“Trying to do the impossible”

My father, in here, is instructing Qigong, inside a yurt, at the “Shanti House” (Hebrew: “בֵּית הַשַּׁנְטִי”) located in the south.

Qigong (Pinyin), ch’i kung (Wade-Giles), and chi gung (Yale) are romanizations of two Chinese words “qì” (氣) and “gōng” (功). “Qi” primarily means air, gas or breath but is often translated as a metaphysical concept of ‘vital energy’. “Gong” (or kung) is often translated as cultivation or work, and definitions include practice, skill, mastery, merit, achievement, service, result, or accomplishment, and is often used to mean gongfu (kung fu) in the traditional sense of achievement through great effort. The two words are combined to describe systems to cultivate and balance life energy, especially for health and wellbeing. (Wikipedia).

This is important to me so because I don’t have any pictures showing my father teaching Qigong. Luckily, I only have a 1min video that one of the participants took through WhatsApp. However, the video is in SD (Meaning, 640×352 pixels) which is very low resolution. Then, I took a frame from this video and I was managed to do the impossible to upscale it to 3840×2112. I’ll will continue work further more on it as well as trying to upscale the video.

Restoration background

As you can see in the video I am trying to figure out how to improve….. The computer screen where I am working on this project, The cellphone I am holding with my left hand where I am trying different AIs (and more…) and on the right is my other cellphone where I have opened another photo of my mom – This is from another occasion and both, this photo and the one I am trying to enhance, are about 1-2 years apart (meaning, my mom looks basically the same).

Update (6:44PM):

Finally, after maybe about 10-15 hours of hard work I was able to improve the photo remarkably, at least in the standpoint of general upscaling while in the faces I wasn’t able to reach a remarkable outcome but a good outcome which makes me happy all the same.

Zoom In

2nd image:

January 28 [6:25AM]

Zoom In

I later discovered that the panorama attachment was nor perfect in the position of my mother’s hands. Thereby, the ‘after’ is the way that it suppose to be. This is also, why when scrolling the bar (between before and after), its not only a transition between a low resolution into high resolution (more details) but also, because of the hands in the attachment process, looks like a glitch.

Update – January 29 (12:26AM):

Finally, after “2d3h50m35s” I have finished rendering the upscaled video. The source video, as mentioned, was a SD (Meaning, 640×352 pixels) and I was managed to do the impossible and finding and implementing a good software of video enhancing, to bring it to a resolution of HD (meaning, FullHD = 1920×1056). True, the quality is not perfectly sharp and it is crispy but even though, when considering this fact and possibly other imperfections that my artistic eye can catch, the improvement was significantly done and I am vey happy.

SD “Source”

FHD “Enhanced”

At this point, when I am examining the entire video, frame by frame, and I am trying to watch specifically upon the faces, I am realizing that perhaps my first enhancement tries, as shown above, might not be accurate enough. As a result, I will take out, again, screenshots from the same time keys (or, frame keys) but in the case it will be from the upscale (enhanced) video. I will enhance these frames again and trying to figure out, by comparting the old finished enhancements to the new tries. This I will have to do as face by face and at the end I will merged (or not?) the entire results into a seamlessly beautiful picture.