Well done.... I'm a retired Architect who used to do all his renders by hand..... I use AutoCAD, Revit, SketchUp and Blender these days....never heard of these AI programs before this video...... thanks for sharing
I have been a Architectural Illustrator since 1976, I think these programs will do the same to us as humanoid Robot Cashiers will do to Walmart Cashiers!😁 I remember bringing a couple of computer Artist from NASA to a group meeting of designers in the late 70's. The remarks after I will never forget,many said “Computers will NEVER be able to replace what we do or as fast!” Which was true to an extent, but if you ignored WEB, Aldus Pagemaker, Illustrator and the king, Photoshop, you would find yourself behind in the business.
I would like to see AI rendered floor plan stylization. So being able to input a 2D black & white floor plan and get a shaded, colored,(maybe landscaped)furnished interior, etc. Thanks for your videos, friend.
Fantastic, Salmaan. You are helping us understand what is today one of the most anxiety-producing yet promising technologies! To be honest, I run most of your videos at .75 speed so I don't miss anything haha. What are your thoughts about Midjourney--have you found better ways to rein it in and make it do more of what you want it to do? Best, James
I believe you would might get a better output by using a shaded/textured image where the render engines can differentiate between glass and other materials. It also seems like you missed VISOID from the list! You should give it a try :D
Fantastic video Salmaan, learnt a good stuff from it. Quick question- i am in Booth design business where we create 3D Booth designs for our clients. At alot of times, we have to show the same design to a different customer and rendering after changinh the logos and color theme is long. Do you have any recommendation on any platform that can help me in swapping the cx logo and some other sections from my initial or input design? Your advise will be highly appreciated.
Very good to get an idea about AI inputs - I am an old generation architect , 70 + years but this really fills me with a lot of curiosity, I want to ask if you do work on Procreate ? If yes do let me know so that I can follow - thanks
I'm glad to hear your thoughts on AI. I am a beginner with Procreate myself and look forward to exploring the software and see how it can benefit Architects!
@@salmaanarchitect actually, iPad gifted by my daughter started me on to procreate and digital drawings world and I really am very keen to explore and do some work - it’s satisfying
Can I please ask which option is best using autocad architectural design of the building and results in decent carry internal rendering/interiors. Are there any output compatible with Meta Quest 3 as well
Best ai platform would be Invoke AI, that you can install for free on your own pc. Of course a graphics card with 12 gb would be good. Here you can select from houndrets of ai files from civit. Ai, created and finetuned with stable diffusion models. Use your image as depth mask, controlnet, or ip adapters, finetuned in rhe canvas parts, inpaint.. More to learn but also better results on the artists friendliest user interface I've ever seen
I don't understand what Sameer is saying. Does he mean that Midjourney cannot accept architectural sketch input and must rely entirely on the text prompt? If that is true, then Midjourney should not be included in this comparison.
What he means to say is, Midjourney does accept Architectural Sketch input but it won't give output exactly the same as input. It adds its own creativity and Midjourney by itself is a platform meant to make creative outputs, not exact ones.
@@stylyai how do you mean this.. if you look at midjurney in all its artistic beauty, it still lacs of many functions.. As an architect, i can take a screenshot of a 3D Cad file, feed it into stable diffusion using A1111 or better InvokeAI, and use it as a depthmap, as a lineart controllnet, and also can import Loras or IP Adapter to get by far more precision for exactly what i want to achieve.. this is what i mean with customasible. Now A1111 has somewhat a clunky interface eeven if there are many funktions, of which some are not realy good for professional workflow, because you dont instantly see that they are restricted or need a permission. Comfy Ui on the other hand is extremly advanced and customazible but very difficult to learn, and not artist friendly. In my opinon, Invoke AI comes with the advantage of both worlds.. first Class interface, all the tools, you need for more conterol, and furthermore nodes if you even want to go furthere using nodes.. Of course this is more complex as the webinterfaces currently available. but its also more precise. And thers another advantage... i can run it in our company localy, without beeing afraid, that images created for international architectural competitions can be misused..
No. I teach this in our 500 employee architectural company. It's allready increadible for art. But for architecture the precision is missing and the ai s are not trained specialized for architects.. Still a long way, but allready helpful for ideas and improving half finished renderings, but not to a level like vray
Just go to stable diffusion programs. Most advanced interface maybe is comfy ui, but only nodes. A1111 is good but a to overloaded interface with many functions. Best.. INVOKE ai.. Brst interface, extremely artist friendly, inpaint, outpaint, ip adapters, controlnet, totaly free and still extremely poweful
If you want people to understand what you are saying then slow down your talking speed. It will give the listeners a chance to get accustomed to your accent. otherwise a great video.
