6/21/08 3:29:07 AM [0x0-0x5f05f].net.nemetschek.vectorworks[541] *** set a breakpoint in malloc_error_break to debug

6/21/08 3:32:01 AM com.apple.launchd[138] ([0x0-0x5f05f].net.nemetschek.vectorworks[541]) Exited: Terminated

VW has run out of memory and cannot recover. I had to kill it.

My frustration with this issue is getting to be overwhelming. VW can only really use 2GB of memory. No matter what I do, the longer it stays open and the more often I update viewports (rendered or section), the sooner the problem shows it's ugly head.

I wish you luck with a fix or a reply from NNA.

RonMan

Link to comment
Looks like you need more RAM, what are your system specs?

System Specs are listed in the 1st post. I have 4GB of RAM (3.48GB in WinXP), which is far more than VW can utilize.

I'm making the render attempt immediately following a fresh load. VW RAM usage sits around 270MB until I attempt to render a viewport, then it shoots for ~1.5GB -> 1.7GB, where it gives the errors. I assume it's failing while trying to allocate a final, large memory block which is between 300MB & 500MB & that's why it's hitting it's head on the 2GB process limit.

I just don't know why it's using so much RAM to do something ,which appears to be, very simple. I can do Final Quality RenderWorks on the entire drawing w/out viewports...but as soon as I try to render just 3 of the entities within a viewport, VW sets its hair on fire & runs around the room freaking out.

Maybe you could post a file, so people could have a look. I'd suspect a geometry issue.

http://www.mediafire.com/?pmdymmrmyly (44.51MB)

Design Layer Rail is the "model space" I'm able to render just fine. Sheet Layer D Rail contains the 8 viewports. The bottom right viewport has been my primary test subject.

Thank you all for your tips & suggestions. All are appreciated.

Link to comment

dovermyer,

The Quadro NVS graphics card may be the culprit (it's intended for 2d rather than 3d work). That might explain the difference in outcomes between design layers and sheet layers.

Another culprit might be a fragmented disk...this might explain the variation in the error messages.

Finally, are any render methods successful?

Link to comment

Yeah, the hardware was purchased without my input...I would have told them to grab an EVGA 8800GT 512MB for $150 instead of this NVS 290 256MB (Not Very Strong?). IMO...you don't touch workstation cards under $200...stick with gaming cards under that.

Disk fragmentation is only at 3% and was defragmented a week ago. Though, I'll give it a try once more since I did recently do the SP3 install.

I did get OpenGL to work ONCE. I don't remember the exact settings, it was only on one viewport, and the system hard-locked 2 minutes later...

Link to comment

In my opinion, 3% is pretty high - despite what windows will say.

It's not really file fragmentation that causes memory problems anyway, it's free space fragmentation (which is not explicitly reported).

If you see lots of little lines when the disk is analyised that can grossly impede memory intesive applications.

Link to comment

I've looked at the file you posted, and have a few observations:

  • 868 MB=very large and unwieldy file
  • Some ways to minimize:
    • purge un-needed colinear vertices in polylines
    • use more symbols for repeated items (one definition=endless usages w/o memory penalty)
    • most importantly, create leaner file structure with DLVPs to reference files with standard elements

    I've been told there is no perceptible difference between For VP settings, I've been told there is no perceptible difference between "Shaded Polygon" and "Final Shaded Polygon" except the latter takes lots longer to render....

    [*]I don't know what the official recommended file size is, but I get the feeling that staying under 250 MB is a good idea. Personally I try to stay under 150 MB.

I was able to render your D Rails Sheet, but it took f...o...r...e.....v...e.......r........

Link to comment

The file size change is an important clue. As I had changed Document Prefs>Save VP Cache in order to retain renders, I increased the file size, but unexpectedly dramatically. With all VPs rendered it was over 1GB!

The root problem is that though each VP is cropped to a close-up, the ENTIRE design layer is calculated THEN cropped, and it is quite a DL, less a model than an aggregate of all pieces/parts.

I would recommend more rigorous and liberal use of layers and classes to isolate, control, and organize visibilities. I played around turning classes off and got much better results, i.e. quicker renders and much less memory usage. I'm not sure of your requirements, but I would put different assemblies on different layers (including an invisible "Reference/Derelict" which I include in every file to dump development stuff), while using classes more as you have, to control types of parts.

IMHO, this is definitely the primary issue.

HTH.

Link to comment

Great assessment! Thanks for your help Gytis. I'm on another project for most of today, but I'll definitely post my results after I've had time to clean that drawing up. I've already begun switching groups to symbols where practical.

I'm new to VW & this drawing pre-dates me. Everyone's help has been greatly appreciated.

Link to comment

I'm having a similar problem,

after getting the "one or more VP could not be rendered..."

followed by "one or more operations aborted due to lack of memeory..."

I went out and bought a bigger machine.

Mac Pro, 2 x 2.8 GHZ Quad Core Intel Xeon Processores (that is 8 of them)

14 GB 800 mhz hard drive

14 gb ram

NVIDIA Geforce 8800GT video card...

the file is 368 mb

running on VW 2008 SP3

viewport size 3'x1', custom render with 1 reflection

dpi set to 100 (only 100)

and still get the same message...

presentation tomorrow,

and the drawings are locked in the program.

ARRRRGGGGHHHHH !!!!!

anyone getting this to work?

Bill

Link to comment

Oh, and I can get the model to render on a "saved sheet" without a viewport...

and so I'm using the workaround, render, save as a jpeg, then import to a blank sheet...

well, it works better than the viewport seems to.

the problem seems to be with the final dump of information changing the wireframe to the rendered image...

back to work.

Bill

Link to comment

after a very long two days and nights...

went into "Time Machine" found the last file that rendered,

went through everything I had added, deleting then seeing if would at least not set the computer spinning (that horrid little rainbow thingie).

and after a few hours found a 3d solid that apparently the system did not like at all.

deleted it. But viewport update still ran out of memory....

and so,

set up the viewport, changed the preferences to save 3 last...

sheet set to 200dpi, custom render, 1 reflection.

saved the file, closed the file, quit VW, restarted the computer.

Opened the file, hit "update viewport"

and it rendered smoothly.

did it again for each of the 8 viewports...

finished just in time to run to the printer.

is there an easier way to do this?

Bill

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...