ALSACAT:

BELMONT (67) 09/10/2006

DATE OF OBSERVATION	09.10.2006
ReGION	Alsace
DEPARTMENT	Bas-Rhin
CLASSIFICATION	B
UPDATE DATE	03.22.2007
TYPE DOF PHENOMENON	Identified phenomenon not indicated

BELMONT (67) 09.10.2006

SUMMARY

Odd image on photograph: probably a vegetal seed.

DESCRIPTION

While emptying his camera, an amateur astronomer observes an image in the last shot that intrigues him. When the photograph was taken 10 days prior, he did not see or hear anything.

After analysis of the photograph sent, it is likely that it is a plant seed passing near the lens when the person took the photograph.

WITNESS
DATE OF THE OBSERVATION	09.10.2006
GENDER	M
REACTION	Emotion - Curiosity, Interest - Active
CONDITION
WEATHER CONDITIONS	Unknown
LOCAL DATE AND HOUR	09/10/2006
REFERENCE FRAME	Country landscape
LOCALIZATION
NATURE OF THE OBSERVATION OR DESCRIPTIVE TERMS	Phenomenon
CHARACTERISTIC OF THE OBSERVATION	Single object
APPARENT SPEED	Unknown
NOISE	None, total silence
EFFECT ON THE ENVIRONMENT	Unknown - Not specified
NUMBER PF OBSERVED PHENOMENA	Only one

From: B [-]M [-]
Sent: Monday, September 25, 2006 10:09 AM
A: Geipan
Subject: Photo

[Attachment, birthday0038.JPG]

Hello

I am sending you a photo, taken on 09/10/06 during a hike. I myself am an informed amateur astronomer, I have a fairly critical opinion on the subject and I have never observed a so-called ufo phenomenon in 10 years of practice. the three shots are taken at the champ du feu in alsace. a top meeting place for amateur astronomers. It was while emptying my camera on Saturday afternoon before going to a star party that I noticed an unusual phenomenon on the last shot. So 10 days later. We saw or heard nothing. you can go to the advanced properties of the images to get the shooting conditions. CANON G5.

B[-] M [-]

From: B[-] M[-]
Sent: Friday, September 29, 2006 9:28 AM
A: Geipan
Subject: RE: RE: Photo

Thank you for your answer.

did you notice that by tightening the thresholds there are black dots not very far from this object. maybe birds however you say an atmospheric re-entry, but of what?

Regards B[-] M[-]

-----Original message-----

From: Geipan (mailto:geipan@cnes.fr)
Sent: September 8, 2006 4:19 PM
To: Geipan
Subject:

Hello,

Thank you for your submission. It's hard to tell what you've captured in the photo. It doesn't appear to be a stray reflection and could look like an atmospheric re-entry that you wouldn't have noticed in the daytime. In that case, what a coincidence! Sincerely.

Jacques PATENET Head of GEIPAN
CENTRE NATIONAL D'ETUDES SPATIALES
18, Av Edouard Belin BP 911 31401 T

From: [-]
Sent: Monday, October 2, 2006 12:51 PM
To: []
Subject: Re: UFO Photo Champ du Feu

[Attachments:] OVNI_cdf_traite.jp ATT15147881.txt

Hello Ladies and Gentlemen,

Mr. [-] [-] sent me a copy of the photograph taken by Mrs. [-] on September 10, 2006, asking me what I think of it.

I processed two excerpts of the unusual image with photoshop software. These two excerpts were greatly enlarged without distortion, then subjected to the same contrast processing for all three color layers.

This contrast processing consists of applying a non-linear modification of the "gamma", that is to say the slope of the curve "RGB brightness of the screen" = f ("digital brightness of each pixel" ).

The curve used, defined manually, is an S very flattened at its extremities, and very steep in its center.

In this manner, the relative contrasts are very strongly reduced in the darkest pixels of the original, and very strongly increased in the brightest pixels.

I placed the edited snippets on two layers, roughly where the original images were.

Thus, you will be able to see that one of the extracts shows the unusual image against the sky, and the other a grass located in the very foreground.

Then, I "flattened" the three layers to make a single image, of which I reduced the overall size, too large for transmission by Email.

Finally, I transformed the "photoshop" format into "JPEG" format, so that it can be read with any type of computer.

What does the final image thus treated tell us?

The highly magnified processed image of the grass in the foreground is very clear.

Therefore the photograph has not been moved. One can perfectly appreciate the slope of the contrast of the edges of the rod having an inclination close to that of the unusual object. (examine with a magnifying glass the variation of the contrast of the edges of the rod, compared to the bottom, in the direction perpendicular to this rod).

