Você está pronto para transformar sua forma de criar?

Welcome and Introduction

Overview of BR Software and Its Focus

• Dalbelo welcomes attendees and introduces BR Software's digital channels, emphasizing their presence on Instagram and LinkedIn for updates on events, including those related to Adobe.

• He invites first-time viewers to comment about their experience with BR Software's live sessions, highlighting the company's nearly 40 years in technology and innovation.

• BR Software is recognized as a key partner among major players like Adobe, Microsoft, Autodesk, and others in various tech sectors including design and security.

Adobe Max Event Highlights

Key Announcements from Adobe Max

• The focus of today's discussion is on recent developments presented at Adobe Max held in Los Angeles, where significant updates were announced for Adobe applications.

• Notable version changes include Photoshop moving from version 25 to 26 and Illustrator from version 29 to 30; these changes are not year-related but rather version upgrades.

Introduction to Firefly

Exploring New Features

• Dalbelo outlines the session's aim: to summarize key features of Photoshop, Illustrator, Camera H, with a primary focus on Adobe Firefly.

• He shares his background as a consultant at BR Software for nearly six years with expertise in creative aspects of Adobe’s Creative Cloud.

Interactive Session Setup

Engagement with Audience

• Dalbelo mentions that today’s presentation will be shorter than usual but will include an interactive Q&A segment via Teams.

• A webpage created using Adobe Express will serve as a resource for sharing information about Firefly; it will be updated regularly.

Understanding Adobe Firefly

Generative AI Capabilities

• The page created serves as a living document containing essential information about Firefly which includes documentation and important links.

• Firefly is described as Adobe's generative model launched in 2023 that allows users to create images, videos, vectors, audio content among other formats.

Historical Context of AI at Adobe

Long-standing Commitment to AI Development

• Dalbelo emphasizes that Adobe has been investing in artificial intelligence (AI), particularly through its technology called Sensei since around two decades ago.

• He clarifies that Sensei is not an application but rather the underlying technology powering various tools within the Adobe ecosystem.

Content-Aware Fill Feature

Practical Application of AI Technology

• An example provided is the Content-Aware Fill feature in Photoshop which utilizes Sensei technology by analyzing image areas for content recreation.

Understanding Adobe's Approach to AI and Creativity

Overview of Adobe's AI Tools

• The speaker discusses the limited reference area they have in photography, highlighting the revolutionary impact of Continental Aware since its launch in 2010. They emphasize its continued usefulness alongside generative tools.

• The speaker mentions that there are numerous resources available, suggesting that if viewers want more details, they can request a dedicated video explaining these tools.

Adobe's Vision for Artificial Intelligence

• Adobe has been working with artificial intelligence (AI) for over 20 years to enhance their software tools. The goal is to optimize repetitive processes, reducing tasks from three clicks to one.

• The company views AI as a means to foster creativity rather than replace it. This perspective positions creatives as primary users of Adobe products, aiming to streamline workflows and save time.

Introduction of Adobe Firefly

• The speaker introduces Adobe Firefly and notes an update in terms of use upon its launch. Users must authorize Adobe to analyze their images for model training purposes.

• It is clarified that while users' content is analyzed by the AI, it does not train on this content directly. Instead, it synthesizes information from user-uploaded images without retaining them for future learning.

Data Sources and Ethical Considerations

• A significant advantage of Firefly is its authorized learning database sourced from Adobe Stock and public domain images. Photographers must consent for their work to be used as references in training the AI.

• Photographers who contribute images receive compensation when their work is utilized by Firefly, creating a sustainable cycle where new content continuously enriches the database.

Intellectual Property Protection

• Users generating images through Firefly are protected under intellectual property indemnification policies. If a dispute arises regarding image ownership, users can claim up to $10,000 if they can prove proper usage credentials.

• Not all versions of Creative Cloud offer this indemnity; only specific plans like Pro Plus or Enterprise editions provide such protections against potential copyright claims related to generated content.

Licensing Choices and Adobe Updates

Importance of Choosing the Right Licensing

• It's crucial to pay attention when selecting your licensing options. Additional resources and links are provided for further information.

