MASTERING composition in AI art (you can set your OWN layout)

Stable Diffusion Tips - With Fooocus
13 Dec 202305:49

TLDRThe video script outlines a technique for achieving better shot composition using Photoshop and AI. It guides viewers through the process of creating a composite image by separately generating and editing two characters, a xenomorph and a soldier, before combining them with a jungle background. The tutorial emphasizes the importance of image quality and depth of field, resulting in a more realistic and dynamic final composition.

Takeaways

  • 🎨 Frustrated with shot composition? Try breaking the process into separate parts and reassembling in Photoshop for better results.
  • 🖼️ Start with a blank project and disable the photo input for a clean slate.
  • 📐 Set the canvas size to 1344x704 for a 16:9 aspect ratio, close to the standard.
  • 🚀 Choose 'speed' for processing performance and use the top three styles for simplicity.
  • 👾 Use specific prompts like 'Hyper realistic xenomorph running towards viewer' for detailed images.
  • 📸 Request 'full body view' and '8K' for higher quality and more realistic images.
  • 🔄 Generate multiple images to choose the best one for your composition.
  • 🌟 Use a service like Creait Pixel Cut to remove backgrounds from your characters.
  • 🌲 Reuse the background from the AI-generated image for a cohesive color scheme and atmosphere.
  • 🖌️ Composite the images in Photoshop, adjusting scale and adding blur to the background for depth of field.
  • 📌 Re-run the composite image through the AI with a new prompt to finalize the scene.
  • 🔧 Adjust settings to 'Quality' for the final AI run to ensure the best output.

Q & A

  • What is the main issue the video aims to address?

    -The video addresses the common frustration of not achieving the desired shot composition when working with image prompts, particularly in scenarios involving a monster chasing a man.

  • What is the proposed solution to the composition issue?

    -The proposed solution is to process the image in separate parts and then composite them back together in Photoshop, which allows for more control over the final composition.

  • What are the initial settings recommended for the blank project in Photoshop?

    -The initial settings recommended include disabling the photo or image input, setting the screen resolution to 1344 by 704 (close to 16:9), choosing 'speed' for processing performance, and using only the top three styles to keep it simple.

  • How does the video suggest selecting the best image for the xenomorph?

    -The video suggests running the prompt with 'Hyper realistic xenomorph running towards viewer' and 'full body view' in 8K resolution, then choosing between the generated images based on preference.

  • What is the purpose of changing the prompt for the second character?

    -The purpose of changing the prompt is to create a second character, a terrified soldier, with a focus on showing fear on his face to add more dynamic and action elements to the composition.

  • How does the video suggest removing backgrounds from the characters?

    -The video suggests using a service like Creat Pixel Cut, which offers a free option, to remove the backgrounds from the characters before importing them into Photoshop for further compositing.

  • What technique is used to incorporate the jungle background into the composition?

    -The technique involves grabbing parts of the xenomorph's background and dragging them towards the middle, erasing the character, and using the resulting image as a prompt to maintain the color scheme and steaminess of the jungle.

  • How does the video suggest adding depth of field to the composite image?

    -The video suggests adding a soft amount of blur to the background image in Photoshop to create a depth of field effect, enhancing the overall composition.

  • What prompt is used to generate the final composite image?

    -The final prompt used is 'Hyper realistic steamy jungle scared Soldier being chased by xenomorph' with the addition of '8K photo' for higher quality.

  • What is the benefit of manually composing the image rather than relying solely on AI?

    -Manually composing the image allows for more control over the layout and elements, resulting in a better outcome than relying on AI, which may not always arrange the elements in the desired way.

  • How does the video encourage viewer interaction?

    -The video encourages viewers to like or subscribe if they find the content helpful, and to provide feedback or suggestions for improvement in the comments section.

Outlines

00:00

🎨 Image Composition with Photoshop and AI

This paragraph discusses the challenges of achieving the perfect shot composition and introduces a technique to overcome these issues. The speaker explains a process that involves generating separate parts of an image using AI and then combining them in Photoshop. The process begins with setting up a blank project, choosing the right screen dimensions, and selecting appropriate settings for processing performance and style. The speaker uses a hyper-realistic xenomorph running towards the viewer as an example, specifying a full body view and 8K resolution for a more photographic image. After generating two images, the speaker selects the preferred one and repeats the process with a different prompt, this time creating a terrified soldier character. The images are then saved for later use in Photoshop, where the speaker also discusses using a background removal tool like Creal Pixel Cut to prepare the characters for compositing.