![We tried to compete with AI... [AI vs. ARCHITECT]](/img/n.gif)
Well done.... I'm a retired Architect who used to do all his renders by hand..... I use AutoCAD, Revit, SketchUp and Blender these days....never heard of these AI programs before this video...... thanks for sharing
I have been a Architectural Illustrator since 1976, I think these programs will do the same to us as humanoid Robot Cashiers will do to Walmart Cashiers!😁 I remember bringing a couple of computer Artist from NASA to a group meeting of designers in the late 70's. The remarks after I will never forget,many said “Computers will NEVER be able to replace what we do or as fast!” Which was true to an extent, but if you ignored WEB, Aldus Pagemaker, Illustrator and the king, Photoshop, you would find yourself behind in the business.
I would like to see AI rendered floor plan stylization. So being able to input a 2D black & white floor plan and get a shaded, colored,(maybe landscaped)furnished interior, etc.
Thanks for your videos, friend.
Nice idea, stable diffusion can do that.
I think they have the same main problem is consistency. Cannot get the same rendering features for different scenes
Fantastic, Salmaan. You are helping us understand what is today one of the most anxiety-producing yet promising technologies! To be honest, I run most of your videos at .75 speed so I don't miss anything haha. What are your thoughts about Midjourney--have you found better ways to rein it in and make it do more of what you want it to do? Best, James
MJ this time got a 2 out of 10 lol
Great explanation, Salmaan. Thanks very much. For me, I guess that MNML is the most complete and friendly AI tool to work with.
THIS IS FANTASTIC!!!!!!!PLEASE MAKE A VIDEO ON AI GENERATED FINISHED FLOOR PLANS.. IT WOULD BE MUCH HELPFUL
Interesting material. For me, number 4 is way better than the previous one. Solid 9
E assim, eu descobri a IA para renderizar meus projetos. Obrigada, pena que não é em português. Obrigada.
You've really done well for me. Thank you so much
Hi, I just wanted to inform you that on Fabrie Ai Per image it takes 10 credits which leads to 30 images per day and not 300 images
It's fantastic to see Salmaan and Sameer working together on this - keep going 🙌🏽😁
Thanks Salman, which one is better?
Insightful video on latest AI updates Salman !! Keep up the good work 👏
Great video 🙌 please make a video on ai generated floor plans generators (specifically for complex residential plots/areas)
Thanks
I believe you would might get a better output by using a shaded/textured image where the render engines can differentiate between glass and other materials. It also seems like you missed VISOID from the list! You should give it a try :D
Thanks Mr. Salmaan
Jazakallah Khair (جَزاكَ اللهُ خَـيْراً)
Fantastic video Salmaan, learnt a good stuff from it. Quick question- i am in Booth design business where we create 3D Booth designs for our clients. At alot of times, we have to show the same design to a different customer and rendering after changinh the logos and color theme is long. Do you have any recommendation on any platform that can help me in swapping the cx logo and some other sections from my initial or input design? Your advise will be highly appreciated.
Hi Salmaan nice to see u
Thanks for ur video but can u choose 1 is the nice website that better then 10
Nice Video Salman 👍 Keep up the good work🙌
Very good to get an idea about AI inputs - I am an old generation architect , 70 + years but this really fills me with a lot of curiosity, I want to ask if you do work on Procreate ? If yes do let me know so that I can follow - thanks
I'm glad to hear your thoughts on AI. I am a beginner with Procreate myself and look forward to exploring the software and see how it can benefit Architects!
@@salmaanarchitect actually, iPad gifted by my daughter started me on to procreate and digital drawings world and I really am very keen to explore and do some work - it’s satisfying
Thank ❤, but way 8/10 if excellent
Very interesting ... Thank you
Can I please ask which option is best using autocad architectural design of the building and results in decent carry internal rendering/interiors. Are there any output compatible with Meta Quest 3 as well
That's very helpful, thank you very much.
Thanks Salman. It was very informative.
Great that's much 100%effective
THANK U for this video
im interested in sampling my design .
Well done mate. Great video. Keep up the good work.
My 3d model is showing wireframe in realistic view.Can you please help?
Great video Salmaan. Keep them coming. 😊
Thanks for this video 😊
It seems that there is still no accurate ai render engine to replace with v-ray and etc
you can achieve this by mentioning "v-ray" in the prompt mentioning using Stable Diffusion. (95% success rate)
does this works with archicad?
Thanks bro
Thank you
Which file format is the image from SketchUp.?
Jpg
Fantastic video!
Good work brother ✌️
fantastic!!
Is there a method to control material & lighting in AI render
Not completely but can be customised with prompts or by using replace option available in a few AI platforms.
