[Chimera-users] level of detail (vrml)

[Damien Larivière]

Many thanks YZ and Tom for your answers!

I was not very precise in my previous messages. I did not clearly say 
that I was looking for 1/ a real-time viewer 2/ working on OSX.

I composed a scene with almost 300 ribosomes (isosurface made of  33 000 
triangles) and transformed it into the OSG format in order to use 
real-time viewer based on the OpenSceneGraph 3D toolkit. Free OSG 
based-software like LifeExplorer or OSG Director enable fluid navigation 
in the scene. They are real-time viewers (and young) dedicated to scene 
exploration, so they do not have high visualization capabilities like 
Maya proposing ray-tracing and many other effects. And they work only on 
Windows and Linux.

I am looking for a "Mac" solution because I lead a short 3D modelling 
seminar next month in France and we will explore scenes showing large 
assemblies of proteins. And half the audience at least will bring Mac 
computers. Thus my question about VRML and LoD: I would have transformed 
the OSG scene into a VRML scene including LoD and I would have proposed 
Mac users to explore the big scenes into Chimera.

The solution I have is to decrease the triangle number, or to use the 
osg scene in the real-time viewer OSG Composer Lite that works on OSX. 
But the latter does not show high real-time rendering performance as 
LifeExplorer or OSG Director do, it has no LoD function, and I must 
anyway reduce the resolution of the scene.

Best

Damien


Yang, Zheng (YZ) a écrit :
> Hi Damien,
>
> 290 ribosomes... Even in Maya, it has to use tricks not showing full 3D, animated file texture can be mapped to particle sprites (flat planes that always face the camera), creating the illusion of crowding.  
>
> Best,
> YZ
>
>
> On Apr 15, 2010, at 10:45 AM, Thomas Goddard wrote:
>
>   
>> Hi Damien,
>>
>>  Chimera actually does not use level-of-detail (display fewer triangles when an object is small on the screen) for displaying surfaces.  It only does it for molecule ribbons, atoms and bonds.
>>
>>  For scenes with 290 ribosomes you are pushing beyond Chimera's capabilities.  But that is exactly what Maya and the Molecular Maya animation software is intended to do.  There is a student edition of Maya for $150 while the standard version is $4000.  Molecular Maya is a plug-in from Gael McGill that allows you to read in PDB models like ribosomes.  A free alternative to Maya is Blender.
>>
>>
>> 	http://www.molecularmovies.com/toolkit/index.html
>>
>>  Tom
>>
>>
>>
>>     
>>> Many thanks Greg for this reply!
>>>
>>> The "problem" I face is the following:
>>>
>>> Chimera enabled me to create an isosurface of a ribosome. Then a script
>>> in Matlab enabled me to compose a scene made of 290 ribosomes (in order
>>> to represent the crowding of the cytoplasm in E. coli) and to export in
>>> the vrml format: one ribosome is the geometry, all the others are
>>> instantiations.
>>>
>>> When I open this 3D scene in Chimera, all is displayed quite
>>> instantaneously but the navigation within the scene is slow. So, LoD
>>> would have been a solution.
>>>
>>> Do you think to something to make navigation faster? I could convert for
>>> example the scene in the Chimera format in order to benefit from the LoD
>>> capabilities of the software? In that case, I would need a part of
>>> chimera format/script to do the conversion (for instance, a script that
>>> instantiate a multiscale model in a given number of copies, at the right
>>> place and orientation).
>>>
>>> My best regards
>>>
>>> Damien
>>>
>>>
>>>
>>> Greg Couch a écrit :
>>>       
>>>> Sorry, chimera does not support VRML's LOD node. Looking at the
>>>> specification, it wouldn't be too hard to implement, just takes time.
>>>> So the question becomes, how important is it to you? And would always
>>>> displaying the first node be okay?
>>>>
>>>> Salut,
>>>>
>>>> Greg
>>>>
>>>> On 4/15/2010 4:35 AM, Damien Larivière wrote:
>>>>         
>>>>> Dear all,
>>>>>
>>>>> Chimera is able to display VRML file.
>>>>>
>>>>> Is it able to display VRML file containing level of detail? (My first
>>>>> tentative was not a success but I do not think the file I used
>>>>> respected the format convention).
>>>>>
>>>>> My best reagrds
>>>>>
>>>>> Damien
>>>>>           
>>>>
>>>>         
>>>
>>> _______________________________________________
>>> Chimera-users mailing list
>>> Chimera-users at cgl.ucsf.edu
>>> http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
>>>
>>>       
>> _______________________________________________
>> Chimera-users mailing list
>> Chimera-users at cgl.ucsf.edu
>> http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
>>
>>     
>
>
>
>