Geotags, aerial synths, street synths… capturing events as they happen. If you can mark it on a map, then this is the place to talk about it.
So far it has not been too difficult to align direction in most cases. My recent effort turns out to be more complex. I added some highlights with captions to a Synth. These captions show up on the overhead view point cloud. However the captions in the overhead do not align with the realworld map positions. How do I do that? Thanks in advance for your suggestions.
There's a little bit of commentary on this here:
The highlight which I reference in the bug report was edited to a different photo after I got a reply and the highlight caption snapped into place based on that photo which focused on the foreground
Would you mind telling me if these highlights were made when you could only add the entire photo as a highlight or more recently when you could select specific parts of a photo as highlights?
The highlight captions that you're referencing are, I think, bound to the points which that photo (or presumably part of the photo) contributed to the overall point cloud. If the photo has locked onto the treeline or horizon in the distance as well as the object in the foreground that your caption is describing, the caption may accidentally bind onto the bit of the point cloud in that photo that you didn't intend.
Hi Nathanael, I had a look at your 'Memorials at Oregon's Capitol'. Very nice and interisting. I feel I do not really understand that point cloud thing properly yet. I see your synth is a rather complex 3d one with 1000+ photos. So far most of mine are simple one spot photography ones.
I have had a look at more overhead views of my synths. It seems to me that distance calculations are a lucky shot.
I feel I do not understand your explanation on moving the highlight to another photo. Also, I fail to understand your question. I can never make a photo make the whole highlight, as I am using a 16:9 aspect ratio.
Do you mean to say that it would be better not change settings of the camera not to have multi focus area?
My synths I am referring to are the St Annabaai synths on Curaçao. The ones that are published on Bing Maps have there pin on the actual camera spot.
I will create an all encompassing synth as well. Curious what that will bring.
Okay, my apologies for not being very clear.
First off, thanks for the compliment on my synth. The process really wasn't complicated. I basically circled different things and then took photos walking from one to the next, which I'm sure that anyone can be taught to do. The results aren't perfect, but then again I wasn't consistent with walking between a couple of the main objects.
As to the point clouds being accurate, it really does depend on whether the positions that you took photos from were near enough to each other. To have your highlights appear at the correct place on the map, you absolutely must have a good clean point cloud to begin with.
About the highlights - when they were first released in April of 2009, when you added a highlight, you couldn't pick a single part of the photo, but rather the highlight just centred you on the photo in much the same way that navigating to a photo and pressing the [Enter] key does. It wasn't until almost a year later in March of 2010 that we were able to pick specific portions of a photo for highlights. From what you've said now, it sounds like your highlights were created recently, though.
As to your 16 x 9 photos not fitting inside of a highlight - when you are first creating a highlight you can click and drag the corners of the highlight frame to any size over the photo. In your case, you could either zoom out of the photo or just drag the highlight frame to meet the longest edges of the photo (including the empty space above or below it), and when the highlight is clicked, that zoom level should meet the edges of the viewer.
Even though a highlight is a specific selection of a 2D image, I believe that what it is actually binding to is the part of the point cloud that the image selection within the highlight contributed to.
With the older highlights when a highlight was simply the entire image, if the image had many different things in focus, the highlight might bind to a different part of the point cloud attached to that image than what you intended.
Creating a modern highlight that points to a specific portion of a particular image seems to me like you have more control over which part of the point cloud that the highlight binds to, but this still means that the point cloud of the thing that you're selecting with the highlight needs to be clear, if it's going to be accurate in overhead view, etc.
Thanks for the explanation. Think I understand. Next challenge is to make better photo sets ;-).