I The Shadow Of An Atom - False Colour Images Reveal Structure

[dt101]

I've been adding some renderings I did of the M87 black hole on a different thread and was asked to elaborate on the source of this technique which was investigating images of the shadow of an atom. You can look up the original experiment on Google, as well as observe the final image. The following images extend upon this by exploring deeper into the shadow for oddities. Hope everyone enjoys.

 

[dt101]

The second image.

 

[dt101]

The third image.

 

[dt101]

Final image.

 

[dt101]

One more...

 

[jedishrfu]

Can you elaborate on the color meaning? Is it some sort of intensity?

As an example, ocean bathymetry charts use colors to indicate depth and there is traditionally a legend that describes the correspondence of depth to color shown.

 

[dt101]

Can you elaborate on the color meaning? Is it some sort of intensity?

As an example, ocean bathymetry charts use colors to indicate depth and there is traditionally a legend that describes the correspondence of depth to color shown.

The colours don't map directly to intensity. The colours are a product of mixing identical images with slight alterations (sharpness, contrast, brightness, exposure, etc) to one of those images. The mixing produces the false colours. The images generally state, there is something to found, this is what it roughly looks like but we have no idea what that is.

 

[jedishrfu]

It would be helpful if you have a reference for this technique. I don't mean something that describes how you did it, I mean something that shows it teasing additional info from an existing image.

I know it looks like it does this. However, the more processing you do on measured data, the greater the risk of introducing false positives in the results and in this case you are processing the results of an earlier processing step that may have thrown out some details already.

Do you get where I'm going here?

Your processing may be adding in information that isn't already present in the base image and your processing isn't extracting hidden information as you may think.

This applies to your M87 photos as well.

 

[dt101]

It would be helpful if you have a reference for this technique. I don't mean something that describes how you did it, I mean something that shows it teasing additional info from an existing image.

I do, I used the sun as a reference and the technique revealed multiple things that are not apparent in the reference image. I added some of that to the thread on M87 black holes. The process didn't invent anything.

I know it looks like it does this. However, the more processing you do on measured data, the greater the risk of introducing false positives in the results and in this case you are processing the results of an earlier processing step that may have thrown out some details already.

Do you get where I'm going here?

Yes, I get where you going with that. I have quite a bit of experience with data mining, neural nets, genetic algorithms, AGI, etc. I'm happy its not happening here, we're getting a fairly decent zoom.

Your processing may be adding in information that isn't already present in the base image and your processing isn't extracting hidden information as you may think.

This applies to your M87 photos as well.

Could be, but I wrote custom code to throw a loss pass filter, with a bit of a bias, at the image and it returned the same structures. I have actually posted some of this at my site. The link is in the other thread.

 

[jedishrfu]

I suspected you had some ML and data mining experience from your comments.

With respect to the zoom, I have to disagree. The analogies I've heard is that to image the M87 black hole photo of 40 uarcsecs is like trying to image a golf ball on the moon.

The problem here is that while your work is very interesting and the images produced are really quite artistic, its not been peer reviewed in any science journal and so none of us here can really comment on what you believe you've discovered other than to say that its a personal theory.

PF has strict guidelines here to not discuss personal theories.

 

[dt101]

I get it, no worries. Interesting though, isn't it?

 

[jedishrfu]

Truly, they are. In some respects they are like those fascinating Julia set fractals.