Found Deceased IN - IMAGE discussion re Abigail Williams, 13, & Liberty German, 14, Delphi

[Billy Mabou]

I wonder if the small dots are chroma clipping in the channel.

I re-created the animated GIF of the 3 BG images, but instead of display in normal RGB, I did a color transform to YCbCr color space. The reason being, most JPG images are compressed and stored in YCbCr color space. Y is the b/w intensity, Cb and Cr are 'chroma' or color axes that define the color.

I contrast stretched the two chroma (Cb and Cr) images by a factor of 2.5 for better display, but left the Y image along.

ALSO: I have no idea what the small points in the Cb and Cr images are due to . I haven't seen that before that I can recall.

 

[Brighty]

In the moving images....i see his left arm behind his back holding something long ,,,maybe a bat or long gun...sticks out at top of his left side . Almost like he is trying to hide something.

Yes or at least he is reaching behind him for something, possibly weapon in back pocket...I am sure he is reaching back or carrying something behind his body to hide it.

 

[BigCityAccountant]

I re-created the animated GIF of the 3 BG images, but instead of display in normal RGB, I did a color transform to YCbCr color space. The reason being, most JPG images are compressed and stored in YCbCr color space. Y is the b/w intensity, Cb and Cr are 'chroma' or color axes that define the color.

I contrast stretched the two chroma (Cb and Cr) images by a factor of 2.5 for better display, but left the Y image along.

ALSO: I have no idea what the small points in the Cb and Cr images are due to . I haven't seen that before that I can recall.

.
When the picture in the middle flashes to the 3rd frame, it appears something is hanging from his left leg. (it's the picture where the left knee looks blurry. Perhaps, an ID badge??

By the way~~Great Job Michael, WebSight, NIN and everyone else contributing!!! :goldcrown::goldcrown::goldcrown:

 

[Billy Mabou]

I'm sorry I just don't understand what you mean by the "small points?"

I'm not 100% sure, but I'm thinking Michael is referring to the little black dots or splotches in the cb and cr slices of the picture.

It looks to me like clipping, where a pixel value exceeds the range of the colour space. These areas appear in the parts of the image where the light is very bright. For instance, the right wrist, which some people see as a watchface. To me, based on my experience, what I see there is flesh that's been overexposed. Ideally, there'd be flesh colour visible there, but the exposure exceeded what the camera could encode, so the area "clips" away leaving a white hole (where red, green and blue are all exceeding and clipping). That hole is what people are seeing as a watch face IMO.

So, I'm guessing that's what we're seeing in the colour channels. Black holes where colour values should be, but are lost.

I'm not a colour scientist, but that's my understanding. I hope Micheal will correct me if I'm misunderstanding what he was talking about.

 

[Billy Mabou]

Also, notice how the blue channel clips in bright areas of blue (the jeans) and red channel in areas of red (skin).

Ok, this is the point where my wife usually gives me 'that look' and I need to stop talking.

 

[mpnola]

Nice. That seems more fittin w other two imo. What's the green patch on the hood of hoodie( for whoever else sees hoodie now) doesn't match background leafage to me personally.

Sent from my SM-J727V using Tapatalk

It has that same green mark behind his left leg at the edge of the bridge. I don't think it is something that really exists

Sent from my SCH-I435 using Tapatalk

 

[mpnola]

I agree with your observations. LE has most likely removed some frames between these images, for whatever reason....frames that could possibly show him adjusting his hoodie.

Imo

Sent from my Commodore 64

Yall. He would have to be doing some seriously fast ninja moves to change his hoodie and get his hands back in the same place in 2 frames. I think it is much more reasonable to assume the hoodie is still on but the difference in light just changed the appearance in the frame. I bet in the next 2 frames it would reappear.

Sent from my SCH-I435 using Tapatalk

 

[Lilowlet]

It has that same green mark behind his left leg at the edge of the bridge. I don't think it is something that really exists

Sent from my SCH-I435 using Tapatalk

Seriously? I dunno what's real anymore😢 really starting to dislike pixles... distortion.. Colors etc. Thanks I'll check it out!

Sent from my SM-J727V using Tapatalk

 

[Elley Mae]

I'm sorry I just don't understand what you mean by the "small points?"

If i am not mistaken the points appear between suspects legs in cb x 2.5, in cr x 2.5 they appear on suspects right hand and face. Jmo

 

[michael.gartley]

I'm sorry I just don't understand what you mean by the "small points?"

Yeah, that was a confusing statement by me. What I was referring to was the small dark and bright pixels in the two Cb and Cr images such as a few I circled in red below.

The Cb and Cr images are typically low spatial resolution on purpose, due to how the JPG compress is designed to work. However, the circled regions I was wondering about seem to correspond to very bright or saturated (max Y value of 255) pixels.

Gonna ask my office next door neighbor (a color scientist) who probably knows why ;-)

 

[michael.gartley]

The Cb and Cr images are typically low spatial resolution on purpose, due to how the JPG compress is designed to work. However, the circled regions I was wondering about seem to correspond to very bright or saturated (max Y value of 255) pixels.

Gonna ask my office next door neighbor (a color scientist) who probably knows why ;-)

I spoke with a colleague that specializes specifically in image transforms through different color spaces and the answer is it's normal. The regions I circled in red are simply a different color (bright white) than the pixels around them and are indicative of actual fine color detail differences in the image. In the case on his cheek, the difference between white shiny reflection and flesh color of his face.

