ok what is it

[ChrLzs]

I'm curious why the streaks in yours are so dramatic, when I also have taken 30 second exposures, and they look like this (you can see the exposure length on the exif in "..." button at right under photo name):

As stated, it's because I was using a slightly more magnified view (75mm, if you are familar with 35mm focal lengths - 50 is approximately 1:1), plus it is a crop - so you are seeing a zoomed view. If you look at the first pic you posted, the trails are in fact of similar angular size if you zoom it up.. In the second shot that image was significantly out of focus (sorry, but them's the facts - take a good look at the original!) So the stars and trails have bloomed out into wide elliptical blobs - reducing that image far enough to make the stars look sharp has caused the effect to be much less visible.

Note that the farther away from the celestial pole you are, the longer the trails will be. If you are shooting (northern hemisphere) towards the Pole Star or (southern hemisphere) Octans, the traisl will get progressively shorter and more 'bent', down to nothing at the dead center. But this effect is still quite noticable for anything other than a telescopic (highly magnified or cropped) view - plus if it was angled at Octans I'd recognise the star pattern - and I don't. As I said, I don't think these are stars. I'd love to see some feedback from the person who took the image, although it may now be a bit embarrassing to fess up to what has actually happened, if they even know..

In all cases and at any magnification, I think you would have to agree that none of these example images look *anything* like the OP image. Something isn't right with it - but without the full-res version and further info, we may never know.

[Kahn]

Hi;

A youtube clip of a geosynchronous satellite flare...

[media=]

[/media]

Which could explain the stars movement and the object remaining stationary.

Another good possibility. I was thinking aircraft strobes on a turning aircraft. That's why I did not even venture to guess at what the object was. Too may possibilities.

[ChrLzs]

Guys, I think several folks missed the point that I made previously about the exposure.

Well, I didn't miss it, but it wasn't correct, I'm afraid!

Now I'm making the assumption that this was taken with a digital camera.

pretty safe assumption nowadays, but without further info...

If you note the image artifacts to the lower right of each "star", I believe these were caused by the camera not being secured on a tripod.

No, they are nothing like movement trails - if they were, they would be at least roughly the same width as the source and relatively smooth, not 'checkered'. Instead they are in precise 8x8 pixel blocks, they show exactly the effects that jpeg block compression artefacts do - you need to re-read Jesse's post - these *are* jpeg block compression artefacts, not real scene detail, not clues as to pasting, not movement trails...

Similar to the effect you see when someone is doing light painting except the camera was moving instead of the source.

As per the above - those trails are smooth, and the correct width for the source... Not like what is shown in the OP image at all.

However, when zoomed by the same amount (500%), we do not see the same effect with the object photographed.

Once you zoom that far (and to do so, you should never use interpolative zooming as it creates even more false detail), you need to be able to recognise compression artefacts. Yes, the object does not show much in the way of compression artefacts, so it suggests that it might have been 'created' at a different time or by a different process to the blocky areas, but there could be other reasons for that.

So the stars "moved", but the object did not

There are several examples of star movement on this page and *none* of them show effects like the OP image. (Although I could resave my images at very high compression rates to demonstrate what happens..) So the conclusion that they 'moved' isn't really warranted.

Does *anyone* here who has done any serious astrophotography believe they are stars from a 30 second exposure? I don't.

Edited November 30, 2013 by ChrLzs

[JesseCuster]

Guys, I think several folks missed the point that I made previously about the exposure.

Now I'm making the assumption that this was taken with a digital camera. If you note the image artifacts to the lower right of each "star", I believe these were caused by the camera not being secured on a tripod. Similar to the effect you see when someone is doing light painting except the camera was moving instead of the source.

Actually, if you examine the stars in the photo, you'll see that the artifacts appear around the star in some cases, to the upper right in others, etc. There's no real consistency. And the artifacts make up neat little 8x8 blocks as you would expect if it was the result of JPEG compression.

The reason you don't see the same obvious effect with the UFO, is that it is large enough and has enough color gradients that the same effect is still there, it's just not as obvious. Like I said above, small objects with harsh contrast against a plain background do not come out well in heavily compressed JPEG images.

