New features, updated old ones and general news from the Photosynth team.
Exploring the world through synths has become an even better experience now that we have our Map Explore (http://photosynth.net/geoexplore.aspx), so we made it even easier for you to Geotag your synths to show on the map.
There is a new setting in the synther that can auto-geotag your synth if you have gps or other location data in the photos you're uploading. If it finds location data it will place your synth on the averaged location on our maps.
You can turn this off by unchecking the box before the "Geotag synth using location data in photos (if present)" text in the bottom of the synther Create window.
Interested in how you get the geotag data into your photos? Programs like Microsoft Pro Photo Tools can help (http://www.microsoft.com/prophoto/downloads/tools.aspx) or some cameras like the Nikon P6000 and the iPhone have the location features built right into the cameras.
this is really some neat stuff you guys are coming up with, I can barely keep up and I think that is good thanks it's worth the investment
Darius, I'm assuming the Microsoft Photo Calibration Tool ( http://blogs.msdn.com/virtualearth3d/archive/2009/02/11/microsoft-photo-calibration-tool.aspx ) will also work.
I've been wondering ever since the update whether Photosynth only reads in the latitude and longitude exif tags or the roll, pitch, yaw, orientation, etc. tags as well. Bill Chen of MSR said that his code has preserved GPS tags since the Fall 2008 release of Photosynth, but it's a welcome update to see Photosynth start reading that and fitting our synths to the map automatically.
It seems that in the case of photos with GPS data from the camera itself, if the collection is synthy enough the lat/long data alone could orient the synth correctly (or very closely to accurately) on the map, but I'd love to see compass bearing, roll, pitch, yaw, etc. read in as well.
I'm curious about how you're handling all the embedded GPS data from the individual photos, since the final Photosynth has a single location. Is the location of the final Synth some centre derived from an average of the location in the photos, or do you map the photo locations into the 3D model and then determine the centre, or how are the EXIF-GPS locations used in aggregate?
@rakerman, I believe the synth location is an average of the locations in the photos.
Wow, it auto-tagged my pictures and also a load of others from flickr.
Really cool feature!
I took a bunch of photos with my iPhone 3GS, thinking that the embedded GPS data would help the synther figure out how the different groups of photos fit together. Some images are the front of a house, others are the back, and some are from the sides. But the tool is having trouble getting them all into one connected synth. I had hoped that the synther would use the new spatial and orientation data to help determine where key frames go, and then build the others around that.
Unfortunately, I gather that is not the case.
But I'm going to recommend it for an improvement.
I also had the same question/suggestion that JHGrove3 presents here: