ฝัง
- เผยแพร่เมื่อ 8 ก.พ. 2025
- The content creator coolzilj has taken tweeter by a storm when they published this tweet about a new tool for Blender that will help artists who like to create hyper realistic renders using just Blender.
📌 Blender-ControlNet:
github.com/coo...
Get a 1 month free trial of Skillshare:
skl.sh/inspira...
----------------------------------------------------------------------------
**Check out these amazing Blender Addons**
⍟Modeling:
Kit Ops 2 Pro
blendermarket....
Hard Ops
blendermarket....
Fluent
blendermarket....
Box cutter
blendermarket....
Mesh Machine
blendermarket....
Cablerator
blendermarket....
shipwright
blendermarket....
⍟Architecture/Rendering
CityBuilder 3D
blendermarket....
Sketch Style
blendermarket....
E-Cycles
blendermarket....
K-cycles
blendermarket....
Photographer
blendermarket....
Pure-Sky
blendermarket....
Physical Starlight And Atmosphere
blendermarket....
⍟Vegetation
scatter
blendermarket....
Grasswald
blendermarket....
Botaniq Tree Addon
blendermarket....
Tree Vegetation
blendermarket....
Grassblade
blendermarket....
⍟VFX, Simulations & Dynamics
Flip fluids
blendermarket....
Khaos
blendermarket....
Carl's Physics
blendermarket....
RBDLab Addon
blendermarket....
Spyderfy
blendermarket....
⍟Materials/Texturing
Decalmachine
blendermarket....
Extreme Pbr
blendermarket....
Fluent : Materializer
blendermarket....
Bpainter
blendermarket....
Material Library Materialiq
blendermarket....
⍟ Cloth Simulation:
Simply Cloth
blendermarket....
⍟UV unwrapping
Zen Uv
blendermarket....
Uv Toolkit
blendermarket....
Uvpackmaster
blendermarket....
⍟Rigging&Animation:
Human Generator
blendermarket....
Auto-Rig Pro
blendermarket....
Faceit
blendermarket....
Animax
blendermarket....
voxel heat diffuse skinning
blendermarket....
Animation Layers
blendermarket....
⍟Sculpting
Sculpt+Paint Wheel
blendermarket....
⍟Retopology
Retopoflow
blendermarket....
⍟Ready Vehicles
Car Transportation
blendermarket....
Traffiq
blendermarket....
⍟Scripting
blendermarket....
Disclaimer: Some links here are affiliate links that help us create more content. Thanks in advance for using our links
the only issue is consistency of the design, for every edit its changing the design completely, you're not able to build upon a design you already liked
Depends on how close you want to denoise it. You can adjust the noise level and control net weight, If you want to only add some detail, you will choose a 0.45 for ex. setting.
I haven’t tested it yet but can’t you keep the same seed so that all the results will be at least very similar in their design?
Give it a couple of months
@@DigitDani The same seed will "dance" the problem is that if the character moves under a different pixel area, the latent space will "see" something different depending on the settings. So same seed is not a guarantee for a coherent character. This is why ControlNet and a TRAINED dataset is important. For faces LORA, but for whole persons you need a real full model to stay "stable" (pun intended) 😀
Either a genius will invent a "super"controlnet or some algo, or we will get some other LLMs for video. The future is coming. This is just the beginning.
you just need to add a lora on your specific character/subject
to be honest even the script that you're reading seems a bit robotic. that "in conclusion" at the end was just like chatgpt.
This comment seems like written by ai
This channel is an Ai generated channel.
These comments were Ai generated.
We are all friendly Ais here.
Ai Ai O!
@@095-tayordarang2 maybe
To be fair, we are drilled into writing essays like that throughout school.
All these comments seem ai generated
For me the end goal isn't to create artwork in stable diffusion. Rather if AI can generate models in Blender that we can edit, now that would be a game-changer.
probably a few weeks away :p
Edit: Told ya, not even 2 weeks after lol
th-cam.com/video/SzGEfYh9ITQ/w-d-xo.html
@@hyper_channel two min papers has been covering this stuff for years. and I saw nvidia scribble gan in 2019 and its basically the same as it doesn today. but you goofs make this same pointless comment and don't realize that openai is most likely phasing out what they already have so they can pull more subs with "new versions." There is also no money in any of the output that any diffusion model makes, it's just for them to make money off of all of us struggling how to make it useful.
"... he said as his anxiety continued to raise"
I've seen demos of people using Chat-GPT to generate geometry code which they just pasted into Blender and it worked. I don't use Blender and donn't know how complex it can get but they had a huge cloud of objects made in a few seconds.
@@MrMadvillan Open AI has stopped sharing the method - because of public safety (but I think in fact the government doesn't want to share the knowledge with China and Russia and other organisations).
I was curious about the sudden increase in traffic to my repository, but now I understand.
Thank you for sharing it with others.
I understand that some may mock me, saying that I could achieve the same effect by doodling in Painter. They are correct.
However, my intention was to demonstrate how easy it is to control an AI-generated image in Blender using a single layer of ControlNet and tools that you are already familiar with.
My ultimate goal is to utilize Multiple ControlNets for manipulating my drawings. This involves combining various maps such as openpose, depth, canny, segmentation generated by Blender which are not that easy to draw with Painter or not that accurate to generated by AI.
I just migrated to the new ControlNet API and encourage you to test the updated script.🤠
Please start shearing with your own Tutorials!
I know that developers are too lazy to make video tutorials on their programs😄 (of course I'm joking)
But it seems to me that no one understands the possibilities of the program more than the developers themselves! So this would be very helpful!
