The Key

[Nate North]

I post a lot of videos, and so many now that I'll sometimes just stack them onto a post.

But this one is special. This video shows the very core of the Save Point Technology working. The big idea that I've spent years claiming will change everything.

I doubt anyone remembers, but starting years ago, I've been saying cryptic, seemingly meaningless things, such as "once I film a person walking down a road one time perfectly, I'll have a shot of every person walking down every road in every country of the world, forever"

Probably sounded like the rantings of a madman. Lol. Here's the system up and running as of today. Take any film clip of any type, such as my unreal engine control plates, and make that clip fit your script. No more budgets, no more walls, no more limits. If I have a clip of two people having a conversation at a table. I now have a clip of ANY two people sitting at ANY table having a conversation. Get a good shot of a motorcycle driving down a highway? Now that motorcycle can be driving down ANY highway, in ANY weather, heading towards ANY city at ANY time of day.

Once one of the transformations you see in the video above takes place, the scene is re-animated using the technique shown in the "Hybrid" video. The new scene will be animated as the original scene was, so no more weird AI movements. It's not automated yet, so it will take me a minute to make the first full proof video of everything working full scale, but it's going to be very soon.

There's always more work to do, faster, better, smoother, more automated, etc. But this is it. If you can fully understand this workflow I've assembled, you can make any movie you've ever imagined. It will still take a lot of work. It just won't take money any more.

(video below shows our in house solution for driving refabricated scenes from real or cgi base clips, I'm going to add one more brain to this stage, and it should retain coherence at photorealism under motion)

I should also note that previous animation videos and lipsync videos were using a temporary solution, and once I've streamlined and automated the current build, all animations should be lifelike, and line delivery natural and expressive.

 

[Spike]

If I have a clip of two people having a conversation at a table. I now have a clip of ANY two people sitting at ANY table having a conversation. Get a good shot of a motorcycle driving down a highway? Now that motorcycle can be driving down ANY highway, in ANY weather, heading towards ANY city at ANY time of day.

Something, here, that's getting Nate excited, a breakthrough that I'm not sure I quite yet grasp. (And I don't think I need, yet, further explanation; maybe better to keep working and let my understanding, with the work, evolve.)

I'm probably off on a wrong track, here, but I find myself wondering: Is there a generic something, in every image/scene/situation, to be extracted and replicated in some infinite number of iterations?

Of course there is: guy on motorcycyle, people at table. But maybe something more, something metaphysical. I'm thinking Plato, thinking of the parable of that poor dude chained up in that cave, so that all he can see is a blank wall, watching shadows of things made by some light behind him and thinking they are real, when, in actuality, they are only reflections of what is really real, something outside his ability to comprehend.

Yipe. Plato's writing is exquisite, as are Socratese' ethics and wisdom. But I've always found the metaphysics--the "forms" business--a little loopy. What problem was Plato solving, in making the world an illusion? But anyway.

I wonder: What would it look like if you lined up a few hundred images, like the ones in The Key above, and played them at 24 frames a second? It would, I'm sure, become a blur, the figures more indistinct. But what if they became, instead, more crisp? Maybe, revealed: the form of guy on motorcycle, of people at table.

Anyway, I'll shut up now.

(and oh, the front half of cat )

edit: and, I forgot to say--the key, the cat, great to look at: distinct, cool, engaging.

 

[Nate North]

Something, here, that's getting Nate excited, a breakthrough that I'm not sure I quite yet grasp. (And I don't think I need, yet, further explanation; maybe better to keep working and let my understanding, with the work, evolve.)

I'm probably off on a wrong track, here, but I find myself wondering: Is there a generic something, in every image/scene/situation, to be extracted and replicated in some infinite number of iterations?

Of course there is: guy on motorcycyle, people at table. But maybe something more, something metaphysical. I'm thinking Plato, thinking of the parable of that poor dude chained up in that cave, so that all he can see is a blank wall, watching shadows of things made by some light behind him and thinking they are real, when, in actuality, they are only reflections of what is really real, something outside his ability to comprehend.

Yipe. Plato's writing is exquisite, as are Socratese' ethics and wisdom. But I've always found the metaphysics--the "forms" business--a little loopy. What problem was Plato solving, in making the world an illusion? But anyway.

I wonder: What would it look like if you lined up a few hundred images, like the ones in The Key above, and played them at 24 frames a second? It would, I'm sure, become a blur, the figures more indistinct. But what if they became, instead, more crisp? Maybe, revealed: the form of guy on motorcycle, of people at table.

Anyway, I'll shut up now.

(and oh, the front half of cat )

edit: and, I forgot to say--the key, the cat, great to look at: distinct, cool, engaging.

I think if you fully understood what's possible here, it would blow your mind. In example, and I've tried to communicate this before, EVERY single FRAME of Save Point carries within it a DNA SEQUENCE. Once any DNA sequence is created, any frame, I can choose to make that the seed from which as many new genetics as I wish grow. The mainstay of my research is extremely similar to the day to day work of geneticists, cataloging and hybriding DNA sequences to achieve a desired result. I'll just do it right now to illustrate, 1 second.

Ok, this is a quick clip from a very simple little mini sequence I'm working on at the moment. I took a single genetic from a single frame, and extrapolated it out to about 500 themed video clips. As the 500 clips grew, I applied alterations to the genetics, creating the simple story. The genetic was this bioluminescent color balanced midnight Amazonian jungle river, and once in a while as the pipeline ran, I would toss in or remove another genetic, pyramids, temples, vines, stairs leading into the jungle, camera instructions, etc. I've long since created a lot of the basic stuff, A perfect sunset is a genetic for example, and can simply be added to the jungle to override it's native time, and create the exact same game location at sunset. Genetics can also be hand crafted on the fly, while the engine runs, so as of a few months ago, I basically have a steering wheel as the crops grow.

Actually, I'll just post it in it's own thread, I haven't uploaded a video here for a few days. Just uploading it now, give it a half an hour.

And to answer your thought about Plato's parable of the Cave more directly, lol, that story was one of the bigger influences on the "main" Save Point plotline. I guess my thinking here, if I'm understanding your musings correctly, is that media, television and movies, are comprised 100% of "shadows on the wall", and by a certain logic, don't need to be real, or have any reality behind them to function identically to their more analog counterparts. All the evidence in the world so far would seem to contradict my view, but right now it's a countdown to when perception shifts permanently to the way I've seen it for a decade. It may be a great actor who created that pixel by reflecting light off of his eye into a lens, but at the end of the day, a pixel is just a pixel. No magic. Just things we don't fully understand yet. (ACC)

And just so I don't accidentally make my original post unclear for some, "the key" video is not focused on the genetics, but rather on a part of the visual control pipeline. It does something totally different. It identifies the sillouette of any object on the screen in the source image (typically a film frame I wish to adapt into X) and then conveys to the brain to constrain the genetic at hand to fit those shapes. So I ran a large sequence of genetics through a single frame choke control, and this video above was the result. The one I'm about to post is just showing a single genetic being mutated to create a coherent segment, and does not use the OP technique, though at some point all future product will use it.