Scott Kelby’s Top AI Secrets for Lightroom & Photoshop | Bild Expo 2025
Introduction to AI in Photography
Initial Engagement
- The speaker opens with a humorous question about attendees' motivations for attending the session, hinting at the expectation of free food.
- Expresses personal concerns about AI's impact on photography but emphasizes finding productive uses for it.
Purpose of the Session
- The speaker aims to share insights on using AI tools effectively within their workflow, highlighting boundaries between useful and unnecessary applications.
- Mentions that many people are unaware that Photoshop has been utilizing AI features for years, which they often consider as standard production tools.
Understanding Generative Fill and Expand
Key Features in Photoshop
- Introduces two major areas of focus: generative fill and generative expand, noting their significance in modern photo editing.
- Shares a personal anecdote about dealing with reflections in glasses during portrait photography, illustrating common challenges faced by photographers.
Practical Application of Generative Fill
- Describes how generative fill simplifies the process of removing reflections from glasses compared to traditional methods.
- Explains the ease of using generative fill: making a loose selection around an unwanted reflection and generating a fix with just two clicks.
Using Generative Fill Effectively
Step-by-Step Process
- Outlines the straightforward steps involved in using generative fill: select an area, click generate, and wait approximately 12 seconds for processing.
- Highlights that users receive three variations after generation, allowing them to choose based on subtle differences like expressions.
Tips for Optimal Use
- Advises leaving input fields blank when wanting to remove elements from images; this is a common practice since early versions of Photoshop.
- Emphasizes that most selections can be made without additional input—just select and hit generate for quick results.
Understanding Generative Fill in Photoshop
Exploring Variations and Options
- The generative fill feature provides three different variations for users to choose from, with the first option rarely being the best. Users can generate additional sets until they find a satisfactory result.
- Each generation creates a new set of elements, including reflections, which may vary in quality across options. It's common for the second or third variations to be more appealing than the first.
When to Use Prompts
- Prompts are useful when wanting to add or change elements in a photo, such as replacing clothing items or adding objects that weren't originally present.
- The speaker expresses discomfort about altering images by adding elements like planes or lighthouses, feeling it compromises authenticity.
Personal Reflections on Image Alteration
- The speaker shares their internal conflict regarding image manipulation; despite having the tools available (like Photoshop), they prefer capturing moments authentically without artificial enhancements.
- They recount an experience of adding a sky to a photo taken in Maine, which led to feelings of guilt over not representing reality accurately.
Reasons for Minimal Prompt Usage
- The speaker tends to leave prompts blank because typing something often leads to significant changes that could disrupt their peace of mind regarding image integrity.
- Their primary use of generative fill is for efficiency—removing distractions quickly compared to traditional methods like content-aware fill or healing brushes.
Demonstrating Generative Fill's Effectiveness
- An example illustrates how generative fill outperforms older techniques by effectively removing distracting elements (e.g., garbage cans).
- A comparison between content-aware fill and healing brush shows limitations in traditional methods versus the flexibility and effectiveness of generative fill.
Enhancing Results with Resolution Adjustments
- After using generative fill, users can enhance results further by increasing resolution if necessary; however, there are size limitations that may prevent enhancement if fixes are too small.
- Users should be aware that clicking enhance might yield warnings if adjustments fall below required dimensions but can significantly improve larger edits.
Generative Fill Techniques in Photography
Overview of Generative Fill
- The speaker discusses the efficiency of using generative fill to enhance images, noting that it can save significant time compared to manual editing.
- Three variations are generated for a selected image, highlighting the anticipation and excitement during the waiting period for results.
Challenges with Object Removal
- The speaker addresses common issues when attempting to remove unwanted objects from images, such as inadvertently adding similar items instead.
- A personal anecdote illustrates this problem: trying to remove a squirrel resulted in a larger one appearing instead.
Solutions for Effective Object Removal
- To improve object removal outcomes, users should type "remove" in the tool's input field to clarify their intent.
- Making a larger selection area is crucial; touching edges of unwanted objects can lead to incorrect replacements.
Practical Example of Selection Techniques
- An example involving a church demonstrates how improper selections can result in undesired edits. Ensuring clean selections is emphasized.
- After adjusting selections properly, successful removal of unwanted elements is achieved by generating fill again.
Advanced Tools for Distraction Removal
- The speaker introduces an additional tool called "find distractions," which helps identify and eliminate specific elements like wires or cables from photos.
- A travel photographer's struggle with tourists interrupting shots leads into practical application: selecting and removing tourists from an image while preserving the main subjects' emotions.
Final Thoughts on Image Enhancement
- Despite imperfect selections, generative fill can significantly enhance images by focusing on desired subjects and improving overall composition.
- The discussion concludes with reflections on how these tools transform photography practices, particularly in portraiture and travel photography contexts.
Smoothing Skin Techniques in Photo Editing
Introduction to Smoothing Skin
- The speaker discusses the effectiveness of smoothing skin in photo editing, emphasizing that it won't drastically change an older person's appearance but can improve gradations in skin tones.
