Originally posted by Post Patriot
"Three they look pixilated when zoomed into. The size is wrong also.If you look at the surroundings the clarity and size arent proportionate to the
distance of the craft."
These are fair concerns. However, since a number of people are complaining about the depth of field, focus, size, and lighting, I want those same
people to analyze this image.
(
Fig. 1 -
2100 x 3000 image /w exif data)
Notice this shot includes similar lighting conditions, there's an overcast sky, and additionally we have several aircrafts silhouetted against the
background.
For those of you who are complaining about the unrealistic "blurring on the edges." Please do the same analysis on this picture. Notice there's a
halo-effect, as one would expect, due to background lighting.
For those claiming the object is too clear, fig. 2, and therefore very close. In the above photograph, fig. 1, pretend we don't know that the
objects in the sky are F-15s. Instead look at them as though they might be the real-world equivalent of a bug from Gallaga.
No don't read further!
Look at those F-15s and visualize in your mind "Gallaga space bug!" Or if you can't envision the silhouettes as bugs instead imagine them as F-15
RC aircrafts.
I hope this impresses on you have how hard it is to measure DoF from a 2D picture when the object is unknown.
The only way to really approximate size is to use a
known object, which acts as a ruler for a single plane of depth, then using focal
length to determine the point of focus (ie/ in this case infinity). Calculating based off linear scaling and focus (circle of confusion) we can then
attempt to determine relative distance between planes (which constitutes a
VERY ROUGH ESTIMATE)!!
Notice the object, in the below picture of the triangle, is not in focus like the lamp-post. It's also less focused than the tree. Which means at
best the object is over the tree. Making it at least ~4.4' on any given side.
(
Fig. 2)
To reiterate,
For the foreground plane the light fixture can be approximated to be 1' in diameter. Using that as a guide we can say within 6 inches of
accuracy that it maps to 74 ± 1 pixels. Based off that and an average lamp-post size approaching 10' (± 2' of accuracy) we can then calculate
upwards to the possible points of intersection with the craft.
For the tree I made the assumption that a leaf would be 3" ± 1" in diameter. After finding what appeared to be a stand-alone leaf I measured out 6
± 1 pixels worth of data. Using that I was then able to use the formula for linear scaling to calculate the approximate distance between the
lamp-post and the tree. With the leaf as a guide I was able to calculate the size from the base of the trunk (assuming a 25' ± 20' tree) to
object.
I'm not saying this picture isn't a hoax. I work with game artists who could easily, with a day or twos work, do something much more impressive than
this.
The only way to truly say whether or not these photos represent something legitimate is to speak with the person who took the photographs to get a
full understanding of the circumstances surrounding the pictures and the persons technical capabilities. Following that a good investigator would
question others in the area to verify elements of the story and establish the persons trustworthiness.
Arm-chair research only tells us so much. Even though the scales lean in favor of something being physically in the sky.
Points in favor :
- The EXIF data appears accurate in the sense that the if you head on over to the weather underground and check
the conditions for the date / time, they match up with the photos.
- Other people are reporting seeing something similar in the
SC area, in and
around, this time period.
- The object correctly fits the lighting and appears to have physical dimensions that follow a reliable trajectory across all three
pictures.
Neutral aspects worth considering:
- The object isn't as in focus as people are making it out to be. For instance compare the lamp-post to the object. There are no internal
details on the craft. Which suggests several possibilities:
- it's a goodly distance behind the trees or even closer to the camera than the
lamp-post,
- no internal details (ie/ it's a solid material through-out) or,
- the object is computer generated
Which you choose will entirely mimic your bias.
- The foreground planes (lamp post / tree) are so close to the camera we innately attempt to line up the object with these reference points.
Unfortunately this makes it very easy to assume based on perceived angle, guessed height, etc., that the object is very close to the observer. However
this is entirely a trick of the mind. Though I admit it makes it very easy to lean against the possibility that the object is further away and
therefore quite large.
Cons :
- The physical dimensions are rough geometrical shapes. The only characteristics that are evident are barely visible "tiny vertical
stabilizers" on the trailing edge of the "craft." However this could just as easily be pareidolia talking.
Without fine details it's hard to calculate scale and determine what the object actually represents.
- The images could have been faked, no doubt about it. So the strength of these photos is directly tied to the trustworthiness of the photographer
and, to a lesser extent, the persons skill-set (ie/ a grandmother homemaker is less suspect than a 20-something 3D modeler)
Leaping to a conclusion in either direction is unwarranted in this case.
[edit on 21-6-2009 by Xtraeme]
In the bottom pic theres a lamp post. Is there a way we can try to size this thing? 
