Jump to content

AI Visualizations - Sketches


Recommended Posts

  • 1 month later...

hello all, 

 

I have been exploring the AI Visualizer tool and am having some difficulty with it. My goal with this tool would be to use it with a massing model to generate sketch elevations to sell a vision to a new client. My problem is that the AI does not seem to get what i am trying to do, add windows, doors and add material textures. The AI always seems to generate a fence and does not seem to keep the shape. Maybe I am writing the prompt wrong, I have tried many ways of writing it. The more detail I put into the prompt the more the AI does not follow the massing model.  These are some of the better images that it has come up with, I have re-generated this about 40 times.

 

I have also tried AI on other files that have been developed more than a massing model and I am still having difficulty getting a quality image developed. I keep getting images that look like they have been poorly stitched together with photo shop. 

Screenshot 2024-05-23 at 12.57.53 PM.png

image.png

image 2 - 50%.png

image 3 - 50%.png

image 4 - 0%.png

  • Like 2
Link to comment
4 hours ago, R-searsarchitects said:

My goal with this tool would be to use it with a massing model to generate sketch elevations to sell a vision to a new client. My problem is that the AI does not seem to get what i am trying to do


Wouldn’t it be faster and more fun to layout and sketch those ideas yourself?

 

Having a computer do the creative part isn’t creative.

  • Like 3
Link to comment
  • 3 weeks later...
  • Vectorworks, Inc Employee

I left my heart....in San Francisco 🎵🎶

Last year's AIA main expo was at this charming city. I thought it would be nice to use it as inspiration for some portraits I need for a new interiors model.

Prompt: Beautiful pen and ink sketch of San Francisco year 2023, photo realism, colored, detailing, daytime.

Screenshot 2024-06-10 at 9.55.18 PM.jpg

San Francisco Sketch.jpg

  • Like 2
Link to comment
  • 5 months later...
  • Vectorworks, Inc Employee

With the release of the latest AI Visualization update, recalling prompts and setting is now possible. I just wanted to share this artistic style that is very appealing to me. I am including a screenshot and the vectorworks file so you can recall the prompt and apply it to your projects. Let me know how it works for you.

A night cafe - artistic.png

A night cafe.vwx

  • Like 4
Link to comment

I opened the file but I'm being denied the AI Visualiser for some reason so can't check the settings...

 

Screenshot2024-11-22at06_45_27.thumb.png.3883e298aafd37e91568fd7f2300ae3a.png

 

But I guess in the absence of a model these were produced using text prompts only...? @Luis M Ruiz would love to see the results were you to use a very basic VW model as the basis for the imagery.

 

Any ideas why AI Vis is greyed-out?

Link to comment
  • Vectorworks, Inc Employee

@TomW I have heard of this problem a couple times and I don't know what the cause is.  We do not do any extra product license checking for this menu (like for Spotlight etc menus to enable them only for certain conditions).  The workaround is to re-add the menu to your workspace.  Sorry for the disruption.

 

Link to comment
  • Vectorworks, Inc Employee

"A boutique store, interior view, an architectural sketch, with a detailed play of light and shadow, the store has an ornate, classic style, the drawing is made using ink and watercolor, a dynamic and creative effect, the contrast of light and dark shades, loose brush strokes, perspective representation, 3D, HDRi"

vwx int.jpg

Int Sketch.jpeg

  • Like 1
Link to comment
  • 3 weeks later...

I played around a bit with AI image generation. I noticed that different styles require different levels of creativity to achieve satisfactory results. For a photo style, 69% seems optimal for me to ensure that the image doesn’t deviate too much from the original plan. However, at this level, humans still don’t appear in the images. In the case of a drawing style, 69% is too much, AI changes too many details. In this case, 29% worked best for me. I created a simple 3D drawing as a base and generated images in different styles based on it, as follows.

I also noticed that it gives much better result if I always refresh image before starting a new process. It seems that without it the previously generated image influences the result.

image.thumb.png.1a120d54f2bcc24089fb5aa4a23ff64f.png

Edited by Zsombor
  • Like 1
Link to comment

Scene description was always the same: 

outdoor city park with playground

trees

benches

playing kids

parents

lots of trees in the background

 

I changed only style description. This image:

 

professional outdoor photo

agfacolor

depth of field 

69%

image.thumb.png.cc00e6782223e078faa388dcf3263230.png

Edited by Zsombor
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...