Here's a sample below of a couple of the stars, blown up 1000% and with the brightness increased to make it more obvious. In these cases, the artefacts are not to the lower right of the star. If it was trails caused by motion blur, it would be consistent across the photo. As it is however, note the neat 8x8 blockiness of the image with the artefacts from each star fitting neatly into one of the 8x8 boxes. Exactly what you'd expect in a JPEG compressed image.

I don't know anything about astronomy photography so really can't comment on the validity of it being the real deal as a photo of actual stars, but the artefacts we're discussing are definitely caused by JPEG compression.

To the OP: Any chance of the original uncropped, unresized file as it came from the camera, is there?

[Kahn]

No, they are nothing like movement trails - if they were, they would be at least roughly the same width as the source and relatively smooth, not 'checkered'. Instead they are in precise 8x8 pixel blocks, they show exactly the effects that jpeg block compression artefacts do - you need to re-read Jesse's post - these *are* jpeg block compression artefacts, not real scene detail, not clues as to pasting, not movement trails...

I would have to differ with your assessment of it being block compression, though you do have a point with the 8x8 (though all I see is 7x7). Having programmed filters for Adobe Photoshop (back in the stone age), the JPEG compression algorithm would not produce bright spots on the opposite side of the block.

Once you zoom that far (and to do so, you should never use interpolative zooming as it creates even more false detail), you need to be able to recognise compression artefacts. Yes, the object does not show much in the way of compression artefacts, so it suggests that it might have been 'created' at a different time or by a different process to the blocky areas, but there could be other reasons for that.

That was my point, though I'm open to your other explanations.

[Kahn]

Actually, if you examine the stars in the photo, you'll see that the artifacts appear around the star in some cases, to the upper right in others, etc. There's no real consistency. And the artifacts make up neat little 8x8 blocks as you would expect if it was the result of JPEG compression.

Interesting! What software are you using, cause ALL my artifacts were in the lower right corner in both Adobe Photoshop CS2 and Paint Shop Pro X4.

[rashore]

Erm... could someone give a laymans description for those of us that don't speak photo tec please?

What's going on in the OP photo?

[toast]

Guys, I think several folks missed the point that I made previously about the exposure.

Now I'm making the assumption that this was taken with a digital camera. If you note the image artifacts to the lower right of each "star", I believe these were caused by the camera not being secured on a tripod. Similar to the effect you see when someone is doing light painting except the camera was moving instead of the source.

However, when zoomed by the same amount (500%), we do not see the same effect with the object photographed. So the stars "moved", but the object did not. As I mentioned before, this could only happen in one of two ways, either the object was illuminated for a shorter period of time than the exposure, or the object was overlaid on the original photo of a star field. I suppose you could do it the other way round, but that would involve more work.

Kahn, re : yr star moving theory. Yr theory might not be right. If you take the 2 "stars" under the object (in max enlargement) as a reference it can be seen that the brightes pixel is in the upper left corner of the 8x8 pixel field. But this matrix did not show up in the other "stars" in the same fashion. So it cannot be judged that the "stars" moved and the object not.

I by myself think that the "stars" are not real ones. I often do camera falling star hunting, earning hundred of pics per trial with lots of stars on it. The stars are always more colored (blue,orange,yellow) than the dots on the schimke picture. Anyway, as already mentioned a few times here, without RAW file and EXIF data and place/time of shooting, no deeper investigation is possible because all trials will end up in a zoserloop.

Edited November 30, 2013 by toast

[toast]

Erm... could someone give a laymans description for those of us that don't speak photo tec please?

What's going on in the OP photo?

This might give a suitable air jacket for the first trials :

http://www.digital-photography-tips.net/digital-photography-terminology.html

[ChrLzs]

Interesting! What software are you using, cause ALL my artifacts were in the lower right corner..

????

Kahn, you need to look more carefully. Check the OP image again and look at ALL the 'stars', eg try the one at coordinates 867, 169. That is one of those that Jesse showed, and on my screen at the same mag, it is exactly as shown..