• Regardless of whether you have 1, 10, or 50 licenses, you can choose from four available versions.

Changes in Adobe's Nomenclature

• Adobe has recently updated its product names; the website now offers a version focused solely on online services without software.

• The previous Creative Cloud is now referred to as Creative Cloud Pro, while Creative Cloud Pro has been renamed to Creative Cloud Pro Plus for Teams.

• For Enterprise users, the former Creative Cloud Enterprise is now called Edition 3, and Enterprise Pro is Edition 4.

New Features and Benefits

• The new Pro Plus and Edition 4 offer unlimited downloads from Adobe Stock, providing excellent value along with intellectual property indemnification and additional Firefly features.

Understanding Generative Credits

What Are Generative Credits?

• Generative credits control the consumption of resources within Firefly. They help manage energy usage since generating content requires varying amounts of energy.

Energy Consumption Insights

• Different types of content consume different levels of energy; for example, generating images versus videos incurs different credit costs based on resolution and duration.

Credit Allocation by Licensing Type

• Each licensing type comes with a specific number of generative credits:

• Edition 3 includes only 25 credits (limited).

• Pro Plus provides 4,000 credits with unlimited standard resource capacity.

Resource Types: Standard vs Premium

Differentiating Resource Categories

• Resources are categorized into standard (unlimited for some licenses like Pro Plus Edition 4) and premium (with monthly user limits).

Purchasing Additional Credits

• Users needing more than their allocated credits can purchase additional Firefly credits in increments ranging from 2,000 to 50,000 per user.

Content Generation Limitations

Size and Quality Constraints

• Content generation limitations regarding size or quality depend on the model used rather than directly on credit availability. Different models yield different resolutions.

Third-Party Models in Adobe Software

Introduction to Third-Party Models

• Discussion on the integration of third-party models within Adobe applications like Photoshop, Illustrator, and Premiere.

• Mention of significant changes at Adobe, including generative tools and conversational interfaces.

AI Partnerships and Model Specialization

• Adobe's focus on conversational AI for editing and template creation.

• Collaboration with major companies such as Google, Runway, and OpenAI to enhance software capabilities.

• Explanation that different AI models excel in various areas (e.g., textures, lighting), allowing users to select the best tool for their needs.

Benefits of Integrated AI Tools

• Users can access multiple specialized models without needing separate subscriptions for each tool.

• Demonstration of how specific tasks can be optimized by choosing the right model (e.g., Firefly for certain elements).

Available Models and Features

• Overview of available models for images, videos, audio; mention of high-resolution video generation capabilities with Luma.

• Introduction to custom models in beta testing for Enterprise clients that allow training on proprietary data.

Testing Opportunities and Future Developments

• Announcement about unlimited testing opportunities until December 1st for various features using Firefly and partner tools.

• Details on what is available during this testing phase: unlimited image generation with Firefly partners and video generation using Firefly without credit consumption.

Creative Production Tool Overview

• Introduction to the Creative Production feature within Firefly designed for bulk actions (e.g., background removal).

• Description of existing tools within Creative Production aimed at enhancing workflow efficiency.

Introduction to Presets and Background Removal

Overview of Presets

• Presets allow users to apply treatment settings such as light, shadow, and color to images.

• Users can apply a preset to up to 1000 images at once, with a maximum file size of 50 MB per image.

Background Removal Tool

• The background removal tool consumes 20 credits per use; multiple tests were conducted without excessive credit consumption.

• The tool effectively removes backgrounds from various image types, demonstrating high accuracy in complex areas like hair.

Firefly's Image Model Five Features

Introduction to Image Model Five

• Firefly's Image Model Five is currently in beta and allows for both creation from scratch and editing existing photos.

Practical Applications

• Users can generate images with improved resolution (up to 4 MB), enhancing the quality of outputs.

• The model enables specific edits like changing clothing colors or adjusting poses, showcasing its versatility.

Testing the Capabilities of Firefly

Experimenting with Edits

• Various tests included removing unwanted objects (like post-it notes), which was faster than traditional selection tools.

• Some edits did not meet expectations, such as altering facial expressions or features that resulted in unsatisfactory outcomes.