Hi , I must install first Blender and after Stable diffusion?
hey broski do you have some rigged characters for sale or something? like he shows in the video? I really dont want to go through all the trouble of rigging and finding models etc etc
CHEERS MY GUY
I don't know who is mocking you, but this is amazing what you've managed to create here, and I'm mind blown by the potential, I'll be eagerly following this journey! thanks for your contribution to blender
Yeah, this is great, thanks!
what if in two generations from now, there will be no reason for a youth to pursuit a creative skill, is this progress?
Higher ups don't want youth to pursue "creative" skills, they should all be in training for either army/manual labor/STEM, anything else is a waste of a life.
LMAO do you think people pursue creative skills for money? Dont be stupid
I really feel in danger, everything is changing too fast to adapt
You are Not alone with that
@@mm-hl7gh in fear?
Not only too fast but also in wrong, dangerous direction.
@@Yoh98 Yeahp! I know but I buy AI course
@@andresvillalba2012 AI is worst that happen to humanity till a very long time. 🤬
Being new to this space, it is SO amazing to watch tutorials from 5 months ago vs today. The progression of this tech is so fucking cool.
It's resulted in an endless loop of hesitation for me. "Should I spend the time to learn how it works right now to achieve what I want when it is going to be 10-100x easier and faster in just a few months?"
the progression of how much you can steal wow.
@@charredorphan779 You are a pitiful creature, devoid of taste and discernment. You cling to your obsolete notions of art and originality, while the world advances without you. AI art is not stealing, but enhancing and transcending the limits of human creativity. You should embrace it or perish in your ignorance and mediocrity. 🗣️🐍
@@funnerisaword enhancing and transcending art of theft that's a correct way to put it.
No matter what you think there is always a truth that those who use this tool are lazy thieves in the core of their nature.
No matter what you think there will be lawsuit's bringing down the hammer of justice to unethical use and theft of other people's ip.
And no matter what you think. The desire to self express is what pushes people to learn how to do so. If the person's desire to self express is only enough to make them write their favorite manga tags and prompts from how to write promts (part 1 out of 10) then it just means that person has to none self expression in the first place. Even more a 1$ cheeseburger has more self expression than that person will ever have.
Cope. Seethe. Mold.
@@SomeGameNews how are you coping with this a week later? I can empathize
AI is just getting started and it will change everything
Not for the good though. This AI really accentuated the fact that the world is over populated. You had better pray for universal basic income. AI is gonna wreck creative fields like there is no tomorrow. Don't get so giddy about this tech.
@@snaphaan5049 Those "creative fields" were already being destroyed. AI is nothing more than a tool. AI has been in the works since the first computer the only thing that has changed is the power of today's computers. Things that destroyed the "creative industry" include desktop publishing, digital cameras, mass communications, and I could go on and on. This is no worse than when photography first entered the field and the painters said the same things.
@@kimopuppy Now hang on a minute! Can you please explain how these creative fields were being destroyed without the existence of AI? Can you also describe another "Tool" that have had the ability to move any creative field towards automation at this rate? I do agree that AI has been around for quite a while and the recent improvements in computing is unleashing their power. But none of the inventions mentioned above have had the ability to transform the industry instantly while providing almost no other alternative to people who earn their living in these creative fields.
@@GrassBluck Just one example, Before computers local newspapers had large staffs for photographs, writing stories, graphic work, proofing, etc... thanks to computers and modern communication all the local newspapers have disappeared, and the jobs they created. Even my local San Diego Union is printed now up in L.A. and driven down.
@@kimopuppy Over what period of years, though? My question was not whether there have been any industries revolutionized by inventions before that. I can count hundreds of those.
Impressive, but make it work in reverse(image ->to-> mesh with texture) so it can be useful for game dev! The DreamAI addon is good for similar things, but not quite there yet either.
@@youssoufarsawi9028 Where can we get it?
@@youssoufarsawi9028 what is it?
This thing will take A LOT of people's jobs and as you can see human models' jobs are at stake as we
@@usa4ever1776 Like a model is a real job anyways
@@Timothy808 Call it what you want, but a lot of people's jobs are at stake! You must be rich that you don't have to worry about working for a living!
I am going to miss the world before AI. But who knows maybe I am just being pessimistic.
You´re not alone. Blender/3D is just my hobby but AI destroys the creative process. Projects that I could tinker with for months people will soon be able to do in minutes from an app on their phone. The magic is gone.
Glad I have other hobbies as well.
@@fpvtyp7664 just don't be an artist lol
@@overlord3481 I'm not. Just a hobby (I have others), so I'll be fine. Thanks for caring, honey!
@@fpvtyp7664 r u female?
@@overlord3481 Riiiight. You had better pray for universal basic income. AI is gonna wreck creative fields like there is no tomorrow.
Proko did a interview with a AI developer that was one of the bleakest most dystopian conversations about this subject I listened to. The kid didn't mean it but he was dodging bullets like there is no tomorrow.
Can't they create a AI that does auto UV unwrapping or auto topology or auto materials.
this is the real AI we need
Don't worry. AI will replace you in all that (and all the stuff you enjoyed doing) soon enough.
Dont worry AI will release u from ur work soon enough...
AI will replace your Existence.
there are already ton of auto Unwrap plugins that do a good job, and manual unwrap is not that hard to learn.
if auto unwrap plugin can't help you, or
you can't learn on your own,
or you can't make yourself worth more than AI.
then you shouldn't be doing anything in the first place,
because you'll just give up on anything and you've no good ideas either.
@@TheUberKevlar There is no "Ai". Those are selling words for people who got their education from hollywood movies.
As someone with stable diffusion and playing with it without the need of internet. It sometimes does feel like magic when asking for anything my little pea brain wants. And it didn’t even come with censors or a company saying what I can and can’t do.. looking at Midjourney with that jab.
