[chimera-dev] suggestions for better STL export - remove internal, reduce atom vertices

Tom Goddard goddard at sonic.net
Mon May 9 10:42:51 PDT 2016


A number of web sources say the program netfabb can “repair" surface meshes.

	http://plato.cgl.ucsf.edu/pipermail/chimera-users/2013-May/008826.html

	https://thingiverse-production-new.s3.amazonaws.com/assets/35/64/72/a2/77/Eduardos_Guide_for_3D_Printing_Proteins.pdf

  Tom


> On May 9, 2016, at 12:32 AM, Jakob Suckale <jakob.suckale at uni-tuebingen.de> wrote:
> 
> Dear all,
> 
> A follow up on Tom's suggestion. I have tested MeshLab. It does indeed have functions that can be used to reduce internal surfaces (https://www.youtube.com/watch?v=9EHJQd3SREE <https://www.youtube.com/watch?v=9EHJQd3SREE>) but it's far from a good solution. It is based on assigning a value to vertices based on visibility (Ambient Occlusion) and then removing the ones that have low scores. In a ribbon model, however, it often opens holes for example on the inside of a helix and it leaves many internal structures untouched.
> 
> All the best,
> 
> Jakob
> 
> On 15 April 2016 at 21:29, Tom Goddard <goddard at sonic.net <mailto:goddard at sonic.net>> wrote:
> Hi Jakob,
> 
>   We discussed the excessive STL vertices in atoms and bonds last week and I put a fix into the daily build and release candidate. 
> 
> 	http://plato.cgl.ucsf.edu/pipermail/chimera-users/2016-April/012134.html <http://plato.cgl.ucsf.edu/pipermail/chimera-users/2016-April/012134.html>
> 
> Apparently you are not using those versions since your images show the high vertex density.  I’ve attached an image using the current Chimera release candidate that shows reasonable vertex density.
> 
>   Unfortunately eliminating internal structures and stitching together meshes for atoms, bonds and ribbons at the seams is very difficult.  Yes $10000 CAD software packages can do it but we don’t have the resources to solve that problem.  Since every 3d printer has to deal with this problem I expect the printer software to be able to handle it, but perhaps it is too hard a problem for the 3d printer companies as well.  Perhaps there is free software that can eliminate internal structures and stitch mesh intersecting edges together, I have not researched it — I’d start looking at MeshLab.
> 
> 	Tom
> 
> <stl_mesh.png>
>> On Apr 15, 2016, at 4:33 AM, Jakob Suckale wrote:
>> 
>> Dear Chimera developers,
>> 
>> Chimera has the great feature of STL export, very useful for 3D printing models for research and teaching. May I make two suggestion to improve this feature even further.
>> 
>> 
>> 1 - Remove invisible internal and therefore unnecessary structures.
>> 
>> Chimera generates a lot of internal structures that complicate printing and are pointless for STLs. This includes inside surfaces at the end of ribbon cartoons (red below).
>> <Screen Shot 2016-04-15 at 13.24.44.png> 
>> as well as copious amounts of invisible internal surfaces for wire and ball and stick models.
>> <Screen Shot 2016-04-15 at 13.27.36.png>
>> If these were automatically removed, STLs would not only be much smaller but also much easier to print without remashing.
>> 
>> 
>> 2 - Reduce the number of vertices for atom models. 
>> 
>> Ribbon conversion to vertices is already good, but as can bee seen above atom models are converted using far too many vertices inflating the resulting STL files. I'm aware that this has been discussed and listed previously as a bug but the problem remains and makes for difficult to use STLs and bad 3D prints.
>> 
>> Best regards,
>> 
>> 
>> Jakob Suckale, PhD
>> Lecturer in Biochemistry
>> University of Tübingen
>> +49-7071-29-73363 <tel:%2B49-7071-29-73363>_______________________________________________
>> Chimera-dev mailing list
>> Chimera-dev at cgl.ucsf.edu <mailto:Chimera-dev at cgl.ucsf.edu>
>> http://www.rbvi.ucsf.edu/mailman/listinfo/chimera-dev <http://www.rbvi.ucsf.edu/mailman/listinfo/chimera-dev>
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-dev/attachments/20160509/9843abe2/attachment.html>


More information about the Chimera-dev mailing list