The Largest AI Image I Did So Far.

Find my AI guides and support me on Patreon / sebastiankamph
Chat with me in our community discord: / discord
Stable Diffusion for Beginners Playlist • Stable Diffusion Begin...
My Weekly AI Art Challenges • Let's AI Paint - Weekl...
My Stable diffusion workflow to Perfect Images • Revealing my Workflow ...
ControlNet tutorial and install guide • NEW ControlNet for Sta...
Famous Scenes Remade by ControlNet AI • Famous Scenes Remade b...

[@reconscout2238]

Secret reptilian base in brazil confirmed

[@StoicBites]

How is this done? can you do a tutorial?

[@AbsurdShark]

The "largest image" would be something that would have no restrictions on which part of it you could zoom in to see more detail.

[@marcozolo3536]

yeh its just embedded image bullshit, not infinite resolution

[@mlegarth]

glad we have you two geniuses to tell us these things@@marcozolo3536

[@tripleheadedmonkey6613]

That would be described as "The largest and most detailed image" rather. I think you set the bar too high from the implication of size alone.

[@fureytha]

@@marcozolo3536 The Largest AI Image I Did So Far.

[@professorlich]

Seriously? That's your concern? Forget zoom. The real issue is missing dad joke. That's my outrage. 😅❤️ It sets the bad precedent.

[@armondtanz]

Dang... for a minute there.... that woman on the towel, was wondering where the zoom was headed.. 😂

[@sebastiankamph]

😅

[@armondtanz]

​@@sebastiankamphhave u seen the joker viral video using viggle & comfyui. ? Looks incredible

[@metafintek]

Amazing Sebastian, great work.

[@exposed007]

need tutorial

[@mikea8858]

Do you happen to know if there are any fractal generation models currently available? I mean I can google but .... Straight to a source is better

[@sebastiankamph]

I haven't seen one and I don't know much about fractal generations.

[@phillipberenz4284]

Even though this wasn't your most popular video, I still think it is rad. What tools did you use to generate the images? Also, is there a way to automatically zoom in for a smoother experience versus using your mouse to manually zoom?

[@mikea8858]

Great idea!! I MUST use fractals as input images after watching this :)

[@hermj]

Wow. Love it!

[@sebastiankamph]

Happy to hear that! Spent a lot of time on it

[@hotdog8008]

Hey, here's a thought. What is we do a community art project to actually create not only the largest AI image of all time, but with the most AI artists as well generating them? You start the project off with a 512x512, and then give the tile to myself and three others to each outpaint another 512 in each direction. Then we keep expanding (each person need only share the tile that they generate, and then you or I stitch them all together afterwards?).

[@sebastiankamph]

I like this idea. Hit me up on discord.

[@juanjesusligero391]

That's pretty cool! ^_^ Will you make a tutorial about this? :)

[@sebastiankamph]

That was the plan, I made a brand new workflow for this that I made myself. But people didn't seem to enjoy the video so we'll see if I do the tutorial. Might do it for Patreons since it wasn't as popular as I thought it would be.

[@Darkwing8707]

@@sebastiankamph Normally I watch your videos within a couple hours of them being uploaded, but this one wasn't in my feed. I was thinking it had been a while since you uploaded, so I went to your channel to check and found this one. Not sure why, but it wasn't in my feed. Maybe this video faced the wrath of the algorithm?

[@sebastiankamph]

@@Darkwing8707 Yeah, so what happens with the algorithm is that videos first get posted to a small number of subscribers, if the video does well, it keeps posting to more and so on. If it doesn't do well with those first subs, it kinda kills it. Which was the case with this one. Or at least that's been my experience with YT and the algo so far.

[@HestoySeghuro]

We need a tutorial sir.