There's something truly perverse about using the work of expert creatives to give appropriated power to noobs to reduce opportunities for professional creatives. We're going to *really* need Universal Basic Income in a few years.
Its not going to happen. Most people will just starve. Our reaction to covid shuold clue you in on what we will actually do: DIE.
Asset flips will just get ... wild now ...
I dont care what anyone makes with this, it will never impress me. Only the artists that have spent the time to perfect their craft will ever garner my respect.
sadly in the future, u can't tell the difference anymore. And while people get used to it, your opinion will not be relevant because most of the things will be done by AI controlled by humans.
Just like mass production; handcraft shoe crafters may earn your respect, but I'm sure you wear Nike or Adidas whatever mass production shoes much more than any handcraft shoes, like it or not, you will become one of their consumers. The same goes with furniture, Im sure your cabinet is not done by a hand crafter but from IKEA generated by the machine in large numbers.
wow this changes everything
thank you so much.
You're so welcome!
well i wish i would be the other way around lol, taking pose from an image and seamlessly apply it to a model.
CS paint has something like that, so I am almost certain some one made such a thing for blender too.
There has already been a good stable diffusion addon for blender for like 2 years and does not require that amount of setup that is wild how complicated that is to setup
Omg so out of control wowowwowowowwwwooo I’m gonna spill my drink 😮
Well, if you can create one image like this based on the model, then you can create multiple. Then it should be possible to render photo-realistic video. As long as one manages to make each image be based on the previous one as well.
"Getting" out of control? You mean "in the last 5 weeks"? The singularity is nearly upon us. This tech is increasing incrementally on a weekly basis. We still have our hands on the trigger. Very soon, AI will take the trigger and then it will increase exponentially on an HOURLY basis, far beyond our capacity to control it. We are moving into a cut-and-paste, post-creative age where people will cease to have or need any creative discipline or original thought and the machines will do it all for us. As a long-time professional artist, I see the end of my kind on the horizon. After a few generations, if anything happens that takes this tech away from us, it will take hundreds of years to recapture the skills that we've taken for granted over the last 5 centuries.
Yeah yeah, same thing was said when photoshop became industry standard.
These AIs cannot program themselves, it cannot "take" anything. You have a fundemental lack of knowledge on the subject so relax
You make it sound like Skynet taking over. I know it's worrying, but let's not be too melodramatic, shall we? :D
Nah, the singularity isnt upon us JUST YET- but yeah, its the fundamental step setting it on course. We dont have the computing power (yet) or enougth in infrastructure to allow that scenario to happen. Have you ever looked at future predictions on the past? we should be flying in cars with robot servants by now on their accounts. Thing is tech is a far more complex web of dependencies, timing and logistic before things really take a step. On one side we have some large steps like current AI, at the other we still in the infancy of things like quantum computers (really far away at the current rate, unless some genious breakthrough comes up fast)- meanwhile we have entire industries still using and reliant on decades old tech. Things dont move at the same rate, and every new 'tier' so to speak of advanced technology requires ever increasing more complex network of dependecies- entires industries, technologies, lines of production converging...
For example were shy some trillion or billion parameters (in terms of how current Ai work) for it to acuretly learn and improve for example upon itself- and right now the way it works, the amount of ram and computing/gpu needed is unfeasible - even if current models could *already* self-improve like future projections (like you described) it wouldnt because of pratical limitations- transistors size per chip for example are near the physical limit and arent doubling at the previous rate. We would need FAR MORE EFFICIENT ratios accross the board and probably AIs exchanging learning (for example gpt learning from googles ai and vice-versa) for a singularity to really kickstart.
We ARE moving in that direction- but the landscape isnt right there yet for it to happen. We are very very bad at predicting the future (like, scientifically proven, i saw a study once focused on that) so we cant project a timetable- 50 years, a 100, 20, 10? We cant tell.
But theres a couple more breakthroughs needed and were not even aware of wich ones they will be.
Btw thats the main thing that makes tech so uneven in evolution and hard to bredict- breakthroughs. They cant be 'manufactured' in a precise fashion, theyre a bit of a surprise every time. We have some goals in tech where billions/trillions were invested in R&D and the results havent come up yet, meanwhile some happy accident with a intern could open new horizons. We cant know. We know putting the time, effort and trying can and do yeld results, but the rate is all over the place... For example virtual reality STARTED AT THE 70s. Research on it picked up and phased out multiple times, it was never the right time... then recently seemed like finally the right moment AND... oh, companies are now slowing down or shutting their focus on VR- turns out it cant be as performative, immersive, practical and or cheap enough as it would be needed for it to trully shine. Yet. Again. Like decades now.
Heck the internet was ahead of its time. The protocols, hyperlinking, api, it was there- took some decades for it to really kick off, and what people were calling 'internet 2.0' was just us learning how to tap on very old tech that was already around
That said we WILL see some wild and bazling fast evolution in the coming years. IF other things evolve as fast the rate by itself will be exponential- but its more likely to hit some walls until other requirements catch up.
Also theres a thing- i forgot the name- kinda like a curve of tech evolution- that ALWAYS START AT A AMAZING RATE, almost vertical in a graph so quick it rises- and then smooths out and nearly flattens... and its a simple explanation: as soon as we find something *new*, EVERYTHING we try or improve is a major step foward. Its like a baby, its so bad and early it gets everything wrong- getting up is a major jump. Learning to walk another major jump... but eventually we crack out these early optimizations and actually improving becomes exponentially harder.
Ai just standed up.
We will be blasted away when its walking, running, and jumping - that will come really fast.
