It looks like you're using an Ad Blocker.

Please white-list or disable AboveTopSecret.com in your ad-blocking tool.

Thank you.

 

Some features of ATS will be disabled while you continue to use an ad-blocker.

 

The Kaikōura, New Zealand 1979 Triangle UFO 16MM film analyzed

page: 2
36
<< 1   >>

log in

join
share:

posted on Sep, 28 2019 @ 03:35 PM
link   
Things to consider when enhancing a video or photo:

The first step is the most important: your screen

Screen? Yes, your workspace. Your screen should be set to neutral. With the wrong settings, e.g. too much contrast, brightness or worse too dark, you lose valuable information as you will overlook the finer details. With an overbright screen pixels will start flowing into eachother and edges disappear. An over-contrast screen will cause the problem that gradients of coulours begin to fade as you go from tertiary, to secondary to primary colours: in other words the number of unique colours decreases.

Second step: choose your software wisely. Photoshop, Paint Pro and a dozen other programmes offer a lot of nice filters but these are not meant for forensic enhancement. In my personal opnion most of these filters do not perform the way they should. These filters intend to make an image look more artistic or beautiful but this over-enhancing of edges, gradients and colours leads to creation instead of enhancing.

Third step: Know that a perfect enhancement does not exist. Not even the high-end applications I (am allowed to) work with do a perfect job but with some knowledge of how these filters behave and what you should- and most of all should not do, you can reach a maximum 95-96 % quality enhancement. Yes only max. 96 % so when you have a complicated object with lots of fine details or faint colours at the end your project still has lost 4-5 % of its data. Point is that with every filter you use you also undo the effect(s) created by a previous filter. E.g. It is impossible to sharpen edges without losing detail as you literally manipulate pixels in order to get a better border between gradients. So in the end you will have a clear line or border but have lost colours in that specific region. That is why I prefer using different colour space and selecting various channels. In each channel you can see lots of different borders but also see the gradients you normally would not see or tend to overlook in a full colour image.



Fact is: Whatever you do with a video or image you will never reach a 100% enhancement. The only thing you can do is to prevent damage as much as possible but in order to that you have to understand how your filters work and what their influence will be on your project. Furthermore each filter has its limitations. You cannot expect from a sharpening filter that it can also maintain gradients or that a microcontrast filter considers additionally generated gaussian (grey) noise. My advice: use only one high quality filter for each single step in the process.
The best filters (imho) are the filters used in (medical) science and of course law-enforcement but latter is of course not available for the general public unless you have a lot of money or as in my case have a license to use it.

Algorithms: The problem with algorithms is that none are perfectly written. Each algorithm has its own priority. One performs better on straight lines, the other is better with gradients or colours but in the end they all fail. So we have to make a choice and for forensics I only use the basic algorithms as they introduce lesser artifacts and also create a more natural look.
E.g. when upscaling an image or video (zoom / increase resolution) I select bicubic, Bilinear or a combination of both. There are the much more fancy ones like splines and variants like Lanczos but these are for amateurs (and professionals) who do not want to spend much time on properly upscaling a video or image. I too used to use them as I did not want to spend days or weeks on a single image / or video in my spare time too but in the end I never was satisfied with the result. I stepped away from going the easy way and now the projects I work on are the result of many years. The Kaikoura project e.g. already started back 2005 but it really began when I received the original films in May / June 2015.

You all probably know that when you upscale an image the pixels are guessed and therefore created. When you zoom in on something the pixels are not there but they have to be created through interpolation. When interpolation is done few things happen. You create new borders but also new gradients and depending on how clear the original was your end result will be different. I prefer a non clear, non sharp image, preferably with fadings edges. Each time a pixel has to be guessed the original colours fade more. Most algorithms do not take this into account so when you upscale an image more than 3-4x (300%-400%) you have already lost a lot of quality colours and the more you lose the more it affects the sharpness and clearity.

To make this fading visible you can draw two lines next to eachother, e.g. one red and one green. Now you upscale the line. What happens is that the border between the green and red line begins to decolorize and turns to grey. This is the colour valley. Now imagine an image with lots of colours. After upscaling the image these colours have faded. In order to prevent that you must also during the process of upscaling correct the colourshift by adjusting the hue, saturation levels, intensity and contrast. You can do this by taking several points (pixels) or a small area within the original image, measure colour value with a colour picker (marking) and after upscaling try to adjust the larger created area of pixels to that same colour value. You will never be able to fully correct the difference but at least you have an image that has not colour faded too much.

The best results you get when upscaling an video or an image is by enhancing a series of frames / images in different colours spaces and than stack them again. This way you will have a more complete image. When you have done things right, your circle will still be a circle and a triangle will still be a triangle and Venus will still be round and not a diamond or square.