This grass still has a few seeds, but most have been dispersed. This is normal on a September 10.

We also note that the unusual image is not completely blurred, because its edges have a rather steep "slope", with a rapid variation in contrast. However, the slope of the edges is much lower than that of the edges of the grasses in the very foreground.

It's about a factor of 10/15, and this, in all directions,

maybe a little more (?) in the direction of the tilted longitudinal axis.

Therefore it appears theoretically possible that the unusual object photographed was located much closer to the camera lens than the lower limit of the area of clear depth of field.

A vehicle, or a meteorite, or a space reentry, located far away, and moving during the shooting, would absolutely not give slightly blurred image edges, as it does, in the direction perpendicular to the movement.

So the object was necessarily very close to the lens, closer than the lower limit of the depth of field. It's a near certainty.

This lower limit of the area of sharpness is however very difficult to know with automatic CCD cameras, because the automatic settings of the camera can only be reconstituted, a posteriori, under conditions of lighting, colors, and central contrast of the image strictly identical to those of the shot.

The quirky image evokes quite well, for me anyway, the fuzzy image of an "upside down" vegetable seed, the seed itself being the dark oval part, and the upper part white and orange (greatly exaggerated color contrast here), being the four "sepals" attached to the seed, and which allow it to be blown away by the wind. There seems to be a kind of fuzzy white "hair" at the tip of the lower right sepal.

The longitudinal blur could be related to an angular movement of the seed during the exposure time (oscillation around the aerodynamic focus).

This is very classic.

Consequently, it seems to me that a plausible explanation for this photograph would rather be to be found in the presence of a very small seed carried away, very close to the lens, by the breeze.

To confirm or refute this hypothesis, the authors of the photo are obviously in the best position to look for seeds of this type in their environment. But it's not necessarily easy 20 days after the fact, especially if it has rained since, which is likely.

Look for very small seeds (about 1 to 3 mm overall), with four "hairy" sepals or not. The best place to look seems to me to be on the vegetation itself, if there are any seeds left that have not yet flown.

Get a magnifying glass and arm yourself with patience, because such seeds are capable of flying over great distances.

I would like to be kept informed of the results of your investigations, because "it is by forging that one becomes a blacksmith"! [French saying.]

Best regards,

10/02/06

From: Patenet Jacques
Sent: Tuesday, October 10, 2006 10:29 AM
To: [-] [-]
Cc: '
Subject: RE: Champ du Feu UFO Photo

Hello,

Thank you for your relevant and complete analysis of this photo. It seems to me, however, that given the apparent size of the phenomenon in the photo, an object as small as a seed should in my opinion be very close to the lens and therefore still much more blurred than it is or even be invisible, as is the case with dust on a lens or a photo taken through glass or a fence with the camera leaning against them. Moreover, you have actually noticed that the transverse blur is less pronounced than the longitudinal blur which tends to show a movement in the longitudinal direction but nothing indicates that this movement is strictly perpendicular to the axis of shooting, this which does not exclude a less pronounced transverse blur as you indicate for the seed. This therefore does not, in my view, exclude a very fast distant phenomenon in relation to the shutter speed which we do not know, any more than the aperture and the sensitivity used. It therefore seems to me that all the hypotheses are possible.

Regards

Jacques PATENET

Head of GEIPAN

NATIONAL CENTER FOR SPACE STUDIES
18, Av Edouard Belin BPi 2911 31401 TOULOUSE CEDEX 9

From: [-] [-]
Wednesday, October 11, 2006 12:37 PM
To: Patenet Jacques
CC:
Subject: Re: RE: Fire Fields UFO Photo

[Attachments:] IM002100.JPG ATT16126284.txt

Hello everyone,

Thank you, Jacques, for your comments.

However, I don't think photography is compatible with a distant moving object, for the following reasons, because I don't agree with this sentence you write: "we don't know, nor does the aperture and the sensitivity used. So I'm going to explain myself, because what I'm going to say is not yet taught: 1 -- Due to the variation in edge contrast over several pixels.

Let's fill in the image dimensions and brightness criteria to start with. All this can be achieved with software like "photoshop":

Upper left edge of the photo: x=0 y=0 and lower right edge x=2586 y=1938 (these are pixel numbers). The jpeg image I received has 180 pixels per inch and occupies 14.4 MB. There are 255 brightness levels in the image, per color, and 3 colors (RGB).