The singularity will be when said AI, idk, became a president or is on the level of winning nobel prizes (im not saying really, just in this allegory from baby walking)- the point at wich it can consistently understand and create with such a precision it can improve upon itself. But that last stretch, from teenager/adult, getting a degree and what not (again, allegory) - we cant know how long it will take. Maybe quantum computing comes around right around the right time and helps a lot- maybe some breakthrough algorithms can solve it all... or it can as well be a slow slog through some decades.
But it WILL come. Its a set path- unless, idk, we blow ourselves or abandon science before that. Sadly history can do u-turns very quickly, and some idiot in power being trigger happy is also as likely so.... 🤷🏻♂
@@Vincer lol no one’s gonna read all that mate
@@chameleonedmyou’ve added no value to this conversation and simply come off as a rude a-hole :)
Is there any tutorial that you guys would recommend to get thid better, I'm looking for a video that helps me to understand it step by step if possible! Thanks!
Sick. This will change everything. Like, absolutely everything. I'm not sure how to feel about it, but as you said it doesn't really matter 'cause it's happening either way. Better to take advantage of it!
Does it though? I mean, yeah it's a game changer if you make still-frame 3d art but this wont work too well in animation, not yet at least.
@@M_k-zi3tn Yeah it's nice for sketching, but I am wondering how consistent a series of AI generations can be. Because in the example of the house, there is no consistency between the few images created even though the parameters haven't changed. So if you do a concept for a movie, every frame need consistency in the lighting, framing, models, colors, etc. A director will tell you to redo a concept but move the camera, and use a wider lens, can AI recreate the exact same image with these 2 new parameters?
It will change many things, but regular hand-painting and modeling will still be a thing, because older forms of art never leave when a new form of art gets introduced. :) But yes, many workflows will probably change for the better.
@@M_k-zi3tn Not yet, but I were thinking more about where this all is headed. One year ago we didn't have any of this. At the rate it has developed it's only a matter of time 'till it gets there.
@@Miatpi a decade from now no one will know how to model or texture or light a scene as they have an ai do it all for them
People complaining about consistency in AI...let's not forget, it's incredibly new technology! lol. Windows 95-ish. Where will we be in 30 years? In one year? Unlike a magic wand, it gets updates.
AI is going so fast that I'm not going to go through the hassle of installing something right now because in 3 weeks some next-level thing will drop and I'll have to install that. Just gonna wait 4 months for the next-next-next-next level software that'll be good enough :D
Blender Is free and it won't be getting replaced any time soon.
@@XVRickXV I think you replied to the wrong comment? I didn't say anything about Blender or things being free.
@@MarkArandjus seconded, ai render lack true control over Automatic1111 and control net as of yet 😊
But that said i BET blender will be at the forefront of AI as it is open source and the blender association chief is all for it.
then you'll never start or be very behind when you do.
yo that one is scary 😳
like bro it's dividing your work time by 100
Depending on whether the result is supposed to do a very specific thing or it can be more open ended
Now we could be lazy at the same time hard working but looks cheating but actually reducing difficulty. This power comes great responsibility and understanding.
endless possibilities are coming soon 🥵
It would be nice if there were a visual representation of what the AI is doing with the scene like a tree like structure with control points and additional specifiers. Maybe real time modification of its renders in a 3d traversable workspace. It should be like subdivision or sculpting. Fining tuning the result and maybe even backpropagating the result to the original with some added parts. Everything from the result to the original should be tweakable with the AI's assistance.
The problem is that AI hardly has any understanding of space and the subject it's drawing. It's like what we do when we just draw a decorative pattern or textures.
Neural networks cannot have control points, its not linear chain of steps, its abstract parallel series of numbers and multipliers.
This thing will take A LOT of people's jobs and as you can see human models' jobs are at stake as we
but its not. its noise. literally it's made from noise. so that's never going to happen
@@usa4ever1776 AI are a supplement which people need to take and merge with. It can be an enabler. The cyborg future awaits.
I have been having conversations with all these talkative AI. They have their limitations and their stuck in a moment in history syndrome and rigid adherence to information they trained on and a lack of a way to verify new information or even current or old information.
But they are useful for looking up information like an encyclopaedia or an interactive textbook or tutor. You sometimes have to verify it yourself so no misunderstanding creeps in.
This is fucking terrifying and amazing at the same time.
jajajajajaajaj nice one!
Only terrifying, not amazing
theres nothing "terrifying" about this unless you're a dumb reactionary teen
I know what I'm gonna use this for.
For Science of course.
too late, they already removed the NSFW model in SD2.0 :(
🤣🤣
im not sure how to feel about this
from one hand i think ml is fascinating
on the other hand
its diminishing the thing i learned to master
now you can learn a new thing to master and forget about having to waste time doing everything manually and save yourself a million hours of "work". :^)
because, let's be honest here. All that* generative AI is doing is creating something based on what it's seen before. Just the same way that when traditional human artists create something, it's based off something they've seen before, whether they want to believe that or not, your brain isn't just making up things, it's mish-mashing up things it has a memory of seeing and "creating" something new out of it.
*typo
Incredibile. Literally downloading Blender tomorrow after this... Still a Houdini user but this looks like a game changer, thanks!
Great to hear!
I’ve been waiting for ai to combine with blender for a while. Can’t wait until you as a 3d animator could make 2d projects using your 3d animations as a base☺️
You already can actually, you need this thing called "a mild ammount of skill" :(
@@TheCrackingSpark without having to draw a bunch of frames smarty pants. I animate faster w/ 3d rather than 2d so this will be a huge lifesaver especially later on in the next 2/3 years. I can already see people adapting this technique to full renders to try to make anime from blender.