I will soon provide my team with the results of the second video, which than will be uploaded in due course. You will see the original frame size and also the enhancements performed on a non-upscaled image, so you can compare the results. You will notice that there will be almost no difference between the enhanced original and the upscaled version.

Finally: Always save an image without compression: Preferably PNG with zero compression and when you perform enhancements using several filters in a row, do not save an image in between in JPG as you will lose quality each time you save. To the eye it may not be visible but the more you enhance the less usuable data will be left over for other filters.

Having said that:

Answers to Phage's question:

The original film: UAP size: 310 x 298 microns, smallest details visible at 5 - 10 microns which is pretty impressive (@ 180 Mb/s)

UAP pixel size: 1:1
Average size object over 100 frames:
sideways: height 39 pix - length 42 pix
Top view: height 42 pix - length 45 pix (max. rotation moment)

File name: 1350.png
Used frame number: 0000-1350 - extraction date: 11-6-2015 / 19:25:48
16,7 million colours @ 24 Bits Per Pixel, 5781 unique colours
Print size: 50.8 x 28.6 cm; 20.00 x 11.25 inches @ 96 DPI
Original size: 1920 x 1080 Pixels (2.07 MPixels) (16:9)
Orginal colours: 16,7 million (24 BitsPerPixel)
Current size: 1920 x 1080 Pixels (2.07 MPixels) (16:9)
Current colours: 16,7 million (24 BitsPerPixel)
(see above image)




posted on Sep, 29 2019 @ 08:21 AM
link   
a reply to: A51Watcher

Keep them coming mate (and keep up the mighty fine research)


Did see these two relevant articles from Bruce - bit too technical for me though.


Flashing Lights

Flying Squid Boats




posted on Sep, 29 2019 @ 01:06 PM
link   

originally posted by: karl 12

Did see these two relevant articles from Bruce - bit too technical for me though.


Flashing Lights

Flying Squid Boats


Ta for those, Karl.
They can make one's brain bleed, can't they?

One of the most important images from the first link is the depiction of the object when the camera was not moving, most closely matching the actual shape described by the cameraman himself:




Spooky.



posted on Sep, 29 2019 @ 01:20 PM
link   
a reply to: A51Watcher

Again, thank you.

I disagree that the only issue with the use of zooming algorithms is a matter of color correction. Information which was not existent in the original image is invented in any method but a "pixel zoom."

I also think that the film demonstrates some effects of atmospheric distortion as well as possible distortions due the window of the aircraft. Very difficult, if not impossible to correct for.

I don't think that zoomed and carefully selected frames can be relied upon in determining the shape of the light source.

edit on 9/29/2019 by Phage because: (no reason given)



posted on Sep, 29 2019 @ 02:29 PM
link   
Some of the images do closely match what Crockett the cameraman himself described as "a sphere with spinning lateral lines" (and estimated to be the size of a house).


As a recent anecdote, co-pilot Bob Guard has kept relatively quiet about the incident and clearly regrets the massive publicity the case generated, but his comments from December last year also indicate how tense the situation was in the cockpit whilst Crockett was filming:



"One of the issues for me is we were just doing our job. We suddenly had to justify ourselves. We didn't know what the hell it was. We didn't expect to see anything. It was a bit tense as it got closer to the aircraft.

I got over it. Have I ever seen anything like that again? No I haven't. Do I believe in UFOs? No I don't. Pilots see a lot of unidentified flying things.

Would I tell anyone if I saw anything like that again? No I wouldn't. It's not worth the hassle."

www.nzherald.co.nz...


The last remark indicates how unpleasant the aftermath was.



posted on Sep, 29 2019 @ 09:30 PM
link   
a reply to: karl 12

Thanks Karl. Yes we have seen Bruce's work and he does get a bit deep in the weeds.



posted on Sep, 29 2019 @ 09:39 PM
link   
a reply to: Phage

Well Phage we will just have to agree to disagree.

Selected frames for clarity are a function of removing 'noise' from a video presentation, just like removing background scatter, gives us a clearer view of the target in question.

Errors created with 'pixel zoom' are to be avoided when possible.

Any identifiable errors with any process are to be avoided when possible.



posted on Sep, 29 2019 @ 09:58 PM
link   
a reply to: A51Watcher




Selected frames for clarity are a function of removing 'noise' from a video presentation, just like removing background scatter, gives us a clearer view of the target in question.

What is it which separates noise from signal? The observer? (Whoa, look! That frame looks like a triangle! The object is a triangle! Any frame in which is is not a triangle is noise.)


