Glitched background on Superlumina site


#1

I don’t think I’ve seen anyone working on the glitched background of the Superlumina main page. There are two images there. The first is Sky.jpg, and on the glitched page with the message, it’s a glitch.jpg. I wondered if comparing the two images could provide a clue for decoding the glitch in the other two, unsolved images.

I haven’t found much yet, but I did an image diff between the two images, and the results were a primarily bright red image, with a number of white dots all over it. Could this be some sort of filter or overlay applied to the image? Could the dots mean something? Or, could the dots be made from the stars somehow, and an artifact from comparing the two?

Maybe there’s more going on here, though, and perhaps comparisons of the two might be useful for someone else.

The images are located at these URLs:

(http://s3.amazonaws.com/cdn.superlumina-6c.com/wp-content/uploads/2017/05/04165945/Sky.jpg)

(http://s3.amazonaws.com/cdn.superlumina-6c.com/wp-content/uploads/2017/05/04200655/glitch.jpg)

And here is the result when I tried a diff process between the two (the white dots are VERY small):


#2

And, just to clarify a bit, the red will be the differences between the two. The white dots are the same between the two. That’s why I asked about an artifact from the stars – maybe the stars each share a bright point that overlaps between the two, but otherwise everything is different…?

Still, just wondering if there’s something to work with between those two background images, but I’m not entirely sure of strategies to compare the two.


#4

Did a couple of comparisons to check exactly what sort of difference filter you had used. Just to clarify, I have listed my results, which match yours, below:

  • White pixels show where neither the color value, nor the alpha has changed; pixels in both images are identical.
  • Red pixels indicate where either the color value or the alpha value has changed; pixels are different.

JPEG images have no alpha channel, so in this case we can conclude that the white pixels show where color values are identical in both images.

Note: The colors red and white in this case are only used to identify a change or not. Any two colors could have been used for this, they themselves have no meaning.

There are several ways to get differences and I had checked these using Gimp, which comes with it’s own layer mode and I have some additional plugins to achieve a variety of difference results. I just wanted to clarify the exact filters/tools used to come to the above result. I am sure other image editing software has similar tools/plugins.

Looking at the glitch.jpg image, it appears a color filter has been used to accomplish the red look. In Gimp I can use ‘Colors > Colorize… > Hue=360(full), Saturation=100(full), Lightness=-50(half)’ to get really close. Additionally we can clearly see about 20px horizontal lines/bars all over, which somewhat resembles what we see on the wakingtitan website (http://cdn.wakingtitan.com/loop-short.gif). I think a layer with these lines/bars was added on top prior to applying the color filter. To me it appears that no glitch effect has been applied to accomplish this result.

Other things/tools tried:

  • StegSolve, nothing clearly hidden.
  • JPEGsnoop, nothing out of the ordinary, besides indicating to be an edited image (duhhh).
    See for yourself for detailed analysis results using this tool.
  • Hex edit, nothing stands out.
  • Spectogram in Audacity (U-Law/A-Law), both Frequencies and Pitch (EAC), nothing stands out.
    Note: Make sure to lower volume if you wish to play the raw data as audio.
  • Photosounder, interesting sound, but nothing stands out.
  • Outguess, without key, doesn’t seem to be anything.

I personally do not see anything special in the background image used. Hope my info is useful in any way, but I doubt there is any hidden clue. I do however not want to discourage others from trying. Good luck


#5

Looks like musical notes.