The odd quality of the images caught my eyes so I found the original video. Googling for it didn't quite work but Google Gemini delivered:-
Here is the official government source for the video, which is the most "original" version available (not a YouTube re-upload):
- Official Congress.gov Video Player: https://www.congress.gov/committees/video/house-oversight-and-government-reform/hsgo00/aLJ7hyTz-Nk (Look for the entry titled "The Deposition of Ghislaine Maxwell on the Epstein Probe" dated 02/09/2026)
That video which I believe is the most original version available to the public has a quality that looks like this:-
and NOT like this:-
or this:-
(different colored eyes, uneven glasses)
which I strongly believe are AI "sharpened" or "enhanced", which is really generated, created, hallucinated, dreamed up, assumed).. in other words fakes.
Who made them, how, and why and with what understanding? Probably someone without the proper know how of these things who thought they could sharpen the face properly using AI tools.
Yes her nose appear quite a bit wider but we have to take into consideration things like age (noses grow) , puffyness... and primarily:-
- Flat Lighting (Washout):
The room utilizes diffuse, overhead fluorescent lighting. This eliminates shadows on the sides of the nose (the "bridge"). Without shadow definition, the brain cannot perceive the nose's height/projection, causing it to register as a single flat, broad surface blending into the cheeks. - Low Resolution & Compression:
The image is a low-bitrate video frame. Compression artifacts blur high-contrast edges. The subtle lines defining the nostril wings and tip are smeared, artificially merging them with the surrounding skin. - Lack of Focal Compression:
While 2 meters prevents "fisheye" distortion, a wide-angle lens at this distance lacks the "slimming" compression of a telephoto (portrait) lens. Telephoto lenses compress features, making faces appear narrower; wide lenses at a distance capture the face flatly. - Subject State:
The subject (in tan) appears unstyled/without makeup. Makeup creates artificial depth; its absence, combined with the factors above, maximizes the "flattening" effect.
Here's a video about lense distortion. However in this case it's not a wide lense close to the face
To me her voice and way she speaks sounds VERY much like her as in this old interview of her and her ears look very different.
Here's an example for nose difference where the reason is camera lenses etc + age
EDIT: a "better" AI upscale to 1) demonstrate AI "upscaling" 2) give e better sense of what she might look like instead of the very poor quality low res+compressed original footage and understanding of the nose