[JesseCuster]

Interesting! What software are you using, cause ALL my artifacts were in the lower right corner in both Adobe Photoshop CS2 and Paint Shop Pro X4.

Photoshop CS5.

If you zoom in to about 800% (zoom, don't resize) and increase the brightness to highlight the artefacts, and browse around the image, you should notice that the speckled artefacts all sit within neat little 8x8 blocks and the stars appear in various random places in the different 8x8 blocks. I'm really not sure why you would be seeing anything different.

If you're actually resizing the image, set the option to 'nearest neighbour' in the dialog box as the other options all interpolate the resized image and will blur out and partly disguise the very artefacts we're discussing.

[sinewave]

Interesting! What software are you using, cause ALL my artifacts were in the lower right corner in both Adobe Photoshop CS2 and Paint Shop Pro X4.

Mine too.

[ChrLzs]

Yours too, sinewave? Please show us - use the two 'stars' at the upper right of the image so we are talking about the same ones.

(If you are not sure how, you can copy a screenshot with PrtSc, then paste *that* (Edit, Paste or Ctrl-V) as a new image into your chosen image editor and save it somewhere, then attach that file to a post here.)

Additionally/alternatively, tell us step by step what you are doing to examine the image.

[JesseCuster]

Mine too.

Hmm. Along with the pair I showed above, here's another couple I extracted from that image. Again, I've increased brightness to deliberately exaggerate the effect.

I don't know why you and Kahn would be seeing those artefacts otherwise than what they appear to me (and apparently ChrLz as well)

And just to further illustrate the point, here's the photographer's watermark blown up and brightened to see the exact same blockiness and associated artefacting around the edges of high contrast areas in JPEGs. (click to enlarge)

Edited December 1, 2013 by JesseCuster

[ChrLzs]

As a reference and to show why those stars 'look funny' to those of us who do this sort of photography, here's an enlarged portion of the image I posted earlier, showing real stars imaged over 30 seconds, with their characteristic trails - click to enlarge.

Note that this image is compressed, but only slightly so there is little sign of any compression artefacts (- I've circled an area where you can just see a hint of them..) Also note (if you have good eyesight!) that the colours of the stars are subtly different, and note the smooth trail effect showing the distance the star travelled (well, it was actually the camera and the earth upon which it is mounted, rotating, hence the curve). For info only, the arrow shows a dark edge effect - that is from in-camera sharpening - it deliberately enhances the edges (I should really turn that off..).

But wait - the plot now thickens!! Remember how I just said the colours of the stars in my image are subtle - you can detect minor hues with slight yellow, orange, red, blue and even green tinges... I just had a quick look at the 'stars' in the OP photo, and almost every one is a perfectly neutral grey - in other words, the RGB levels are identical, eg 209,209,209 or 156,156,156. As for every star in my (real) image, those numbers always wander, eg 129,134,138, etc. But not in the OP image - they are perfect grey... There are a couple of fainter smudges that are exceptions, but there are way too many that are absolutely dead on neutral grey to be coincidence. Now this could happen if the image was deliberately taken in greyscale (black and white), but the centre 'thing' is most definitely in colour, so that cannot be the case. That, more than anything suggests to me that they were deliberately added, probably with the logo which is also done in exactly neutral grey and white.

So, imnsho, the 'stars' are DEFINITELY some type of manipulation or (very, very slim possibility) a failing sensor issue. It's time the owner of the image fessed up to what has been done here. Dingo, any further news? Has she been pointed over here? Did she actually present this image as a real image of something amongst stars, or has it just been plucked off her web presence? Has anyone seen the original or the other similar image?

Edited December 1, 2013 by ChrLzs

[Kahn]

Photoshop CS5.

If you zoom in to about 800% (zoom, don't resize) and increase the brightness to highlight the artefacts, and browse around the image, you should notice that the speckled artefacts all sit within neat little 8x8 blocks and the stars appear in various random places in the different 8x8 blocks. I'm really not sure why you would be seeing anything different.