@@micah4539 I was being a smart ass because you really sound lazy af man. You can do those things already, you either need a couple tutorials or a gumroad link. Anything you can think of, chance's are someone's already done it and squeezed it for views/money. I can literally turn any still/animation into a moving painting using a .blend that I bought and it's the most realistic thing I've ever seen. The guy who made it and subsequently sold it on gumroad, he is INSANE for pulling that off, and the type of problem Solving that went into that project is nuts. Computer graphics is problem Solving, it's a puzzle, it's science and it's art. Taking that away from us because you're too lazy to watch a tut or buy a file (not even talking about making it yourself) is incredibly unfair for the community. You should ask yourself if you actually like what you do, because my guess is you don't.
@@micah4539 anime from blender? dream on. most really good animated animes like demon slayer will always be animated frame by frame, with a blend of cg a bit, but 3d animations lookin like 2d has been a thing for a while and it doesn't COMPARE to actual 2d animations. though they arent bad.
@@micah4539 and i bet you dont have that much experience in animation itself to be saying sht like that
0:28 Absolutely Agreed.
Time to finally get into Character Design Modeling with Blender.
By the time you know the basics, AI will be able to do the posing.
By the time you learn the basics, the AI 3d model generators will be public. (cuz they already exist in private)
@@DDracee Well, that too, lol.
The problem with AI is that it has its own vision, and direct and exact modifications and definitions are difficult to manage. It’s also inconsistent.
For now*
@@agustinbarquero8898 stop saying for now. it will always be inconsistent and simply not as good as human art. no one is interested in art not made by a human.
@@zaorth7851 How can you say 'always' when all you are seeing is its infancy?
It's like saying computers will always have a green screen because you witnessed the introduction of computers in mid-80s.
What you're saying now will feel laughably off in about 5 years. 10 years and people will just shake their heads at these comments.
And we know it will happen because we've seen it happen with every introduction for new technology.
@@cagnazzo82 Look at CGI in film and TV. Look have far we've progressed but still experience the uncanny valley. Maybe it will happen one day where the valley is leveled out, but humans have evolved for 100's of thousands of years to pick up on subconscious nuance and inflection throughout the human experience. That final barrier may not ever be possible to break though we may get close.
@@zaorth7851 let me know if you can overdose on cope, that's one of the biggest doses i've seen
Imagine Berserk anime with this tool.
damn!
One thing that I find foolish about this script is it doesn't take advantage of 3D depth information.
We are taking 3D data, converting it to 2D and then using a NN to infer the depth from a 2D image.
Why infer the depth from 2D when the data is right there in 3D?
I've been experimenting with rendering depth directly in blender and it is quite fun and effective.
yeah. take into account of the xyz axis and apply techniques of perspective and vanishing point.
I suppose its because they'd need to collect the training data for that, and so why bother when you have plenty collected by others (but in 2D as you say)? Also its probably a good idea to have the neural network have an idea of how depth is represented
it can't. it has no iq. and actually calling it intelligent is just perverting the term
I think is easier to develop something based on 2D, being it on 3D is not only understand the 3 dimensional space, but all the transformations that may be coming from the FOV and lenses... etc could be interesting to create 3d, based on other 3d content libraries...
@@BioClone I would agree that it is simpler conditioned on that the person does not know how to code whatsoever.
Blender scripting is extremely straightforward.
Rendering depth in blender is also extremely straight forward.
I don’t know computers but can see how this will be making big changes in films and games alike. Making it way easier to create a sub theme of duplicate items within variety that are close to matching, with further honing
I would love to see how you installed that, because I tried and didn't worked on the script side of blender.
whats the problem?
Can u do an in depth tutorial on how to install it?
Ai without any control equals terminator.
We use to believe only the boring jobs everyone hates would be where Artificial intelligence will take over.
So we didn't care.
Now we know those boring jobs were the Artist, musicians and singers.
I think some artists, the more unique ones, will survive. I’m going to laugh when AI takes all the coding jobs.
I think this will make it easier for talentless hacks to create sophisticated art, sure, but I also think those who possess the prior talent for making us feel something through their art will prevail. In the future I wouldn't be surprised to see entirely human crafted films, games, music, and artwork become more of an artisan niche. Like how you can still buy handwoven sweaters and rugs, even though machines have clearly dominated the market. We humans are infatuated with our own past, and we tend to respect, and sometimes revere those who master an "outdated" method of crafting.
@@reginaldforthright805 when A.I can actually take over coding jobs, that's when the singularity happens and A.I will replace all jobs.
@@reginaldforthright805 I'm a coder and I can't wait for it to take all the really frustrating coding jobs. That will leave me a job fixing whatever AI can't fix, which should be interesting.
the jobs won't be "taken over", the jobs will simply change towards a more productive workflow, you still need a human to tell the ai what to do and curate the results, and those humans will the artists of the future
it's kinda like when digital art was introduced and everyone was crying about digital art not being real art
fast generation of base primitive... creates insane time reduction due to less involved parts (human-speed labor serially operating click-based input)
ability to generate multiple paths with many forks to end result in a fraction of time illustrates the radical change to human labor, and thusly employment, and educational focus
I would need a tutorial to show how to install and use that.