Exploring Firefly Boards

Concept of Firefly Boards

• Firefly Boards serve as a creative workspace where users compile references and elements for projects.

Integration with AI

• The boards facilitate brainstorming by allowing users to combine traditional reference gathering with AI capabilities for enhanced creativity.

• New features include virtual try-ons for clothing items using selected models within the boards.

Exploring New Fashion Technology

Introduction to the Fashion Application

• The speaker selects a clothing item for application and mentions that this feature is available until December 1st, encouraging users to test it extensively due to its potential in fashion product ideation.

• The process involves cloud-based composition, which may take some time; however, the live demonstration shows relatively quick results.

Demonstrating Clothing Modifications

• Various modifications are showcased, including changing a model's outfit (e.g., jacket, shoes), illustrating how the technology can alter images while maintaining the model's identity.

• The speaker discusses using presets to merge multiple images together, highlighting the ability to combine elements like shirts and accessories into new designs.

Advanced Features of Image Generation

• Users can change poses and angles of models by selecting reference photos and describing desired actions (e.g., "walking left"), showcasing flexibility in image manipulation.

• The tool allows for generating variations of existing images by describing them; this speeds up creative processes significantly.

Utilizing AI for Creative Production

• A feature enables users to describe an image for prompt generation, facilitating quicker adjustments without needing extensive manual input on descriptions.

• Users can modify exposure settings (wide or portrait), enhancing control over final outputs based on their preferences.

Overview of Additional Tools and Resources

• The speaker emphasizes that there are numerous features within the platform worth exploring before the deadline, suggesting a deep dive into tools like Creative Production and Model 5.

Sneak Peeks at Upcoming Features

Insights from Adobe Max Event

• Sneak peeks at Adobe Max showcase upcoming features developed by engineers; audience reactions may influence whether these become official tools.

Turntable Feature in Illustrator Beta

• Introduction of a new "Turntable" feature shown during sneak peeks is currently available in beta form within Illustrator. Users are encouraged to install beta versions for testing purposes.

Illustrator's New AI Features

Generating Vectors with AI

• The official Illustrator 30 includes a feature to generate vector graphics using AI, allowing users to create shapes and request specific designs, such as "an astronaut dog."

• Users can select different models for generation, including Firefly, GPT, Ideogram, and Jamini. This flexibility enhances creative possibilities.

• A reference image can be uploaded (JPEG or any format), which serves as inspiration for the generated design. Each generation consumes a credit.

Content Credentials and Image Validation

• The content created with AI is tagged with "content credentials," which helps track alterations made to images or videos.

• Users can validate if an image has been modified by checking its content credentials on the portal; it also specifies the model used for generation.

Live Demonstration of Vector Generation

• During a live demonstration, the speaker attempts to generate an astronaut dog vector again using different models due to previous failures in generating desired results.

• The speaker notes that while some models may not support Portuguese yet, they are working towards improving accessibility.

Introduction of Turn Table Tool

• A new tool called "turn table" is introduced for better visualization of selected vectors. It processes actions based on user selections in real-time.

• If the contextual taskbar does not appear automatically, users are advised on how to access it through window settings.

Software Preferences and Language Use

• The speaker shares their preference for using software in English due to more comprehensive resources available in that language compared to Portuguese.

• They encourage viewers who have questions about translations or specific tools in Portuguese to reach out for assistance.

Exploring Adobe Bridge Features

• While waiting for processing times during demonstrations, the speaker discusses Adobe Bridge as a useful file navigator that provides previews of files like PSD documents that typically lack them in standard file explorers.

Overview of Adobe Bridge and Photoshop Features

Introduction to Adobe Bridge

• Adobe Bridge provides detailed information about files, including photography settings like aperture, ISO, shutter speed, and dimensions. It offers various resources and filters based on file types.

• The speaker highlights that Bridge is an excellent browser for users with Creative Cloud subscriptions, whether they use all applications or a single app.

Demonstration of Photoshop Features

• The speaker shares their screen to demonstrate features in Photoshop, specifically focusing on generative fill. They plan to open an image for illustration.

• After some technical difficulties, the speaker successfully opens the software and begins discussing the "turntable" feature which generates versions of a vector from different angles.