If you're actually resizing the image, set the option to 'nearest neighbour' in the dialog box as the other options all interpolate the resized image and will blur out and partly disguise the very artefacts we're discussing.

Followed your directions to the letter. Looks like I happened to pick the two stars that were the worst examples (they do have the halos offset to the bottom right). I do however see that the red star above the object has the artifact center top, and the others do have some variation. So kudos to ChrLes, but on his new point of no variation I do see one red and two blue stars using JC's technique. Do you confirm this ChrLes?

[freetoroam]

As a reference and to show why those stars 'look funny' to those of us who do this sort of photography, here's an enlarged portion of the image I posted earlier, showing real stars imaged over 30 seconds, with their characteristic trails - click to enlarge.

Note that this image is compressed, but only slightly so there is little sign of any compression artefacts (- I've circled an area where you can just see a hint of them..) Also note (if you have good eyesight!) that the colours of the stars are subtly different, and note the smooth trail effect showing the distance the star travelled (well, it was actually the camera and the earth upon which it is mounted, rotating, hence the curve). For info only, the arrow shows a dark edge effect - that is from in-camera sharpening - it deliberately enhances the edges (I should really turn that off..).

But wait - the plot now thickens!! Remember how I just said the colours of the stars in my image are subtle - you can detect minor hues with slight yellow, orange, red, blue and even green tinges... I just had a quick look at the 'stars' in the OP photo, and almost every one is a perfectly neutral grey - in other words, the RGB levels are identical, eg 209,209,209 or 156,156,156. As for every star in my (real) image, those numbers always wander, eg 129,134,138, etc. But not in the OP image - they are perfect grey... There are a couple of fainter smudges that are exceptions, but there are way too many that are absolutely dead on neutral grey to be coincidence. Now this could happen if the image was deliberately taken in greyscale (black and white), but the centre 'thing' is most definitely in colour, so that cannot be the case. That, more than anything suggests to me that they were deliberately added, probably with the logo which is also done in exactly neutral grey and white.

So, imnsho, the 'stars' are DEFINITELY some type of manipulation or (very, very slim possibility) a failing sensor issue. It's time the owner of the image fessed up to what has been done here. Dingo, any further news? Has she been pointed over here? Did she actually present this image as a real image of something amongst stars, or has it just been plucked off her web presence? Has anyone seen the original or the other similar image?

I am certainly no photographer and all this number and rgb, prtsc, blar blar blar stuff is well over my head, but even without the technolgy of camera/photo stuff, it was obvious the stars were exactly the same. But it is interesting how you can work it out, cos this is solid proof, which holds much more weight than just an eye on its own.

_only's third photograph of the real stars showed that stars are not the same, so in the op it seemed strange that even to the naked eye, they were identicle.

[_Only]

But wait - the plot now thickens!! Remember how I just said the colours of the stars in my image are subtle - you can detect minor hues with slight yellow, orange, red, blue and even green tinges... I just had a quick look at the 'stars' in the OP photo, and almost every one is a perfectly neutral grey - in other words, the RGB levels are identical, eg 209,209,209 or 156,156,156. As for every star in my (real) image, those numbers always wander, eg 129,134,138, etc. But not in the OP image - they are perfect grey... There are a couple of fainter smudges that are exceptions, but there are way too many that are absolutely dead on neutral grey to be coincidence. Now this could happen if the image was deliberately taken in greyscale (black and white), but the centre 'thing' is most definitely in colour, so that cannot be the case. That, more than anything suggests to me that they were deliberately added, probably with the logo which is also done in exactly neutral grey and white.

So, imnsho, the 'stars' are DEFINITELY some type of manipulation or (very, very slim possibility) a failing sensor issue. It's time the owner of the image fessed up to what has been done here. Dingo, any further news? Has she been pointed over here? Did she actually present this image as a real image of something amongst stars, or has it just been plucked off her web presence? Has anyone seen the original or the other similar image?