- The technique is particularly useful for addressing hot spots or uneven gradations in lighting on a subject's face.
Demonstration of Techniques
- A demonstration shows how quickly hot spots can be removed using generative fill, taking only seconds compared to traditional methods which may take 25 minutes.
- The results from generative fill are noted to be superior to previous techniques like frequency separation, providing better skin texture and detail retention.
Enhancing Reflections
- The speaker explains a method for creating reflections in Photoshop without generative fill, highlighting the simplicity of flipping layers vertically.
- Challenges arise when dealing with angled subjects; however, generative fill simplifies this process by producing realistic reflections effortlessly.
Generative Fill Options
- Users can type specific requests (e.g., "reflection") into the tool to generate various options for reflections, allowing for easy selection of preferred outcomes.
- If initial results are unsatisfactory, users can regenerate options without needing to reselect areas, streamlining the editing process.
Fixing Compositional Mistakes
- The speaker emphasizes the utility of generative expand for correcting compositional errors made during shooting, such as cropping too closely.
- By selecting areas and activating generative expand through the crop tool, users can adjust compositions effectively without extensive manual adjustments.
Compositional Mistakes in Photography
Fixing Cropped Images
- The speaker discusses a compositional mistake where the bride's arms are cropped awkwardly, emphasizing the importance of proper framing.
- By adjusting the crop, the speaker demonstrates how to create more space above the subject's head, showcasing options for improvement.
- The speaker reflects on personal mistakes made during photography and highlights tools available for correcting these errors.
Generative Expand Tool
- The generative expand feature is introduced as a solution for fixing cropped images, allowing users to generate additional content around subjects.
- Multiple generated options are presented, with a preference for specific versions that enhance composition.
Restoring Old Photos
Utilizing AI Tools
- The speaker transitions to restoring old photographs using similar techniques discussed earlier, focusing on enhancing damaged images from past decades.
- An example of an old family photo is used to illustrate how cropping can be adjusted to include missing elements like hands or feet.
Generating New Content
- The process of generating new content around existing images is highlighted as a way to improve visual quality and restore lost details.
- Users can repeatedly generate variations until they find satisfactory results, demonstrating flexibility in digital restoration.
Exploring Neural Filters
Introduction to Neural Filters
- A shift towards discussing neural filters in Photoshop introduces AI-based features that enhance image editing capabilities.
Colorization Feature
- The colorize filter is showcased as an innovative tool that automatically adds color to black-and-white photos, illustrating its effectiveness through personal examples.
Sky Replacement Techniques
Hesitance Towards Sky Replacement
- The speaker shares personal hesitations about using sky replacement technology due to preconceived notions about its authenticity and effectiveness.
Real-Life Application
- A story involving a conversation with a respected landscape photographer reveals insights into perceptions surrounding sky replacement and its impact on photography.
Sky Replacement in Photography
The Reality of Sky Replacement
- A photographer shares a conversation where he learns that many professionals replace skies in their photos, which challenges his perception of authenticity in photography.
- He reflects on the effort involved in capturing the perfect shot, only to realize that if conditions aren't ideal (like no clouds), those images are often discarded.
- After speaking with other photographers, he acknowledges that sky replacement is a common practice and admits to adopting this technique himself for better results.
Techniques and Tools
- The speaker recounts a trip to Arizona where he struggled with clear skies; he began using past images from Arizona or even Hawaii to enhance his shots.
- He introduces Photoshop's built-in sky replacement feature, emphasizing its simplicity and effectiveness for enhancing landscape photographs.
Understanding Photoshop Features
- The speaker notes that "sky replacement" is one of the few features in Photoshop aptly named for its function, contrasting it with other tools that have less intuitive names.
- He warns against using sample images provided by Photoshop on social media due to potential backlash from viewers who recognize them as stock images.
Transitioning to Lightroom AI
- The discussion shifts towards Lightroom's AI capabilities, noting they are more limited compared to Photoshop but still useful for photo editing.
- He introduces "adaptive profiles," a new feature designed to improve automatic adjustments based on image content like the sky.
Evaluating Adaptive Profiles
- The speaker explains how adaptive profiles analyze each photo individually and make adjustments accordingly, improving upon previous auto-adjustment features.
- By comparing traditional auto settings with adaptive profiles, he demonstrates how adaptive profiles provide a better starting point without altering sliders unnecessarily.
AI-Powered Photo Editing Features in Lightroom
RAW Photo Limitations
- Currently, the AI features discussed are only applicable to RAW photos; JPEG and TIFF support may come in the future.
Generative AI Removal Tool
- The removal tool utilizes generative AI technology from Photoshop and Firefly, allowing users to paint over unwanted elements for removal.
- Users must ensure complete coverage of the area to be removed before clicking 'remove' and can adjust their selections afterward if needed.
Image Generation Options
- After using the removal tool, users can generate three different options for how the image looks post-editing.
- If unsatisfied with any option, users can regenerate additional variations similar to Photoshop's functionality.
