Can’t seem to get something to synth right? Curious about what lens is best? Ask your fellow Photosynthers here.
Has anyone got any experience of trying to stitch together a synth at an event by asking multiple people / audience at an event to send their pictures in from different mobile devices? Would this be a feasible project? I am relatively new to the concept of photosynth and am not sure whether a synth has to be stitched together from shots by one camera. Any advise would be v gratefully received.
Multiple cameras should work just fine.
Although Photosynth uploads the full sized photos to the site to look at (bigger is always better to look at), it only uses a 1.5 megapixel version of each photo to match them to each other, so if your audience's devices take higher than 1 1/2 megapixel photos they should synth together as well as any other images as they're spread all around the stage pretty evenly.
IMPORTANT: The biggest concern that you should have is that the photos are in focus and not blurry... which could be difficult to achieve at a concert or similar venue. Just be sure to tell your audience that you need good clear photos, so hold that cameraphone steady!
I'd love to see this done at concerts around the world.
Roughly how many pics would we need to receive to make it a worthwhile synth do you think?
It would completely depend on your venue, I think and exactly what you want out of the synth.
If it were me, (for the purposes of my explanation I'll just assume musical scenarios, but this could expand to anything) I would want to spend some time before the audience arrives, perhaps before and after sound check, getting a couple of arcs of photos all facing the stage at different distances all the way from one wall to the other. An arc from the front, middle, and back of the venue should provide enough of a lattice for everyone else's photos to hang onto.
I would also consider a few panorama style series of photos from the stage sweeping from one side of the theatre to the other - one set zoomed all the way out, one set zoomed in all the way to get all the way to the back (separate sets if there is a balcony and a ground floor at the back), and lastly some photos that connect the shots on the stage to the photos shot from the audience's perspective.
I wouldn't say that all of the above is 100% necessary, but in my opinion it's better to have a few too many photos than you can use in one synth and trim some of your own out than to not have enough to hold everyone else's together.
If you can get the photos from the stage to connect to the photos to the floor and get someone to snap some photos from the stage while the speaker or band is playing, the audience will love being able to float from their photos to up on the stage looking back out at them which should be a great experience.
Honestly the best way to find out how many photos you need is to just head down to the location for your event and make a test synth - maybe more than once if your first try isn't hooking together like you want.
Thanks very much once again for all this information - this is exactly what i need as a complete synth novice.
You are correct in assuming a musical scenario - what i have in mind is a large scale music event with a crowd of approx 12000, 8000 of whom will be sat in seats around an auditorium facing a main stage. There is also a secondary stage in the middle of the arena, joined to the main stage by a walkway, and 4000 other audience members will be on the floor of the arena, milling around the second stage and walkway. So this is quite a complex set up with more than one main point of focus - do you think this makes it more tricky to get effective synthing action?
Also, do i need to ask the audience to focus on one particular area for their photos or is it a freestyle situation where they can take photos anywhere in the arena?
One more thing, is it better to coordinate the photography into a specific moment ie, does everyone need snap simultaneously?
This synth might give you an idea of what it's like to synth a huge event:
There are lots of things that will a crowd sourced synth come together better:
1) Good lighting
2) Everyone taking pictures of the same thing
3) Everyone taking pictures at the same time
That said, Photosynth works pretty well (in 2D mode) for veiwing photos of an event even when they don't synth together.
As long as there is good overlap (60%) between any two photos, the lighting is good, and the photo is not blurry (motion or focal) the two should synth together. So, to get everyone's photos to synth together, they all have to be linked like this.
A simultaneous snap would be extremely cool - they tried something like that at Obama's inauguration, which you may be thinking of and if you get enough photos back you would start to get something similar to the setup used for the bullet-time sequences in The Matrix.
As far as the auditorium, getting shots that wrap around the seating is really the key and if I'm imagining it correctly the elevation that the seating affords should make tracking each of the stages far simpler. My suggestion at this point is to just give each stage its own pair of arcs - at least one from high in the seating, perhaps one from the front row of the elevated seating and one from the floor.
I think that I would get at least a near and far panorama sequence from both stages and shots walking both ways along the walkway. The trick is to provide coverage that is concise but will offer Photosynth an angle similar enough to any shot that an audience member could take that it can match it.
The issue with the "freestyle" situation is that all your images will tend to be of the same thing (the stages), with few if any photos connecting the area between them. You could make this work by taking a set of images that includes everything to serve as a sort of scaffold. There is no requirement to have everyone shoot at the same moment, but, depending on the moment this may be a neat effect. Where and when is this taking place?
And there you go. Matt linked to it while I was typing. =]
Again, my method is good if you care about capturing the shape of the venue (and in my experience this is the best way to get the possibly random shots from everyone else to tie together as cohesively as possible) but with direction like Matt mentions (something along the lines of "Everyone focus your cameras on the small stage. Ready? Take a photo in 3, 2, 1, Snap!" etc.) you might get a good result even without a framework like I've been describing.
I'm just the sort that likes to have options and also wants to capture the structure of the place as it provides better odds of Photosynth placing shots taken there more correctly. The only danger is having too many shots to put into a single synth.