We value your input! Please participate in Archicad 28 Home Screen and Tooltips/Quick Tutorials survey
on 2024-10-02 03:30 PM
Our solution to address the intricate requirements of visualization is to offer a versatile and adaptable set of tools. Read more about the rest of the spectrum of our offerings here. One of the items that we dwell on in this toolkit is AI Visualizer.
This article is based on the valuable insights shared during the Archicad AI Visualizer Webinar.
For detailed information on recent updates and key changes to AI Visualizer read this post on Graphisoft Insights.
In the fast-paced world of architecture, the ability to quickly visualize and iterate on design concepts is essential. With Archicad 28’s new AI Visualizer, architects can now elevate their conceptual design process by generating visual representations directly from mass models with the help of AI. This powerful tool integrates seamlessly into the Archicad workflow, enabling architects to explore a wide range of design possibilities from the early stages.
.
Getting Started with AI Visualizer
The AI Visualizer comes in Archicad 28 and now it is hosted in the Cloud. It is available to Archicad Collaborate users and students. It is designed to assist architects during the early conceptual stages of design. It takes an image of the basic mass model, which is essentially a simplified version of a building model that outlines the general shape and volume of the structure, and transforms it into a detailed and visually appealing representation of what could be. This capability is particularly useful for architects who need to present initial ideas to clients or stakeholders without investing significant time in developing full-fledged 3D models.
To begin, create a simple mass model in Archicad, which can be as simple as a few blocks representing the major volumes of a proposed building. Once the mass model is ready, you can use text prompts to guide the AI in generating different design alternatives. After running Generate with your prompt, the AI creates an image of the 3D view and begins refining it.
For example, a prompt like "Modern office building with wooden surfaces" could yield an iteration of your mass model, façade aesthetics, or other iterations that could be tried with different materials, each with different aesthetics. For this test, it resulted in the following example:
Exploring Design Variations with Creativity Slider
One of the standout features of the AI Visualizer is its ability to generate diverse design variations from a single mass model. By adjusting the Creativity Slider, you can control the degree to which the AI deviates from your original model. Lower settings will produce designs that closely adhere to the mass model’s geometry, while higher settings allow the AI to introduce more creative elements, such as innovative façade patterns, unique structural forms, or unexpected material combinations. For example, here we have a prompt “Create a realistic image of a modern office building with wooden surfaces and glass” and the same prompt is used with different percentages of the creativity slider:
Creativity at 20 percent:
"Create a realistic image of a modern office building with wooden surfaces and glass"
Creativity at 50 percent:
"Create a realistic image of a modern office building with wooden surfaces and glass"
Creativity at 80 percent:
"Create a realistic image of a modern office building with wooden surfaces and glass"
Refining and Locking Design Elements
"Modern office building with brick facade"
As you explore various design options, the Lock Result feature becomes invaluable. This function allows you to preserve the design that resonates with your vision while continuing to experiment with other aspects. For example, if the AI generates an accurate representation of your idea, you can lock this aspect and explore different configurations for the building’s form or the integration of outdoor spaces, ensuring consistency across multiple iterations. When you have a design you like, the Lock Result feature remembers previous prompts, and you can further customize the image. Here are some examples of this feature:
Facade Option 1:
"Add some wooden strips"
Facade Option 2:
"Change the facade to anthracite zinc"
This feature is particularly useful during the conceptual phase when presenting different options to clients. You can maintain a consistent design narrative while demonstrating flexibility in other aspects, such as materiality or spatial organization
These settings allow for fine-tuning and include Iterations and Prompt Strength They can be used in conjunction with the Lock Result feature.
Iterations:
This controls the number of times the generation process is repeated. A higher number of iterations results in more accurate outputs but takes longer to complete, while fewer iterations are faster but less precise. The default value is set to 35, which provides a good starting point, though you are encouraged to experiment within the range.
Prompt Strength:
This setting determines how closely the AI adheres to the given prompt. A higher prompt strength ensures the AI follows the instructions more strictly, potentially reducing creative flexibility. Lowering this value gives the AI more freedom to be creative. Adjust the prompt strength to find your ideal balance.
This parameter controls the resolution of the generated image, directly affecting the time it takes to produce. Higher resolutions result in better image quality but require more time to generate. The default size is 800x600, with a maximum resolution of 1024x1024 available for use.
Combining this with the Upscale Image will allow you to export your image at a larger scale: the long side will be increased to 4096 pixels,
and the shorter size will be proportional to the original image size. Activate this function before generating the images. If you wish first to generate a lower-resolution image as a test and then upscale it, be sure to turn on Lock Result to preserve your original result.
Practical Application for Early Design Stages
Overall, Archicad 28’s AI Visualizer is a powerful tool for architects working on early-stage design projects where rapidly flowing ideas and iterations are crucial. Starting with a basic mass model, you can use the AI Visualizer to quickly generate a series of design options. Each option can explore different facade options, all guided by simple text prompts.
Create the Mass Model: Develop a basic mass model in Archicad that represents the overall shape and volume of the proposed building. Which the AI Visualizer will take a picture of. Make sure to adjust the 3D view angle.
Apply Text Prompts: Use descriptive text prompts to guide the AI in generating different design variations. For example, “Create a realistic image of a modern office building with wooden surfaces and glass.”
Adjust Creativity: Experiment with the Creativity Slider to explore both conservative and bold design variations.
Lock Key Elements: Once there is a particular design you like and would like to stick to similar designs, use the Lock Result feature to preserve it while continuing to refine other aspects of the design.
Creating the Options: You can create different options with the previously selected Lock Result feature, change the prompts about the model to look like you would like it to look, and run it multiple times.
Save the Options: If you see any design that you like, you can choose to refine this image with the Upscale Image option and afterward save it.
Whether you’re working on a high-rise, a cultural center, or any other large-scale project, the AI Visualizer offers the flexibility and power to push your conceptual design inspirations to new heights and help you develop new ideas about your concept and results can look impressive.
This concept seamlessly integrates into interior design projects, offering a fresh approach to creating aesthetically pleasing spaces. Explore further by reading an insightful article on Enhancing Interior Design with AI Visualizer for Archicad, which explains how AI tools enhance the design process.