Black pixels have brightness R=0 G=0 B=0. Totally white pixels (saturation limit) have R=255 G=255 B=255. center of the "sepals" of the "seed" which are just at saturation (R=253 G=254 B=249).

Otherwise the sky background is in the vicinity of R = 205 G = 211 B = 223. The JPEG compression is almost nil, because a raw three-color image of this size, with 255 levels per color occupies 2586 . 1938 . 3 = 15 million bytes (one byte = 255 levels, I'm not telling you anything).

We can therefore consider that it is the raw image, seen by the CCD matrix through the lens.

Let's look at a horizontal photometric section passing approximately through the center of the "seed", along the upper edge. We adjust y = 344 = constant. We note the brightness B (blue) which is maximum in the sky:

We obtain the pairs: x = 917 / B = 218 ... 919/207, 921/202, 923/189, 925/186, 927/178 ... We therefore observe a slow transition in the brightness of the pixels, spread over 6 pixels on average. We find almost the same thing all along the upper edge and the lower edge of the "seed". For the extremities, the slope is perhaps a little weaker, but it is very difficult to measure, because we are then in the "noise of quantification" present everywhere in the sky. I don't think we can say that there is a real blur anisotropy all around this image. Indeed, with a CCD matrix, the quantum fluctuations of the photons appear. If the pixels receive an average of 10,000 photons during the exposure time, in an area of the image (the sky), then quantum physics tells us that the natural quantum fluctuations of this number, from one pixel to another, follow the Laplace-Gauss statistical distribution (bell curve) with a standard deviation equal to the square root of 10000, i.e. 100 photons. This means that

63% of pixels receive between 9900 and 10100 photons, 90% of pixels receive between 9800 and 10200 photons, and 99% of pixels receive between 9700 and 10300 photons. Consequently, if, as is the case here, the 255 levels correspond to approximately 15000 photons, it is naturally normal that, without apparent cause, the intensity of the pixels of the sky vary between 9700 and 10300 i.e. 600 photons, this which corresponds to approximately 10 brightness levels out of the 255.

Note that the statistically significant variation of the previous photometric section is therefore between x=919 and x=927, where we are far from the level 218 - 10 = 208 . So it actually spans 927 - 918 = 9 pixels. I am therefore cautious in announcing 6 pixels.

Now let's do the same horizontal photometric cut on the top edge of a distant tree just below the "seed". Same process, with y=815=constant. we get:

x = 917 / B = 175 ... 918/165, 919/132, 920/85, 921/75, 922/112 ... This time, the transition is brutal, on less than one pixel.

Excellent camera!

It is statistically significant below 175 - 10 = 165. We go from a level B = 132 to a level B = 85 in a single pixel. The edge of the tree branch is therefore resolved to the nearest pixel.

This means that the "seed" has edges at least six times wider (blurred) than the distant objects. By carefully examining the edges of all the distant objects in the image, we can write: the "seed" has edges at least six times wider (fuzzy) than ALL the distant objects.

2 -- The second reason has to do with the geometric dimensions.

The pine trees, not very far (500 m), from the forest edge below the "seed", have an average height of 110 pixels. Suppose these pines are actually 10 meters high. This would mean that at this distance, each pixel of the image would correspond to a lateral distance of the order of 10 centimeters. Consequently, the lateral "blur" of approximately 6 pixels of the seed would correspond to a distance of the order of 0.5 meters, if the photographed object had been at the same distance as these pine trees. However, the photographed object measures 131 pixels long and 40 wide, which would correspond to approximately 13 x 4 meters at the distance of the first pine trees. Obviously much more than that for a distance greater than that of the pine trees.

Under these conditions, for the same distance as the pine trees, the lateral velocity should have moved the object laterally by 0.5 meters during the exposure time.

In general, digital CCD cameras have an exposure time of the order of 0.01 seconds (more on this later). So the blur would correspond to a horizontal speed of the order of 50 meters per second at the distance of the pines (180 km/h). Or 10 times more for a distance ten times greater, of course.

It seems to me that this is not a considerable speed, and that a craft going at this speed would have attracted attention for quite a long time.

Indeed, considering the angles (we'll talk about it later), the lateral "blur" corresponds to about 6 minutes of arc. For an average exposure time of 0.01 second, the angular speed would be of the order of 10 degrees per second, or about ten seconds to cross the visual field, which is neither the angular speed of a meteorite, nor that of an airplane or a rocket. I suggest you follow such angular velocity on the sky with your finger to appreciate. This

corresponds to a finger movement of 10 centimeters per second at arm's length.