 

[Billy Mabou]

I spoke with a colleague that specializes specifically in image transforms through different color spaces and the answer is it's normal. The regions I circled in red are simply a different color (bright white) than the pixels around them and are indicative of actual fine color detail differences in the image. In the case on his cheek, the difference between white shiny reflection and flesh color of his face.

Still looks and sounds (based on what the colour science guy said) like clipping to me. Not sure what "actual fine color detail differences" means. I see clipping a LOT. It's ugly.

 

[No it's not]

I remember seeing this photo when the case first got going. It then disappeared (I can see why now according to your explanation). One of the WS members found it and posted it again but it got lost in the threads. I did a cursory look for it at one time but it never came up in my searches. Glad to have it back. Still another "look" for BG and possibilities for sleuthing his person.

Nin, I see how your sequencing "works" but funny how my mind wants to say 3-2-1 because it looks more like a complete step sequence to me: the end of lifting of the right leg (the kicking out right before putting it down, which moves him forward), putting that right foot down (has moved forward the one step) as the left lifts in back, and then the left knee coming up getting ready to take the next step (in which the left leg would now kick out to move forward).

Also, with that tree clearly visible in the back and then not in the other two frames, it seems he has stepped away from it. Again, it may just be my mind sees the tree in one but not the other two so it feels like he's moved on from it, especially since we know he does move forward and away.

Also I wonder if perception of BG's position could be slightly skewed because of the possiblility that Libby herself was moving or at least the camera was moving while she was filming. For example, she could have been moving off the platform and making her way down off the bridge or was at the end of the bridge but trying to move further away.

What would be really great is if we had been privy to the wider shots before they were cropped. That tree is probably in all three backgrounds and would probably reveal the sequence easier.

Another point that this additional photo brings to mind: we have three stills, three, which almost constitute a complete step sequence and LE would not release a video clip that at least showed this? Why on earth not? Nothing has been revealed here but one step.

Okay, I have edited this at least 8 times and must stop. These are just my thoughts only.

WS, when you look at the three frames, compare or measure the distance between the platform RRT and the leg of the POI/BG. Frame 1 is closest. For frame 2 and 3 look at the planks and notice where the left foot of the POI/BG is in frame 3. Frame 2 shows the left foot up and in the process of getting there (where left foot is located in frame 2):

Frames 1-2-3 ( NF and the 2 original frames by ISP), platform RRT as reference point:

Frames 1-2 (NF and one original frame by ISP) showing sequence:

Frames 2-3 (2 original frames by ISP):

All IMO

-Nin

 

[J Lee]

Yall. He would have to be doing some seriously fast ninja moves to change his hoodie and get his hands back in the same place in 2 frames. I think it is much more reasonable to assume the hoodie is still on but the difference in light just changed the appearance in the frame. I bet in the next 2 frames it would reappear.

Sent from my SCH-I435 using Tapatalk

Valid points...although we don't know if he stopped walking mid-stride or if the wind blew his hood off.
Without viewing more of the video we will probably never know what happened on that bridge, at that time.
It is all speculation because right now it is all we can do with these poor images.

Sent from my Commodore 64

 

[No it's not]

Yeah, that was a confusing statement by me. What I was referring to was the small dark and bright pixels in the two Cb and Cr images such as a few I circled in red below.

The Cb and Cr images are typically low spatial resolution on purpose, due to how the JPG compress is designed to work. However, the circled regions I was wondering about seem to correspond to very bright or saturated (max Y value of 255) pixels.

Gonna ask my office next door neighbor (a color scientist) who probably knows why ;-)

Hey, you got some nice neighbors! ;--)

-Nin

 

[Peace777]

Could the bag /pouch be a tactical belt pouch? Here's some examples at this site they come in all colors shapes and sizes
https://www.aliexpress.com/nylon-belt-pouch_reviews.html

 

[LaneGirl]

In the frame by frame it makes it look like he is making a turn to me. He seems to be leaning too far to the left (if you were behind him) to be walking straight.

 

[WebSight]

WS, when you look at the three frames, compare or measure the distance between the platform RRT and the leg of the POI/BG. Frame 1 is closest. For frame 2 and 3 look at the planks and notice where the left foot of the POI/BG is in frame 3. Frame 2 shows the left foot up and in the process of getting there (where left foot is located in frame 2):

Frames 1-2-3 ( NF and the 2 original frames by ISP), platform RRT as reference point:

Frames 1-2 (NF and one original frame by ISP) showing sequence:

Frames 2-3 (2 original frames by ISP):

All IMO

-Nin

Thank you Nin for the explanation and for reposting the visuals. I totally "see" it now. Thx. :fence:

 

[J Lee]

In the frame by frame it makes it look like he is making a turn to me. He seems to be leaning too far to the left (if you were behind him) to be walking straight.

I agree. He is either
1.) turning.
2.) walking carefully, possibly avoiding hole/obstacle.
Or...
3.) about to sprint.

Sent from my Commodore 64

 

[Lilowlet]

I agree. He is either
1.) turning.
2.) walking carefully, possibly avoiding hole/obstacle.
Or...
3.) about to sprint.

Sent from my Commodore 64

4) intoxicated? (No I don't think it's RL just adding a possible)

Sent from my SM-J727V using Tapatalk