If you want to enhance a coarse render, which engine and model do you have to specify for your reference image? The idea is to ask the artificial intelligence to make your render more realistic while preserving the textures and perspective of the scene.
I think if I understand correctly you would use a reference image but a very low creative setting so the AI takes from the ref image but doesn’t lose your existing textures and stuff
If you keep the same prompt and remember the exact setting you used, you should be able to make changes and have similar results! But sometimes even with a similar prompt things will change
Great info! Thank you so much! Do you think if we do a low quality 3d renders (3ds max) with our specific furniture and textures, AI would be able to enhance the image to a more realistic final image (using the right text prompts)? Also, are we able to add a specific furniture during the process? Thanks!
Just subscribed to your channel haha... Absolutely agreed with you Rundiffusion's controlnet is awesome. Except when I tried to sign in to my own gallery, it kept saying wrong user and password. Has anyone experience it? Or was it only me? haha... Or was it still on Beta version
Each time you run a session, it provides a username and password they you can enter on the right side where you should be able to access your gallery too. Lmk if this doesn’t work
Search it on google and you should be able to download from hugging face. Otherwise this link should work, huggingface.co/SG161222/Realistic_Vision_V2.0
Awesome video - starting getting into this as a fellow architect, but unfortunately don't have the option for Scribble under model (next to preprocessor), any idea why?
@@thearchitecturegrind by d way i think these applications can b included easily in platforms such as twinmotion d5 or lumion as an extra optional aspect of having a pre composition general atmosphere / textures auto apply system..... D assets r already there and it's way less computing demanding from d motherboard cs it deals with putting together multiple pics assets with a 3d understanding rather than importing actual 3d shapes.... Who knows maybe d merge will happen someday between the traditional 3d visualisation platforms and Ai😎
![We tried to compete with AI... [AI vs. ARCHITECT]](/img/n.gif)
You're on the same path I am with attempting to make architecture easier in rendering and simple drafting
Best of luck!
Thank you!, do you know about any tool that would help create a 3d model of a Land Subdivision (Town Plan) to 3d?
This is so awesome!!
Thank you, great presentation, subscribed
Thank you very much!
If you want to enhance a coarse render, which engine and model do you have to specify for your reference image? The idea is to ask the artificial intelligence to make your render more realistic while preserving the textures and perspective of the scene.
When you say coarse render, are you referring to one that is close to final?
I think if I understand correctly you would use a reference image but a very low creative setting so the AI takes from the ref image but doesn’t lose your existing textures and stuff
I made the video!
awesome video super cool!
what will if there are revisions can i get the same original mood that i like?
If you keep the same prompt and remember the exact setting you used, you should be able to make changes and have similar results! But sometimes even with a similar prompt things will change
Hey ,,i need your help could you plz one image rendering for my college project, coz i haven't laptop
Yes I can help you
Great info! Thank you so much!
Do you think if we do a low quality 3d renders (3ds max) with our specific furniture and textures, AI would be able to enhance the image to a more realistic final image (using the right text prompts)? Also, are we able to add a specific furniture during the process?
Thanks!
Yea I think that would be a great thing to test! I don’t think you will be able to add specific furniture unless you put it in the 3D max model
Just subscribed to your channel haha... Absolutely agreed with you Rundiffusion's controlnet is awesome.
Except when I tried to sign in to my own gallery, it kept saying wrong user and password. Has anyone experience it? Or was it only me? haha... Or was it still on Beta version
Thank you! I will look into that issue, but I have not had an issue with that
Each time you run a session, it provides a username and password they you can enter on the right side where you should be able to access your gallery too. Lmk if this doesn’t work
good tutorial bro
Thank you!
Superb tutorial...does it require heavy pc or graphics or ram
no, it does not, That is what's great about it! Although I recommend having a good amount of free storage if you plan on downloading stable diffusion.
@@thearchitecturegrind thanks
where I can get realisticvision V20.ckpt?
Search it on google and you should be able to download from hugging face. Otherwise this link should work,
huggingface.co/SG161222/Realistic_Vision_V2.0
❤❤
good video ! what sort of quality are renders produced ?
Thank you! Very high quality
Awesome video - starting getting into this as a fellow architect, but unfortunately don't have the option for Scribble under model (next to preprocessor), any idea why?
I will look into it Joshua, weird problem you’re having
Do you pay per rendering?
you pay by time, so on average its $0.50 per hour, but for 15 minutes it is like $0.13
Is it only work for images or can we make whole house video too?
Does it work with proper measurements of the land ?
Unfortunately it only does images for now. Some are beginning to create 3D panoramic views though.
It does not work with measurements or dimensions yet, this is a purely visual program for now.
❤❤❤
wow
Thanks
Coolll
thank you!
@@thearchitecturegrind by d way i think these applications can b included easily in platforms such as twinmotion d5 or lumion as an extra optional aspect of having a pre composition general atmosphere / textures auto apply system..... D assets r already there and it's way less computing demanding from d motherboard cs it deals with putting together multiple pics assets with a 3d understanding rather than importing actual 3d shapes.... Who knows maybe d merge will happen someday between the traditional 3d visualisation platforms and Ai😎
@@carvingtheway That would be something if what you said started to work and someone made that
Does it work with nsfw drawings?
Not at the moment
Rendering using vray takes years and years.
Yes Ik it requires a lot of effort, good quality tho
this is not AI.This is made by a human, i did this in 1 year of school in 2002
Nice 👍🏼
That's what lagging for AI... it can't figure out things like human😂
Yes Ik. When it does figure out humans it might be weird tho
türkiyedeki özel şirketler içinde amin
😡😡😡😡