05:03

🌲 Composite Image Prompt with Steamy Jungle Background

In this paragraph, the speaker continues the image compositing process by discussing the addition of a background to the composite image. The speaker opts for a jungle background from one of the previously generated images and describes a method to capture the color scheme and atmosphere of the jungle for use as a background in the composite. The speaker then moves on to Photoshop, where the two characters (xenomorph and soldier) are layered, with the xenomorph in the background and the soldier in the foreground. The background is slightly blurred to create depth of field. The final step involves using the composite image as a prompt in the AI to generate a detailed, high-quality image of a realistic, steamy jungle scene with a scared soldier being chased by a xenomorph. The speaker emphasizes the effectiveness of this manual compositing method over relying solely on AI for layout and composition, and invites feedback and engagement from the audience.

Mindmap

Keywords

💡Composition

Composition refers to the arrangement of elements in a work of art, photography, or other visual media to create a unified and aesthetically pleasing image. In the video, the main theme revolves around improving the composition of a shot by carefully positioning characters and background elements. The speaker is trying to achieve a more realistic and engaging scene by adjusting the composition of the image.

💡Photoshop

Photoshop is a widely used software program for image editing and manipulation. It allows users to alter images in various ways, such as adjusting colors, removing backgrounds, and compositing multiple images. In the context of the video, Photoshop is used to combine separately processed images of a xenomorph and a soldier to create a final composite image with a desired composition.

💡Prompt

In the context of the video, a prompt is a specific instruction or request given to an AI or a software to generate a particular image or result. The speaker uses prompts to guide the AI in creating the desired characters and scenes, such as 'Hyper realistic xenomorph running towards viewer' or 'Terrified soldier running towards the viewer'.

💡Image Processing

Image processing involves the manipulation of digital images to achieve desired effects or outcomes. This can include tasks such as enhancing image quality, altering colors, or removing backgrounds. In the video, the speaker processes the images of the xenomorph and the soldier separately before combining them in Photoshop to achieve a higher quality result.

💡Background Removal

Background removal is the process of separating the main subject of an image from its background. This technique is used to isolate characters or objects for easier compositing with other images or backgrounds. In the video, the speaker removes the backgrounds from the generated images using a tool like Creal Pixel Cut to prepare them for further editing in Photoshop.

💡Depth of Field

Depth of field refers to the range of distance within a scene that appears acceptably sharp and in focus. It is a photographic technique that helps create a sense of depth and dimension in an image. In the video, the speaker adds a soft blur to the background image in Photoshop to achieve a depth of field effect, making the scene look more realistic and three-dimensional.

💡AI

AI, or artificial intelligence, refers to the simulation of human intelligence in machines that are programmed to think and learn like humans. In the video, AI is used to generate the initial images of the xenomorph and the soldier based on the prompts provided by the speaker.

💡Image Prompt

An image prompt is a visual or textual cue that serves as a starting point or inspiration for creating a new image or piece of content. In the video, the speaker uses image prompts to guide the AI in generating the desired scenes and characters, and later as a reference for editing and compositing in Photoshop.

💡Character

A character in this context refers to a person or creature depicted in an image or visual narrative. The video focuses on creating and compositing two characters: a xenomorph and a soldier, with the aim of creating a dynamic and engaging scene.

💡Blur

Blur is a visual effect that softens or distorts the details of an image, often used to convey movement or to create a sense of depth by selectively focusing on certain elements. In the video, the speaker applies a blur to the background to simulate depth of field, drawing attention to the foreground characters.

💡Layering

Layering is a technique used in image editing where multiple images or elements are stacked on top of each other to create a composite image. This allows for greater control and flexibility in editing, as each layer can be manipulated independently. In the video, the speaker layers the xenomorph and soldier images on top of a jungle background to create a final composite scene.

Highlights

The trick of processing an image in separate parts and then compositing them back together in Photoshop for better composition control.

Setting up a blank project with specific dimensions and processing performance settings for optimal results.

Using a hyper-realistic xenomorph prompt with full body view and 8K photo setting for a more detailed and photographic image.

Randomizing the image output to have a choice between the best results.

Changing the prompt for the second character to a terrified soldier running towards the viewer to add emotional depth.

Utilizing Creal Pixel Cut for background removal, even with its free version, for the purpose of reinserting characters into the scene.

Stealing the color scheme and steaminess from the original xenomorph background to maintain visual consistency in the composite.

Using the Defocus feature to create a background from the original image, capturing the desired jungle atmosphere.

Scaling and positioning the xenomorph and soldier characters in Photoshop to create a dynamic composition.

Applying a soft blur to the background to achieve depth of field in the composite image.

Reusing the composited image as an input prompt to refine the final output with AI, enhancing the overall composition and depth.

Opting for the Quality setting over Speed in the AI processing for a higher quality result, despite the longer processing time.

The final composite image demonstrates a better result than relying solely on AI for layout and composition.

The method provides a practical application for improving image composition in a creative and controlled manner.

Engaging with the audience by encouraging likes, subscriptions, and comments for feedback and improvement.