Enhanced People Masking Tools
- New masking tools allow for more precise selection of subjects compared to previous versions, enhancing user control over editing.
- Users can select specific facial features or body parts (e.g., skin, eyes, hair), enabling targeted adjustments through sliders after creating a mask.
AI-Based Noise Reduction
- A significant advancement is the introduction of AI-based noise reduction that effectively manages high ISO images without losing detail.
- Even at 8,000 ISO, noise is significantly reduced while maintaining image clarity; however, estimated processing times may vary greatly from actual times.
Super Resolution Feature
- The super resolution feature doubles image resolution using AI technology and is particularly useful for large prints while enhancing sharpness beyond original quality.
AI Features in Lightroom and Photoshop
Enhancing Image Size with AI
- The speaker demonstrates how to enhance an image size of 6,000 by 4,000 pixels using the "Super Resolution" feature in Lightroom.
- After processing, the image size successfully doubles to 12,000 by 8,000 pixels within a few seconds.
AI Tools in Photography
- The speaker emphasizes that both Lightroom and Photoshop's Camera Raw include AI-based features for photographers.
- Adobe has developed "content credentials," which embed metadata into photos detailing any modifications made, promoting transparency in photo editing.
Transparency and Ethical Use of AI
- There is a discussion on differing opinions regarding transparency; some prefer privacy while others advocate for openness about edits made using AI tools.
- The speaker encourages responsible use of AI as a production tool to enhance workflow without compromising integrity or feeling deceptive.
Time Management for Photographers
- A quote from Richard Abdon highlights the philosophical aspect of photography: "All photos are accurate. None of them is the truth," suggesting that editing does not negate authenticity.
Audience Interaction and Questions
- Following the presentation, audience members are invited to ask questions about the discussed topics.
Question on Generative Features
- An audience member asks if it's possible to revert back to previous generations after regenerating images in Photoshop.
- The speaker explains that unless saved as a layered PSD file, previous generations cannot be retrieved once closed.
Inquiry About AI Limitations
- Another question arises regarding issues with fingers in generated images.
- The speaker clarifies that Adobe's generative model was trained on fewer but more curated images compared to other platforms like Midjourney, leading to better results with hands and fingers.
Discussion on AI in Image Editing
Ethical Considerations in AI Development
- The speaker discusses how some companies, unlike others, have taken an ethical approach to AI development, particularly in image editing. They mention that while many companies produce visually appealing results, they often struggle with specific details like fingers.
- Adobe's strategy involved using a limited dataset of images from individuals who signed agreements for their work to be used, contrasting with other companies that scraped vast amounts of data without consent.
Challenges with Image Generation
- The speaker notes ongoing issues with generating realistic images of fingers and mentions the recent beta release of Photoshop which promises improvements but has not yet been tested by them.
- A question arises about noise reduction techniques when working between Lightroom and Photoshop. The speaker advises performing noise reduction before exporting to PSD format.
Tools and Techniques for Noise Reduction
- The speaker recommends using a plugin called On1 Noise AI for efficient noise reduction, highlighting its speed and effectiveness compared to Lightroom’s built-in tools.
- They clarify that this recommendation is based on personal use rather than sponsorship influence, emphasizing the tool's performance.
Color Profiles and Output Settings
- A question about color profiles leads to clarification that Adobe RGB and sRGB settings are relevant only during file export processes, not during initial editing or selection stages.
- The speaker explains that standard profiles remain available but are located differently within the software interface.
Addressing Distortion Issues in Photography
- An inquiry about photographing subjects with thick glasses reveals the speaker's lack of experience with generative fill addressing distortion caused by lenses.
Future Improvements in AI Tools
- A participant raises concerns regarding AI tools' effectiveness on stylized artwork versus realistic images. The speaker suggests visiting Adobe’s booth for insights into potential future developments.
- They encourage direct engagement with product managers at Adobe for more information on upcoming features or improvements related to stylized image generation.
Selection Tools in New Software Updates
- Discussion shifts towards new features in Photoshop's beta version where selection tools have been enhanced. Users can now access improved masking options directly from the top menu instead of navigating through multiple steps.
Photoshop Innovations and Modern Compositions
Advancements in Photoshop Features
- The latest version of Photoshop introduces advanced masking capabilities that allow for more detailed editing, including the ability to manipulate clothing items separately from the background.
- A discussion arises about traditional composition techniques, emphasizing how older methods focused on composing images entirely within the viewfinder.
Evolving Composition Techniques
- The speaker suggests that photographers should adapt their approach by leaving space around the center of an image to facilitate post-processing in Photoshop, challenging old rules of composition.
- With advancements in camera technology (e.g., 50-megapixel cameras), viewing images at 100% is deemed unrealistic; instead, working at smaller sizes (like 25%) is recommended for practical editing.
Impact of Social Media on Image Presentation
- The rise of platforms like Instagram has changed how images are perceived and edited; it allows for more leniency in detail when viewed at smaller sizes, making it easier for less-than-perfect images to still look appealing.