Smithsonian X 3D launch

[Taylor]

The Smithsonian Museum has launched an on-line 3D viewer they refer to as X 3D, with examples of objects from their collections.  The on-line viewer has some interesting features ("advanced tools"), described here.

 

The X 3D web viewer is "powered by Autodesk" and it appears to remain proprietary for the time being.  Their 3D effort is being managed by the Digitization Program Office, which states,

"The Smithsonian digitization challenge and opportunity can be measured by the total number of collection items: at 137 million objects, artworks and specimens, capturing the entire collection at a rate of 1 item per minute would take over 260 years of 24/7 effort. At the present moment, the Smithsonian has prioritized the digitization of about 10% of its collections for digitization. To rise to this challenge, the Digitization Program Office is promoting rapid capture photography workflows for two-dimensional collections, and exploring innovations to speed up the capture of our three-dimensional collections, preferably in 3D."

 

Such a high-profile effort has the likelihood to set standards for other institutions, and I'm curious to what extent the Digitization Program Office is coordinating with other organizations to conform with open standards for this type of dissemination effort.  For example, they state, "For many of the 3D models, raw data can be downloaded to support further inquiry and 3D printing."  I wonder if the raw data will be provided with complete metadata?  I would hope and expect so, since they envision the 3D future as extending into field data collection efforts.  I haven't tried downloading any of the data sets, but some examples are available here.

 

I've noticed (e.g., here and Carla's response) that the University of Leuven has developed their own viewer and file format for RTI, CyArk has developed their own 3D viewer for point clouds, and other institutions like Oxford University are engaged in massive efforts to digitize portions, or the entirety of, their collections for on-line dissemination.  With public-private partnerships such as the Smithsonian's X 3D project, it would seem there's a downside potential to have fragmented, proprietary standards and restrictive copyrights for certain subsets of data collections (e.g., CyArk), while there's an upside potential to do groundbreaking research by providing access to collections across multiple institutions and even private collections, if open standards can be used to better harmonize these data sets and make them available. 

 

Laser scanning appears to be the core technology for the X 3D project, while this video suggests they're also using CT scanning and photogrammetry.  Scanning might be fine for some objects, but may provide incomplete information for others (such as manuscripts, paintings, etchings, and items with significant fine details).  If this is the future of museums, how will the public and scholars know the extent to which the objects rendered on their computer screens accurately represent the objects in the collection?  What are the guidelines and standards for digitizing objects (e.g., density of point clouds, and accuracy of mesh reconstructions and RGB data)?  Will the Smithsonian integrate other technologies, such as RTI, multispectral imaging, photometric stereo, white-light interferometry, confocal microscopy, etc.?  The Smithsonian recently held a conference built around the rollout of X 3D, and I'd be interested if anyone who attended the conference has any thoughts about these issues.