Turntable Feature Insights

• The turntable feature allows users to create multiple views of a complex vector design. For example, it can rotate images like a dog character from various perspectives.

• A simpler illustration (a lion character) demonstrates how this feature works effectively even in its beta version by generating new angles quickly after initial rendering.

Generative Fill Comparison

• Transitioning back to Photoshop, the speaker discusses using Firefly 1 for texture generation in images. They compare it with older tools like Content-Aware Fill from 2010.

• The speaker explains how generative fill can replace unwanted elements in photos more effectively than traditional methods by utilizing larger reference bases.

Limitations and Advantages of Tools

• When using Content-Aware Fill, issues arise due to limited reference data leading to visible defects in generated areas. This method struggles with complex backgrounds.

• In contrast, generative fill understands lighting conditions better and produces more seamless results without harsh lines or inconsistencies.

Practical Application of Generative Fill

• The speaker showcases how generative fill can recreate natural elements like grass while maintaining proper lighting context within an image.

• Users can request specific additions (e.g., sand next to grass), allowing Firefly's AI capabilities to search through stock images for accurate blending techniques.

Conclusion on Model Performance

• Different models yield varying results; while Model 1 performed best for certain tasks, others did not meet expectations. This variability emphasizes the importance of selecting appropriate tools based on project needs.

• The discussion concludes with insights into resolution capabilities within Firefly models as they relate to vector graphics and overall image quality considerations.

Image Resolution and AI Techniques

Understanding Image Resolution Challenges

• The image in question has a width of 7,000 pixels. Attempting to fill the entire area at once can lead to resolution loss.

• When expanding an image with a maximum resolution of 2000 pixels, the quality diminishes significantly.

• A recommended approach is to make selections of up to 2000 pixels at a time to maintain better resolution.

Practical Applications of AI in Image Editing

• Using AI tools like reling brush can quickly fix minor issues, such as filling gaps in images while preserving depth and detail.

• The second attempt at filling an area resulted in improved resolution compared to the first attempt, which had significant quality loss.

Third-party Models and Their Limitations

• The speaker discusses using third-party models like Flux for tasks such as removing reflections from glasses, noting that Firefly lacks sufficient data for this task.

• Caution is advised against selecting entire images due to potential resolution loss; smaller selections yield better results.

Case Studies on Reflection Removal

• Initial attempts with larger prompts led to undesirable outcomes where the glasses were entirely removed instead of just the reflections.

• A more focused selection around the glasses produced satisfactory results without altering other facial features.

Subscription Levels and Access

• Availability of third-party models depends on subscription levels; most users have access except those on Enterprise edition 3.

• Creative Cloud Pro allows testing until December 1st; after that, premium credits will be required for continued use.

Testing Results with Different Tools

• Various tests were conducted using different tools (e.g., Google), highlighting inconsistencies in maintaining original features while removing reflections.

• Adjustments made during testing showed improvements but still faced challenges with retaining natural appearances post-editing.

This structured summary captures key insights from the transcript regarding image editing techniques using AI tools while addressing challenges related to image resolution and practical applications.

Adobe Tools and Techniques for Image Editing

Overview of Image Selection and Editing Tools

• The speaker discusses a slight distortion in the mouth's shape but notes that the teeth's characteristics remain intact. They mention improvements in selection processes with the use of "flux."

• A rectangular selection was made to remove an appliance from an image, resulting in a satisfactory outcome. The effectiveness of different models (Firefly 1 vs. Firefly 3) is highlighted, emphasizing the importance of testing until December 1st.

• Users are encouraged to provide feedback on Adobe tools directly through properties options, which can help improve functionality. Feedback can be given in Portuguese.

Object Selection Tool Features

• Introduction of various selection tools including a relatively new object selection tool that identifies people within images, allowing users to select specific parts like clothing or body features.

• The speaker demonstrates selecting a person entirely for editing purposes, showcasing how detailed selections can enhance image treatment.

Generative Fill and Model Performance

• After making selections, users can utilize generative fill features to change clothing or other aspects within an image. The speaker mentions varying results based on model performance.