[@Crazy_Truth]

Can you please help me to get excess of ltx studio

[@sebastiankamph]

If you've signed up, all you can do now is wait :)

[@Crazy_Truth]

@@sebastiankamph :(

[@antoninsvrdlik7526]

I've seen this type of animation that even had some kind of sense

[@marcthenarc868]

And of all those images, the thumbnail had to be the one about the girl in the bikini.

[@ADMNtek]

well duh. everyone knows AI art is for making Waifus.

[@aldiergreen]

So what?...

[@lcmiracle]

As it should be, that which is proper and right.

[@DaCashRap]

Looks cool, but the castle in the cat's eye appears too flat on a black background. I would expect a seamless transition through the borders between environments.

[@styrke9272]

pretty cool, this video wasn't on my feed 🤔

[@sebastiankamph]

Glad you liked it!

[@Supremish]

Im gonna get this tattooed on my back.

[@DrNothing23]

Gonna get it tattooed on MY sphincter.

[@teqwise1608]

Coolest sh*t ever!! Was AI used in making this? @sebastiankamph, could you pls reveal what AI tool was used in making this?

[@sebastiankamph]

Thanks! Yes, I used a1111 and comfy to create this. The zooming was me manually zooming with the mouse.

[@TheBagOfHolding]

Its an animation.

[@sebastiankamph]

Guess it depends how you see it. Since I am scrolling with my mouse and showing you a video, I guess then that's an animation. But it's not an AI zoom animation. I could have zoomed in anywhere on the still image. So the actual work is a still image. There is no zoom or camera generation, just me mouse zooming and talking as I scroll.

[@southcoastinventors6583]

Feel like the ending to MIB except with less marbles

[@LouisGedo]

Nice!

[@sebastiankamph]

Thanks buddy!

[@MrAwesomePoopz]

the death star is from the lord of the rings two towers

[@generichuman_]

Bruce Willis was dead the whole time

[@Collants]

Cool! Greetings and thanks!!!!!

[@sebastiankamph]

Hello! 🥰

[@HTRO-EG]

Stability Matrix

[@MrN080DY]

[@Bockiiey]

Wait... what . Thats amazing :O

[@sebastiankamph]

Thank you, glad you liked it :)

[@jayk5549]

Fractal loop worlds

[@coloryvr]

Cool!

[@sebastiankamph]

Thank you :)

[@JaredQueiroz]

keep in mind that our universe is kind of like this, but everypoint of it has the same amount of detail..... but instead of a static image, the whole thing moves Imagine a video with this type of ultra resolution... How much TB would that single file have?

[@sebastiankamph]

Oof, that would be huge!

[@SynthSoundscapes23]

Link to this image pls

[@sebastiankamph]

Too big to link.

[@SynthSoundscapes23]

@@sebastiankamph haha nice sarcasm

[@sebastiankamph]

@@SynthSoundscapes23 I wasn't being sarcastic. I'm working on a similar project where the image is 10gb, so having trouble finding places where I can upload and have people zooming anywhere they want.

[@sijello]

I’d love to learn how to make these, but I like my sleep too much

[@sebastiankamph]

They're quite time consuming if you want to be able to manually zoom with the mouse like I did.

[@silversobe]

Fractal AI! The double death star ruined it though..

[@samuelgomez4101]

Brilliant hahah

[@barriemain]

1 word - How?

[@ThinkCMYK]

Magic of photoshop 😂

[@MarioMastar]

Generally all images are rendered as an array of pixels up to a certain resolution (the screen resolution), so it's all just a visual trick where as you zoom in, the outer layer of the larger image gets removed out of memory while the smaller images gets loaded into memory and scaled accordingly. Really it just kind of loads like 3+ layers at a time so it can quickly replace the pixels as you zoom in and out using dynamic image management to always have what you're able to see ready. Then just add the right section of image during the scene redraw function. The nice thing about AI though is you don't have to pre-create these, the AI can generate them on the fly based on the colors of the section. Though this is why the example in the video shows the transitions being so wild, cause unlike when these are human made, the AI doesn't know the theme of the image, only the color of the section so it just guesses randomly at what to put there. Knowing how long it takes to render these images at scale too, it seems to be using a combination of the generation and scale up while restricting the speed of the zoom so it has time to generate the next image and scale it. Basically it's all sleight of hand, but it IS legitimate. If this was a mock up, it's more a proof of concept of what AI could do, but I think Sebastion did make this legitimately using AI as he likely wouldn't have randomly put cheese on a cupcake if this were manual. XD