Now consider the case of a real "seed" very close to the lens. Digital cameras have impressive depth of field. Mine (which is not very efficient), gives sharp photos from 10 centimeters to infinity! I checked it. So it is perfectly possible that the photographed object was extremely close to the lens. Assume a distance of the order of 10 centimeters. Given the dimensions of the image (131 x 40 pixels), we can determine an order of magnitude of the dimensions of the seed for a distance of 10 cm: total length = 4 mm, total width (sepals) = 1 mm, diameter dark seed = 0.2 mm. So it would be very small, and very close, so it wouldn't have attracted attention.

And the 6 pixels of blur would then be a mixture of optical blur and movement speed. For a fair distribution of the blur between these two possible causes, we would have a speed of 3 pixels in 0.01 second or less than 10 mm per second (36 meters per hour). It seems consistent to me.

3 -- The third reason is photometric.

But to analyze this you need to have an idea of the angular characteristics of the image. To evaluate them, I looked at my own digital camera, since most of the angular fields of these devices are analog, and the CCD arrays have about the same sensitivity, which I'm familiar with for using them in astronomy. I measured the field of view of my own camera and found 23.3 x 18.5 degrees (that's 29.5 degrees diagonally). If the camera used had similar characteristics, this would mean that its field was of the order of 430 square degrees.

A solid angle of one steradian is 3283 square degrees. So the 2586 x 1938 pixel camera was photographing a solid angle of 0.13 steradian. Each pixel of the sky therefore corresponded to 2.6.10-8 steradian. You will understand where I am getting at in a moment.

Incidentally, if 2586 pixels = 23.3 degrees, then one pixel = 23.3 x 60 / 2586 = 0.54 minutes of arc.

Indeed, we know from experience that the blue sky illuminates us, in visible light, by a number between 0.8 and 3 million billion photons (i.e. 10 to the power of 15), per square centimeter of our skin, each second, and for each steradian of sky considered.

Digital cameras have a lens with a pupil size of the order of 0.018 square centimeters (just over one square millimeter). Therefore, taking into account the "sky surface" seen by each pixel of the image, it can be said that the photography of the sky corresponded to approximately 380000 to 1400000 photons per second of exposure time, on each pixel of the image. Simple product of three numbers. CCDs have excellent quantum efficiency nowadays, averaging 70%, which means that 100 photons give 70 electrons in each pixel. Thus, the luminous flux coming from the sky could undoubtedly create a charge of the order of 266,000 to 980,000 electrons per second of exposure on each pixel, depending on the actual luminosity of the sky that day. However, the sky pixels in the photograph are not saturated, they are on average at a level of 83% saturation. But CCD arrays are photometrically very linear, which means that the pixels in this image should theoretically be saturated with 320,000 to 1,180,000 electrons per second.

Given the time of the shot, and the reduced sunshine that we see in the photo, it is probably the lower limit of this statistical range that should be considered, but let's continue like this.

Indeed, there are physical limits to the number of electrons that can be stored by each element of the CCD array, given the surface area of the pixels. For example, a CCD matrix of 512 x 512 pixels of 25 x 25 microns each can store a maximum of 500,000 electrons per pixel. But 5 x 5 micron pixels can only store a maximum of 20,000 electrons. It is proportional to the area of the pixels.

This is also probably the order of magnitude (10 to 20,000 electrons) of pixel saturation for a 2586 x 1938 pixel CCD matrix such as that of the photograph. We quickly learn that this is so.

Simple divisions then give us an order of magnitude of the possible range of the exposure time 20000/320000 = 0.06 seconds 10000/320000 = 0.03 s and 20000/1180000 = 0.017 seconds, 10000/ 1180000 = 0.008 s.

It all makes sense. The exposure time (automatic) was of the order of a hundredth of a second.

But the significant luminosity of the grass in the foreground also leads us to believe that the camera's automatic flash probably went off.

Indeed, these grasses have no other reason to correspond to brighter v (green) pixels than the sunny grassy beaches located a little further away (V = 180 against V = 95 at 109°). In physics, we learn that extended sources of the same nature (grass) lit by the same light (low sun plus sky) must correspond to the same backscattered flux towards the lens, therefore to the same brightness of the pixels.

This law can be verified very well in the photo, with the various patches of grass.

The greater brightness V of the pixels on the foreground grasses cannot be justified by the natural illumination of the scene. So either the flash, or artificial lighting, or a reflection of the sun by a window increased the brightness V in the foreground.

So that would explain perfectly why a very small white area of the sepals of the "seed" is almost saturated. We could even calculate an order of magnitude of the distance from the "seed" if we knew the luminous flux of the flash.