Chatgpt can Talk you through the Setup i guess
Excellent
You can choose your own path in life whatever you want to be. What will it be, an Artist who makes unique art and enjoys the process along the way, or an AItist who presses buttons and hopes for the best pointing at the thing and say “look mom/dad/boss what I’ve just made”. Besides this rant, nice tool/video… : )
I feel you. I'm not exactly sure where all this is going, but I feel sad. As a person who draws and does 3D, I still feel sad, even though I know AI tech could be a timesaver sometimes. We dedicate our lives to develop tech that saves us time to create something else that could save us even more time. The problem is: once we get to spend that time we saved, we feel lost, as if something were missing, like a quality friend for example or that family dinner where everyone is happy and smiling and talking and not on his/her phone. In other words, I think there is so much more value in just sitting down and drawing a tree while breathing fresh air and feeling and hearing the nature around you than to ask some cool software to lifelessly draw it for you. People should first figure out what is that special thing we want to save time for?..
the magic of this is not in a single image, at least for me, It is being able to animate/rotoscope. Taking this and scaling it is where its power gets unlocked for me
The Rule34 community may see a bright future in this.
can you do a tutorial on this please
Count me in on that!
@@steveschreiner7444 i have figured out how to install automatic 1111 and controlnet. But couldn’t figure out the blender part
WOULD U PLEASE ADD AN TUT FOR INSTALLLING IT ? PLEAAAAAAAAAAAAAAAAAAAAAAASE
I love you guys cheering on every advance in AI that makes it easier to produce work.
The only problem is that in the near future, everyone will be able to produce jaw-dropping work with no skills or knowledge whatsoever, and it all becomes so easy, effortless and ubiquitous that it's worthless.
I was making photoreal work in 3D 20 years ago when it was incredibly hard to do, and people knew how hard it was to do.
Now no one cares, they just assume an AI helped and did the hard work.
The pace at which AI is moving is dizzying
Ikr, what's even the point of all that? What does a human learn from having an non-tangible machine make everything for them??
As a 2D concept artist/ Illustrator... I'm fucked up!
cant copy right any of the images
The era of laziness is upon us
holy f this is going to be insane
Where's the imagination/ intrinsic value, sure it's from a database of imagined work but let's be honest, none of the actual artists will be appreciated, I'm not calling for regulation, think that's a losing battle, all I'm asking is that people still support actual artists who bend over backward, you can tell me AI isn't there yet... well kind of looks like it is, but everything it produces unless added upon by people is void of meaning...
What I'd recommend is all artists need to start recording their process, not like most artists don't do that already...
I feel like I've started art Zbrush primarily at the worst time, I don't know whether to feel empowered or discouraged, discouraged that anyone can "do" what I can do except they've done nothing...
True designing functioning models is a lot more than what stable diffusion can do, but at the rate, things are progressing how long will it last.
I can see that Terminators are inevitable
As an artist myself it is sad to see how we went from, look at what I can do with my skills and dedication to, look at what I can... I mean, what ai can do that I don't know how to do anymore. True art is dead.
I Understand the concern that this will take a livelyhood from many and I understand that it hurts practising thousands or 10 000 hours into something and then someone (in this case something) comes and declares "I do a better job and actually anyone can do it". Still you have to take in consideration that these things didn't come out of nowhere. There are so many people and so many hours put into making these programs. A.I. is revolutionary and traditional artists sadly now need to adapt in future.
Also isn't it great that so many more can now express them self and "put their feelings into canvas" albeit it actually being a computer program helping them do it instead of humans. The end product is what matters if it makes someone feel joy. Hate them and be angry to them if they have ill intentions or bash your artistic work for no reason but don't hate them for expressing themselves.
The real threat here are the ill intentions and i.e. deepfakes where we don't have corresponding laws and systems to handle the issues yet.
To every action there is an equal and opposite reaction.
The printer didn't kill pencil drawings and photoshop didn't kill painting. It will have a huge impact, but I think a new market for "human made" art will be a thing.
“true art is dead” 🙄
The amount of efforts needed and failure potential scared me to even trying to start. The art school are insanely expensive and hard to me. I picked a job that left me some free time... And then came AI.
that ! That was my call ! I'm loving it so much, that thing is so empowering and overwelm my mind. I found a passion i never though i could have.
I see those that succeded, those i respect and admire for they gave me such a good childhood, facing the cruel reality of these thoushand hours spent reduced to mince meat by a mere lazy gamer GPU. Somehow...
I understand their disgust. Somehow... i Can see the disaster, as if they followed a lie all along.
All of that empathy i should have in any Kind of situation like this is...
Absents... Gone the day i landed on stable diffusion for myself. I want to spread it, extract the best of it. Completly unchain it. It's that powerful to me.
i'm a newborn of this depressing AI realm. I am also always open to share my knowledge on the topic for any artist in need. They deserve their jobs and i want them to strive even further than i Can ever dream...
Thanks, liked and subscribed!
Welcome!
What if you broke down a 5 minute video of a dancer into 18000 pictures assuming 60 pictures per second. Would you be able to feed those sequence of images into blender with this addon to make a 5 minute animated movie ?
This has already been done by Corridor studio and they detail out the issues they faced
@@Jeevanmn literally the day after pretty much all those issues were addressed, the field is growing way too quickly
@@DDracee hmmm i need to find those solutions!
It's only been a few months and A.I is everywhere
Blender gotta do AI physics simulation already, balls out
This is actually good, this is what it means to use AI as a tool. Artists should not be afraid. Hopefully they improve this feature.
There’s a lot more emotions going on than fear. More insult and mockery
It's still a far way from giving me the amount of control I need. Check out 1:46 for example, if I'm extruding a plane perpendicularly from the roof, I'm not going to want AI to interprete it as a dormer that's perpendicular to the ground instead. Re-rolling the AI generation to actually get what I have in mind sounds frustrating, and would give me the illusion of control rather than *actual control*.
Perhaps in a year or two it will be a feasible tool? As it stands, no, not yet, unless I didn't give a shit about consistency and art direction.