Couldn't the photo provider have selected the UFO area in Photoshop, inverse selected the rest of the photo, and desaturated it to make all but the UFO grayscale? I do this sometimes in photos for artistic effect, but it could also be used to make something in a photo stand out, or make other things less prevalent, for whatever reason the editor wants. I mean, it's already clear that the photo was edited with a signature made to look nice. This is also possibly the type of person who would make every editing attempt to make this photo look as good as they knew how to.

Then again, someone who could do something like that would also know how to use the healing tool or clone stamp to get rid of compression (edit: compression artifacts, I mean), unless they got lazy for whatever reason.

Edited December 1, 2013 by _Only

[ChrLzs]

Followed your directions to the letter. Looks like I happened to pick the two stars that were the worst examples (they do have the halos offset to the bottom right). I do however see that the red star above the object has the artifact center top, and the others do have some variation. So kudos to ChrLes, but on his new point of no variation I do see one red and two blue stars using JC's technique. Do you confirm this ChrLes?

Yep, I did mention a couple of them are slightly coloured. But for so many of the bright ones to be exactly neutral grey? I've never, ever seen even ONE star that was a perfect RGB triplet, unless it was blown to white (255,255,255), and none of these are... Those few slightly coloured ones could be another jpg compression effect, or they could be part of the 'base' image onto which the others were added.

_Only, yes that selective desaturation technique could have happened, but again, it is deliberate manipulation of an image that is being offered for analysis (or at least we think it is..). Plus you'd have to ask why on earth someone would do that on an image that is already very lacking in colour. That technique is really only something done for artistic effect, and it fails dismally if the coloured area is not actually very colourful.

I just can't conceive of any way that someone genuinely wanting help to identify something, would deliberately come up with these manipulations no matter which of them was done. It just doesn't make sense - even if it's a deliberate hoax it still has me puzzled, although I guess I have to admit - I've seen worse...

[Kahn]

LOL, love that pic!

[ChrLzs]

And I've now done a quick bit of digging.. It's on the "Higgins Storm Chasing" (HSC) facebook page. The OP image, which was forwarded to them, is supposedly the uncropped image but obviously reduced (and very badly compressed). We don't know what camera, but it has been stated (by HSC) it was taken on an 18-200 lens. That suggests it is probably an APS-C-type DSLR. (Sorry for the tech speak - that means it's a digital single lens reflex camera, with a medium size sensor - it should give results similar to mine, so my example image should be comparable..). We also don't know if a RAW (better quality than jpeg) file exists but it has been requested.

Nettie apparently lives just out of Toogoolawah, according to a facebooker who claims to be her friend. That would mean it's in a 'country', rather than 'remote' region - it's not a huge distance from the city of Brisbane, but far enough away to avoid most light pollution.. Toogoolawah is a smallish country town (just big enough to have a golf course!). Esk is to the South so the camera was supposedly pointing that way.

Interestingly, Nettie has posted a few comments, but none are particularly helpful (although she is happy to joke about it being Santa...) She has completely ignored the many requests for the original image to be made available or the EXIF data to be posted. She claims to have already sent the HSC people all the details, but clearly that isn't the case.. Her last post (9 hours ago approx, and it's about 5pm here in Queensland now) reads:

Just to clear things up.. it is not photo shopped.. there are no street lights where I live it isnt a plane.. the dropzone was not out flying around as they where rung and confirmed they weren't. .. all details have been passed on to HSC.. of the photo.. have a lovely sunday everyone..

And that's about it.... I suspect the mystery might be resolvable if we ever get access to the original image and the 'other' one.

PS - the 'dropzone' comments relate to the fact that there is a local skydiving club and some folks suggested they might have done a night jump..

Edited December 1, 2013 by ChrLzs

[Kahn]

Good digging!

[WhispersInTheAttic]

It certainly a good image, but yeah I see what you by how people are editing there's photo's and trying to make it out like it's real. Makes people all excited that it's the real thing but in fact they just edit their photo's. It bums me out whenever someone does this. They should just keep the original photo and not try to add more to it.

[toast]

.

[zoser]

.

That's exactly what I thought.