[@Cola-42]

It's been a trend for iPad users in the last couple of years. Look for infinite canvas apps for iPad. Usually, people draw them, and they look way better than this example. I'd assume you can do the same with AI generated images, but you need to use those infinite canvas apps. Photoshop doesn't have that feature yet.

[@SyntheticSurrealism]

Should have posted this yesterday (April fools) ☺

[@mikesheehan5946]

The panda introduction seemed pretty lazy!

[@krectus]

I'd say about half of them are lazy or just plain bad transitions.

[@mikesheehan5946]

@krectus agreed, but I thought it was quite acceptable until, "oh, there's a random panda plonked on the canvas..."... and then, "oh, there's a weirdly miniaturised pineapple next to that woman..."... Still cool though. I couldn't do it!

[@sebastiankamph]

Thank you, very kind both of you.

[@ClaimerUncut]

​@@sebastiankamph"very kind". it's A.I nonsense which benefits those who jump onboard like yourself, and acts a as detterent to the rest of society. Very kind of you.

[@mikesheehan5946]

@sebastiankamph still cool though! But I know you could do better (I didn't know you'd made it so now I feel terrible...)

[@ZergRadio]

So much

[@sebastiankamph]

Glad you liked it :)

[@yatbond]

Very cute

[@sebastiankamph]

Thanks :)

[@GeekL30N]

xD muy bueno

[@sebastiankamph]

Glad you like it :)

[@bo-xt1ps]

This is stupid. It isn't a single image, it is just images inserted into other images. Any 8 year old could do this.

[@sebastiankamph]

Thank you! This 8 year old spent days and days putting it all together.

[@bo-xt1ps]

@@sebastiankamph lmao yikes, insane it took you that long.

[@sebastiankamph]

@@bo-xt1ps That's mostly because each image is over 12000x12000px in size. Used together they formed extremely large canvases. I can actually zoom into other parts of the images too for wild detail.

[@NotThatOlivia]

???

[@Late_AyeF]

Mouth open the whole time

[@JonRowlison]

This is kind of cool, but there are dozens if not hundreds of these on TikTok...

[@sebastiankamph]

This is not a Deforum zoom. And it's also not endless canvas painted. So please show me one other AI made.

[@JonRowlison]

@@sebastiankamph Sorry if that came across as rude... there are tons of "infinite zoom"-type images that people have made with MidJourney. I don't know what technically they have done different than what you had done in yours... but they look the same.

[@dibbidydoo4318]

​@@JonRowlisonthose infinite zooms are fake.

[@Jay-pw7pg]

Serious Question 🙋🏼‍♂️ 🔰🔰🔰🔰🔰🔰🔰 Dear AI Aficionados: Can you be so kind as to help me figure out a few things as a newcomer? 1) What are the best AI assistant options today, paid, for helping me to research a subject - let’s say AGI? So for example let’s say I ask the AI assistant to make me a list of the most relevant, respected, experts currently on the subject of AGI and mental health. And then ask for a list of the most critical journal articles, book chapters, videos, groups etc on the same subject? I’m willing to pay a monthly subscription of a $100 or so. 2) I should ask if this kind of AI assistant even exists yet? And if so, are there different kinds of AI assistants - for example academic experts, sports medicine experts, etc - or is it basically the same AI assistant tech can do all subjects and needs?

[@HTRO-EG]

Stability Matrix

[@HTRO-EG]

Stability Matrix