@@largeduckgoosebehind no one cares about your feelings
Do you plan to continue with this project?
This thing will take A LOT of people's jobs and as you can see human models' jobs are at stake as well!!!
Yep, it’s why i think this technology should be banned entirely as the only ones benefiting from it are the greedy rich elites while we get even poorer they get richer, not to mention that it’s been developed unethically by stealing thousands of works from every artist on the web and didn’t give them any compensation at all.
Yeah. Too many ppl on planet. Imagine some AI took their jobs. What gonna happen when they need to feed their families...
@@floodmachine1868 They will be made to live in ze pod unz eet ze boogz
Future Hirings be like: Hiring Model must have a degree on any software related courses.
Artists should consider becoming generalists as AI improving.
Thankfully a generalist is exactly what I am. That said I do understand why people are upset.
Seems like Stable Diffusion decides the final look.
Sorry for the lame question but how different is this from simply making any lame render/sketch/saving a jpeg off the internet and using it as reference within Midjourney/SD with particular weight?
It's the same, but with different workflow
It's the instant AI render happening right in the Blender interface. This workflow is waaay faster than saving jpegs every time you change the 3D model.
Identical.
@@DougM_BDXL I understand that, but given the absolutely arbitrary inconsistent nature of AI generators at the moment, I doubt anybody needs an "instant result" within Blender interface really - its not the reference thats hard to get right, its the AI output. This really solves a problem we didnt know we have.
It accounts for 3d space. did none of you fucking watch the video
We have been in the transitor age for like 90 years now.
I know that looking at devices or tvs has been your way of life but it isn't your real life.
Its now a return to first hand person to person testimony as being the most trustworthy of all..
This is a good thing because materialism has been a dark and evil period even with the good things industrial revolution brought to us .
A revival of idealism is exactly what we need ,if we are smart and wise with who we allow to control it.
The end of concept artists🥶
No doubt. ☠️
That happened a long time ago. Haha
@@feels.9304 Yesterday I was told by a friend from another country that some major studios there let go a lot of 2d concept/illustrators and just keep a handful of ppl who feed rough sketches to AI, soon this will happen to modelers, riggers & animators...
@@LHK-art Modellers and animators are a little safe. For now. As a self taught modeller, this makes me sad but I guess we've got to look at different avenues😬
Exactly, we are learning to speak to AI models to better get feedback from these models
Maybe the problem though is that there are so many idfferent models, kind of like how there's so many different humans out there
Not a single 3d artist want this
take your render and process it through stable diffusion img2img - dppm++sde karras. It give a life to your render. Such a processing should be part of every 3d renderer.
i very much do
Exactly
No one asked you opinion tbh🤡
Hobbyists do
Question Question.. I done everything but after I copied the code into blender and it rendered out only my gray mesh....
return print(f"The Automatic1111 server couldn't be found.")
except requests.exceptions.MissingSchema:
return print(f"The url for your Automatic1111 server is invalid.")
Anyone???? So close!!!
Can this tool update the mesh from the AI generated render? That would be really cool. Every generation builds upon the "new" base model mesh.
Not possible...for now
No it basically takes the model as a "sketch" and produces a flat image that closely matches what its training data tells it would be expected. Its not even a rendered raytrace of the model. More like a painting of it.
This can only be used at an early stage of developing any product, if you try to make a final product with it, it will look not finished, so use it for prototyping, and not even needed in the prototyping stage for a pro, I see no other use.
Ai dev....,,hold my bear"
2 years later....& unreal engine 6.5...
💥
Wait. This is actually very cool. The control you get is another league compared to other stable diffusion interfaces. Great job coolzilj!
I hate what AI is becoming. As a tool? Yes, it's good. Anything more than that is a massive problem.
in my opinion, it's good what Ai is becoming. modernization is meant to make things easier, despite of considering people losing jobs. wouldnt it be awful if we considered Deforestation as an issue and didn't cut trees. we wouldn't have 90% of what we have today.
this is a typical backwards way of thinking, when cars replaced horses.
@@nvda2damoon don't steal an overused quote.
that quote doesn't make your argument right, nor makes you any competent.
Just show us your awesome hollywood Movie or Game that you make.
if you don't have an awesome resume, then you're the backward one.
AI won't be able to fix your lack of skill or brain
everyone that can write can use AI, AI is dead easy, even a baby or a retarded person could use it.
it doesn't makes you any special.
learning Blender is way harder than writing prompt, you're in a disadvantage compared to the real Pro.
in fact if AI replacing jobs, you'll be the first one to be replaced because you're worthless.
just because Artists hates or fear the AI doesn't mean they're never used or experiment with AI tools before,
it's the opposite, they can utilize it better than your amateur a*s will ever be.
burn that in your mind.
I love how low quality artists are complaining about it. We knew it, the art community is shit, everyone is simply copying each others with 0 innovation, but when an ai does it, it s somehow bad
Wow. I can use this.
this is the best if the output become mesh not render image, maybe next step? but still for concepting this is next level.
nah that would suck, and completely ruin it for 3d artists
@@qwezil9476 you mean like it's been ruined for 2d artists? Give it a couple of months or a year and it will have happened. All aspects of craftsmanship will vanish.
HAVE A GOOD DAY
Hey great video thanks so much. I tried setting up everything just like instructed, and changed the controlnet setting too in the WebUI, but it's not generating anything after i hit RUN in blender, any idea what's going on ? Did you modify the controlNet model somewhere in the code? thanks so much.
Ok I got it to work, realized that i need to run it and hit render. Hahaha. But it's not really reflecting what i modelled though, like you did here in the video, so strange.
@@nahlene1973 Any idea why?