All these arguments are in favor of a kind of tiny seed passing unnoticed very very close to the lens.

To check whether it is optically possible, I took pictures of my computer screen with my digital camera (1786 x 1192 pixels), getting closer and closer, until obtain a blur of 6 pixels all around the characters of this text. The result is clear: distance = 5 centimeters.

This means that with a resolution of 2586 x 1938 pixels, the distance could be around 7 centimeters for the same result, with the original photo. And then the "seed" would be no more than 2.5 to 3 mm in length!!!. There are many of this type in nature.

The fact that a CCD array camera has a considerable depth of field is not an "optical anomaly". On my own device I read that the lens focal length is 6.3mm and its aperture is F/8, which implies a lens diameter of 6.3 / 8 = 0.79mm (!).

However, the diffraction angle of such a lens in the visible is 1.22. 0.55. 10-6/0.79. 10-3 = 8.5. 10-4 radians. At the

distance of 6.3 mm, this corresponds to a blurred zone of 6.3 . 8.5. 10-4 = 5 microns. i.e. roughly the likely pixel size. This calculation reveals that these cameras are very well optimized, and that if they have remarkable performance, it is because of the extraordinary sensitivity of the CCD matrixes.

I personally had the opportunity to follow the evolution of these components, in the astronomical context, which they have revolutionized. (I even built my own CCD astronomical camera cooled to -40 degrees about fifteen years ago).

If I insist so long on the particular case of this photograph, it is to make all four of you aware, and particularly you, Jacques, of the particularities of these digital photographs.

Indeed, film photo is now almost finished, so the GEIPAN will now receive digital photos taken by CCD cameras.

These photos have very interesting particularities, thanks to the quality of the productions of the semiconductor industry. First of all, they are remarkably precise and linear both geometrically and photometrically. Measurements of distances, angles, levels of illumination are accurate to within 0.1% at the output of the CCD matrix, and JPEG compression does not degrade them too much (we generally obtain better than 0.4% finally). Moreover, the transmission of these images does not degrade them at all.

This kind of digital photography replaces a theodolite and a photometer located on the scene of the shooting. If there is, in the landscape photographed, a photometric reference (the blue sky, a street lamp, etc.) then a calibration of the shot can be obtained a posteriori. And the three-color photometry is then generally excellent.

For example, the angular field can be calibrated a posteriori by photographing a graduated ruler located one meter from the lens. This makes it possible to calibrate the angular size of the pixels (here, I know that we have for example about 0.5 minutes of arc per pixel). This allows distances to be calculated. For example, the 110 pixel high pine tree (60 arc minutes or 0.017 radians) is about 10 / 0.017 = 570 meters from the camera if its size is on the order of 10 meters.

Each photometric level (there are generally 255) corresponds to about sixty photons in the RGB color band of the matrix (indicated by the manufacturer), which allows very skilful calculations to be made. For example, on this shot, we can clearly see the random noise of photonic quantization from one pixel to another, which makes it possible to ensure that there has been no trickery.

In short, a digital photo says a lot... And we must be aware of it and practice using it for the type of analysis that we have to do now in our field of specialty. This is what photo analysis can bring.

It seems to me personally that the hypothesis of a tiny seed, located very close to the lens meets all the conditions to explain this shot.

Of course, this is not an absolute certainty. Besides, are there certainties of this kind in our specialty? We gather facts, clues, confirmations or disconfirmations, and it is on the basis of these facts that we form our opinion.

As far as digital photographs are concerned, we have to implement a whole new and relatively complex analysis process. That was my point.

Best regards to all,

10-02/2006

Here is my photo, taken at a distance of 5 cm between the lens and the screen. The edges of the characters are blurred on average by 6 pixels, as in the photo of the "seed":

Champs [sic] du Feu (67) - Sunday September 10, 2006 at 04:42 p.m. - ASTROSURF Forum

This case does not have much data, it's just a photo which was taken on September 10, 2006 at 4:42 p.m.. Witnesses have noticed this phenomenon when transferring the photo onto their computer .. When shooting, no sound was noted. The witnesses had given their picture on the website of the Astrosurf forum. Twenty five pages of comments on the forum. Someone had noticed blur of the .phenomenon, yet the rest of the picture, from the foreground to the background was crisp. This information let suggest that the phenomenon had to be very close to the lens and in this case, be very small. The phenomenon has found its explanation: a plant pistil passing in front of the lens at the time of the shooting. We can only congratulate the people for their insight.