• While generating images live, the speaker notes inconsistencies in results; sometimes it works well while other times it does not. Users should pay attention to prompts during generation.

Background Removal and Efficiency Tools

• Discussion about Adobe’s background removal tool which simplifies the process by automatically selecting subjects and applying masks without multiple steps.

• This tool enhances workflow efficiency by reducing clicks and streamlining actions necessary for background changes in images.

Contextual Bar Features and Harmonize Tool

• The contextual bar provides options for generating or importing backgrounds seamlessly while understanding previous layers' context.

• Introduction of the Harmonize feature now officially available in Photoshop, aimed at improving integration between elements within edited images.

This structured summary captures key insights from the transcript regarding Adobe's image editing tools and techniques while providing timestamps for easy reference back to specific points discussed.

Harmonizing Images with Photoshop

Introduction to Harmonize Tool

• The speaker introduces the Harmonize tool, explaining its purpose of harmonizing images by adjusting lighting and color grading.

• The tool analyzes the background image's lighting to apply corrections to the foreground image, although it may not work perfectly for all images.

Functionality and Limitations

• While the Harmonize tool can save time in most cases, users should be prepared to make manual adjustments as needed.

• It is emphasized that user-generated content is not used for training AI models, allowing creators to use their images without concerns about copyright issues.

Demonstration of Image Adjustments

• A demonstration shows an attempt to harmonize a chair image with a background; however, technical difficulties arise during the live session.

• The speaker shifts focus to previously prepared images due to ongoing issues with the current one.

Example of Successful Application

• An example is provided where a dog on the beach photo is adjusted using Harmonize. Three options are presented for lighting adjustments.

• The second option is favored as it effectively applies shadows consistent with the scene's lighting, enhancing realism.

Advanced Features: Upscaling Images

• The speaker transitions to discussing upscaling features available through a partnership with Topaz plugins integrated into Photoshop.

• A personal anecdote illustrates challenges faced when working with low-resolution images and how alternative solutions were sought in Adobe Stock.

Generative Upscale Tool Overview

• Users are introduced to the Generative Upscale feature that allows resizing images without distortion, offering three different models for enhancement.

• Limitations are noted regarding maximum image sizes for effective upscaling; examples demonstrate differences between traditional resizing methods and generative techniques.

Image Enhancement Techniques and Tools

Overview of Image Enhancement Tools

• The speaker discusses the capabilities of Bloom, a tool that creatively redesigns images but is not recommended for faces due to its tendency to alter them significantly.

• An example is given of an old photo from 2009 with low resolution, illustrating how enlarging it without enhancement leads to poor quality and artifacts.

• Various tools are compared: Firefly Up Scaler maintains some contrast but isn't ideal; Gigapixel improves resolution slightly; Bloom's high creativity setting distorts the image too much.

Neural Filters and Restoration Techniques

• Before upscaling, the speaker emphasizes using neural filters for restoration, which can remove JPEG artifacts and enhance photo quality.

• A demonstration shows that applying neural filters before upscaling results in better outcomes when using tools like Firefly Upscaler and Topaz Gigapixel.

AI Sharpening Features

• The speaker highlights the use of AI sharpening as a new feature in filters that enhances image clarity efficiently without manual adjustments.

• This feature saves time by automating sharpness adjustments, showcasing significant improvements in previously blurred images.

Training Opportunities for Adobe Users

• Information about Adobe's training programs for corporate clients is shared, emphasizing free access to various software courses through corporate emails only.

• These trainings provide valuable skills in different Adobe applications and offer certificates upon completion, enhancing professional development.

Conclusion and Call to Action

• The speaker encourages participants to test out new features available until December 1st, highlighting their potential benefits for daily work tasks.

• Suggestions are welcomed for future live sessions focusing on specific topics or tools within Adobe products.

What Would You Like to Learn More About?

Engagement and Learning Opportunities

• The speaker expresses interest in understanding the audience's learning preferences, indicating a desire for feedback on topics of interest.

• A suggestion is made for viewers to follow both the speaker's LinkedIn and that of Abra Software to stay updated on upcoming events and live sessions.

• The speaker thanks the audience for their questions, encouraging further inquiries even from those watching the recorded version of the session.