Errors created with 'pixel zoom' are to be avoided when possible.
Pixel zoom does not invent information not present in the source. Any other algorithm does. The greater the zoom, the more invention. This is not opinion.
edit on 9/29/2019 by Phage because: (no reason given)



posted on Sep, 29 2019 @ 10:27 PM
link   
a reply to: Phage

If the majority of frames show a triangle and a few do not, then which are more likely to be the true representation?



posted on Sep, 29 2019 @ 10:30 PM
link   
a reply to: A51Watcher



If the majority of frames show a triangle and a few do not, then which are more likely to be the true representation?

Can you provide the statistics for that? In the entire sequence, how many frames show a "triangle" (keeping in mind that other than a straight line there is only one shape which three arbitrary points can define). Your video shows a few. How many total frames are there in the entire sequence?

I find your analysis far from conclusive. This is an opinion.
edit on 9/29/2019 by Phage because: (no reason given)



posted on Sep, 30 2019 @ 01:04 PM
link   

originally posted by: Phage
a reply to: A51Watcher



If the majority of frames show a triangle and a few do not, then which are more likely to be the true representation?

Can you provide the statistics for that? In the entire sequence, how many frames show a "triangle" (keeping in mind that other than a straight line there is only one shape which three arbitrary points can define). Your video shows a few. How many total frames are there in the entire sequence?

I find your analysis far from conclusive. This is an opinion.


Sorry but time does not allow atm to go through thousands of frames and count one by one triangle vs non triangle. Perhaps you should get a copy and count for yourself.


Pixel zoom does not invent information not present in the source. Any other algorithm does. The greater the zoom, the more invention.


You have just contradicted yourself in the same sentence. If you disagree with the statement provided about pixel zoom, I suggest you show the exact statement provided (not a paraphrased comment) to an expert in the field of image analysis and see if they disagree (with detailed response) with any of the explanations.



posted on Oct, 5 2019 @ 01:12 PM
link   

originally posted by: ConfusedBrit

originally posted by: karl 12

Did see these two relevant articles from Bruce - bit too technical for me though.


Flashing Lights

Flying Squid Boats


Ta for those, Karl.
They can make one's brain bleed, can't they?

One of the most important images from the first link is the depiction of the object when the camera was not moving, most closely matching the actual shape described by the cameraman himself:




Spooky.



"One of the most important images from the first link is the depiction of the object when the camera was not moving, most closely matching the actual shape described by the cameraman himself:"


Actually what they see as one of the most important images is not the triangle UFO but one of the other 4 craft that were also filmed.

More on that in the 3 followup films soon to be released.



edit on 5-10-2019 by A51Watcher because: the usual



posted on Oct, 6 2019 @ 09:20 PM
link   

originally posted by: ConfusedBrit


One of the most important images from the first link is the depiction of the object when the camera was not moving, most closely matching the actual shape described by the cameraman himself:






Here is the processed version of that photo which the upcoming videos in this series will show that it is only 1 of 4 craft captured on film in this footage -






edit on 6-10-2019 by A51Watcher because: the usual



posted on Oct, 6 2019 @ 11:36 PM
link   
a reply to: A51Watcher

Candy corn.

Just in time for Halloween.



posted on Oct, 8 2019 @ 12:09 AM
link   
a reply to: Phage

It's a tasty thread



posted on Oct, 10 2019 @ 03:00 PM
link   

originally posted by: A51Watcher
a reply to: karl 12

Thanks Karl. Yes we have seen Bruce's work and he does get a bit deep in the weeds.





Ha yes he does mate.


Also a great thread from Isaac below describing how the official Kaikoura investigation was a 'whitewash'.



(5) The official files contain a considerable volume of material on the well-known Kaikoura lights documents, including comments from military personnel expressing the view that the official New Zealand investigation into that sighting was a “whitewash”.

New Zealand X-Files : Official “fib”, superiors “mystified”, + Kaikoura “whitewash”





posted on Oct, 10 2019 @ 03:34 PM
link   

originally posted by: karl 12

originally posted by: A51Watcher
a reply to: karl 12

Thanks Karl. Yes we have seen Bruce's work and he does get a bit deep in the weeds.





Ha yes he does mate.


Also a great thread from Isaac below describing how the official Kaikoura investigation was a 'whitewash'.



(5) The official files contain a considerable volume of material on the well-known Kaikoura lights documents, including comments from military personnel expressing the view that the official New Zealand investigation into that sighting was a “whitewash”.

New Zealand X-Files : Official “fib”, superiors “mystified”, + Kaikoura “whitewash”




Great link Karl!

Issac provides a very methodical and detailed job as usual.




top topics



 
36
<< 1   >>

log in

join