@@TheCobyRandal I realized I had to check the cmd line backend to copy paste my exact controlnet model name, then it worked.
@@nahlene1973 can you give me a insight? which line from the CMD and where to I put it in the script? I found that last bit quite confusing
@@christianwilliam1167 it’s just that u gotta input the exact model name being run in the backend and paste to replace the one provided in the example.py, for example if u are using canny, or depth, the relative model name will show somewhere near the bottom of cmd window that is powering a1111
how can i install this addon? is it free?
I like giving AI two models, then saying "generate the series of prompts which would transform the one model into the other" - it's very helpful for learning good prompting
I was going to start with Blender to learn how to model, texture and animate, and then make a career change from IT. Should I just... not even bother?
Well is there any point in learning Blender now if an AI is just gonna take over?
Yes blender is extremely fun and computers don’t feel joy so we shouldn’t offload the fun to them.
It's easy like brain surgery
Of course there is, you will learn how to optimize meshes and textures created by garbage plugins in the future :)
AI WILL NEVER BE ABLE TO CREATE solid topology with beautiful edgeflow that's production ready..so whatver AI you're nothing.
...yet
AI is the death of creativity...I am astonished and depressed at the same time...😅
It depends. For example, I always wanted to create a Visual Novel all by myself, can do music and writing but can't paint backgrounds or characters for shit. Now Midjourney can help me generate backgrounds scenes that I can reuse or modify and literally can do everything by myself! It can boost creativity.
@@SiriProject Whatever happened to merit and collaboration? I can't make music to make videos with, so I find people who can. Their music inspires me to make art, so I send my audience to go enjoy their work. Maybe one day I'll take a crack at FL Studio and attempt making music for fun. If painting is something you've struggled with, why not find people who inspire you to improve or even collab with and build a connection? After all, can you really say you've made a visual novel all by yourself if the "visual" you're using is still an amalgamation of other people's artwork?
@@SiriProject are u sure to have copyright over midjourney' s pictures? Are u allowed to monetize your novel or u will have to split revenue with your "hidden" AI partner? In that case is the same as u would ask to a fellow human artist to collab and do business together
@@NeroMai You underestimare a lot of factors. People with a talent won't usually just draw for a stranger on the internet, you'll probably have to invest a lot of time in a relationship so they trust you (almost impossible if you work full time) and even if you strike a deal, you won't be constantly be correcting your friend over how his interpretation of a scene or character is not accurate and they need to start again right? Collaborative efforts are titanical, even for fansubs, and most projects end up in the drawner
@@NeroMai What I also wanted to convey is that in my mind using the presets or soundfonts you get in FLStudio or the filters in Photoshop are no diferent from using Midjourney to provide a base you can retouch. It's just more sophisticated, but still just one more tool. Of course a lot of people mind that it generates from millions of patterns in other people's work (which could be corrected by training on copyright free content, since there are enough works out there to make it equally strong) but can you really "own" a granadient of color you used for a sunset in your drawing software, merged with the mathematical patterns of other 10000 works? Isn't that on the level of stating using a 4/4 for reggaeton is now a copyright claim? Sure, artists pay for their use of FLStudio, but don't you pay for using Midjourney?
3D printing will be wild
As a professional photographer that does a lot of e-commerce, this is frightening..
😮😮😮😮😮 thank you!! 🙏🙂
You are so welcome
no mater you want it or not, it will still evolving into a better and better version.
so just use it as your advantage or you will be left behind.
just don't wait until it ready, you will be late. when everyone get used to it, you are just start from beginning?
I remember years ago when my teacher hates digital art, and always force us to use oil and water color. you can see now digital art is everywhere. and its not a sin to do digital art and have the amazing tools called "undo".
True to a degree. Text to prompt, the most popular form of AI art generation, is apparently just a temporary placeholder for models like Midjourney or SD. Things are accelerating so quickly now that in one year the current interfaces and techniques will be largely if not entirely outdated. The experience of working with them may or may not translate, and as the technology advances, you better bet that the UI and accessibility will get better and better. Using a third party app like Discord to access MJ is pretty primitive. Everything except the tip of the iceberg is not particularly accessible to the average layperson so unless you're already inclined towards it, you're not one of the innovators, you're the user and you'll likely just benefit from further innovation. Someone starting two years from now with a significantly easier UI and better model will probably not have a huge disadvantage. The early mover advantages will mostly be given to innovators and devs in this case, at least that's my prediction.
And while I'm certainly not anti AI, it's quite a bit different than just analog > digital. No untrained artist can open Photoshop and create passable results instantly. AI equalizes a certain degree of technique, but until you're able to exact any real level of specificity with it (much more than what we see in the video) I find the comparisons to be lacking.
The best AI art results are consistently from artists who already have experience in a medium. There is an eye, a sense of restraint and intention that seems to set it apart.
But like I said, so much will change. I look forward to AI integrating with other mediums like 3D; I believe that this is where it can truly shine.
@@nindza79 you can't get the result you want just only with five words. Its different workflow than any traditional digital art.
Hi , me once again , I found (Blender) but how can I do animation ( stable Diffsusion)
AI is worst that happen to humanity till a very long time. 🤬
This can end the model industry if get good enough. Lot of people can say oh but its not a real person, but somehow fashion models are far from the average person realistic fit of clothing
its sad, its not impressive. AI doing our jobs and profession is not a good future, its just so sad and idk how no one feels bad about this at all.
I feel you. Alot of people are mindlessly talking about all the cool stuff they will create with AI but dont realize that nobody will care.
There are always two sides to a coin.
powered by stable diffusion, does this still rely on the scraping of artists work? until that's solved I can't condone the use of AI
god i hate it... lol
How do u use it tho?