Nice. I am from Pakistan
Can this AI rendering work with archicad or is it only on SketchUp?
You can take screenshots from any modelling software and use with these AI
Hey Salman =, Can you help me to make one 3d render for my commercial building
Good job
Your Rating criteria is very strange by the way thanks for information.
Hi, can you have your own tree models and background buildings…?
You could try adding them within the model before taking a screenshot into AI. And hopefully it should work the same way!
Can it work off line
All the platforms mentioned here works online
Best ai platform would be Invoke AI, that you can install for free on your own pc. Of course a graphics card with 12 gb would be good.
Here you can select from houndrets of ai files from civit. Ai, created and finetuned with stable diffusion models.
Use your image as depth mask, controlnet, or ip adapters, finetuned in rhe canvas parts, inpaint..
More to learn but also better results on the artists friendliest user interface I've ever seen
I don't understand what Sameer is saying. Does he mean that Midjourney cannot accept architectural sketch input and must rely entirely on the text prompt? If that is true, then Midjourney should not be included in this comparison.
What he means to say is, Midjourney does accept Architectural Sketch input but it won't give output exactly the same as input. It adds its own creativity and Midjourney by itself is a platform meant to make creative outputs, not exact ones.
@@salmaanarchitect Simply put, it is different from the other AIs compared in this video.
i think is better to save the initial picture without shadows. shadows spoil some zones.
Hi, why don't you try stable diffusion which is completely free?
Exactly, and there the best interface for architects is invoke ai
@@tomschuelke7955 Not true, stable diffusion is not customasible and easy to use
@@stylyai how do you mean this.. if you look at midjurney in all its artistic beauty, it still lacs of many functions.. As an architect, i can take a screenshot of a 3D Cad file, feed it into stable diffusion using A1111 or better InvokeAI, and use it as a depthmap, as a lineart controllnet, and also can import Loras or IP Adapter to get by far more precision for exactly what i want to achieve.. this is what i mean with customasible.
Now A1111 has somewhat a clunky interface eeven if there are many funktions, of which some are not realy good for professional workflow, because you dont instantly see that they are restricted or need a permission. Comfy Ui on the other hand is extremly advanced and customazible but very difficult to learn, and not artist friendly.
In my opinon, Invoke AI comes with the advantage of both worlds.. first Class interface, all the tools, you need for more conterol, and furthermore nodes if you even want to go furthere using nodes..
Of course this is more complex as the webinterfaces currently available. but its also more precise.
And thers another advantage... i can run it in our company localy, without beeing afraid, that images created for international architectural competitions can be misused..
@@tomschuelke7955
@stylyai
You should try 'Forge Webui" or SwarmUI. Much faster than Automatic 1111.
I'm an Architect and been in the field for over 10 years. Seems like its my first time hearing about AI. Anybody else with an opinion?
Can you some elevation renderings of my house? I’ll pay you
Pls email salmaanmohamed18@gmail.com
is not free
how to release your good idea n design on internet for free
Do you think architectural 3D visualizers will lose jobs due to AI in the future?
No. I teach this in our 500 employee architectural company.
It's allready increadible for art. But for architecture the precision is missing and the ai s are not trained specialized for architects.. Still a long way, but allready helpful for ideas and improving half finished renderings, but not to a level like vray
Prome ai
which ai among this is truly free..
Except Midjourney, all of them have limited free trials.
Just go to stable diffusion programs. Most advanced interface maybe is comfy ui, but only nodes. A1111 is good but a to overloaded interface with many functions.
Best.. INVOKE ai.. Brst interface, extremely artist friendly, inpaint, outpaint, ip adapters, controlnet, totaly free and still extremely poweful
Xona AI is better than all of these
выглядит не жизни способно, в таком жить работать? функционал?
Bhayya tumhari video bahut acchi hai lekin ye AI ekdam baklol hai...
Kuch bhi kayde ka kark nahi deta ...
It’s not working properly, I try paid version
If you want people to understand what you are saying then slow down your talking speed. It will give the listeners a chance to get accustomed to your accent. otherwise a great video.
😂 the guy puts up a pretty nice video, but people want to control even his talk 😂
Brother, you can slow the playback speed as well 😅 if you fail to follow his speed.
This video is a work flow not a tutorial. The speed and the video is perfect
I personally had no problem following along......what's wrong with his accent
Captions are an option, my dude. I don't have a problem understanding him.
And as other said, slow thr playback speed. Xenophobe.
If anyone said to me that he renders with AI and not create the model themselves I'm gonna skip them. This is pure 💩💩💩
wow
What is Samir’s instagram ID?
@sameer.arch on IG