Generative AI Made Easy: Video, Audio, and Image Creation

A valuable and playful and very necessary update - this lecture explores the top AI platforms for generating professional-quality video content. You will learn how tools like ElevenLabs, Runway, Firefly, and ComfyUI compare in terms of usability, flexibility, and output quality. The session focuses on real-world production considerations rather than surface-level features. By the end, you will understand which platform best fits your workflow and technical requirements.

Seedance 2.0 is the latest model from ByteDance who provide us TikTok.  The model updates the excellent Seedance Pro 1.5 which is still available at a very affordable price.  The newer models are fully multimodal and have multiple pathways to success.

Purpose of the lecture:  to establish the use of Seedance 2 for producing a futuristic musical video.

Key insights: prompting methods using text, audio, and images for a single outcome.  How well the model handles frequently censored elements and methods for working with censorship constraints.

Evaluation of outputs.

There are likely to be elements of any video that can be improved.  The processes differ with every category of improvement.  This video explores common issues that are encountered in Seedance 2.0 and relate these to the model architecture and solutions are suggested.

This lecture gives detailed instructions on text to video functionality which provides less precise influence, but which may be a timesaver when the peculiar censorship that affects Seedance 2.0 comes into effect.

The reference video to video function is used in this demonstration to achieve a guided second-take to improve a previous video. 

What we'll cover in this section

Purpose of lecture: Introduce ElevenLabs as an accessible, production-ready AI tool for images, video, and audio. You will see the full workflow from first login, to exploring community examples, to creating your own assets and judging whether the outputs genuinely match your intent.

The core technical focus is on how ElevenLabs handles prompts across different visual models, how much text the models realistically read, and how settings like resolution, aspect ratio, and duration affect both quality and cost.

What you will learn: You will generate a prompted image, tweak its background and style, then turn it into a talking-head style video with motion and audio, evaluating quality against your own description.

• Why it matters technically: This lecture clarifies model reading limits, resolution and duration settings, and how different ElevenLabs visual models behave under realistic constraints, including cost and performance.

  
  

In this lecture you will learn how ElevenLabs structures its pricing, account tiers, and credit system. We examine the free option alongside starter, creator, and higher-end plans, noting which features and project limits you get at each level. We'll share quality experience-based observations about quality differences between free and paid usage, how credits have changed over time, and when upgrading can be worthwhile.

By the end, you will be able to approach ElevenLabs subscriptions more tactically instead of guessing.

In this lecture I put everything we have seen so far into a structured evaluation of ElevenLabs. I define clear criteria for image, video, and audio quality, as well as for cost in credits, speed, and ease of use. We then apply these criteria to the clips and assets we have already generated, asking where ElevenLabs genuinely fits professional or semi-professional workflows. By the end, you will have a simple evaluation checklist you can reuse with ElevenLabs and any other AI platform you test.

• Purpose of lecture: I want to show you a practical, realistic way to extend ElevenLabs videos beyond a single short clip.

• Learner focus: We take a talking-head video and use Extend Video to add further dialogue, adjusting prompts, duration, and audio settings to keep the performance natural.

• Technical importance: You see how timing, model limits, audio toggles, and credit usage interact, and how glitches and random music can appear when clips get longer.

• Audience: This is for you if you plan to build multi-part scripts, sales pieces, or narrative segments and need to join several AI-generated shots together reliably.

Get started with Sora, from OpenAI - a generative AI platform that simplifies video creation. This lecture introduces Sora’s capabilities and how it integrates into creative workflows

Learn how to critically evaluate the quality of videos generated using Sora, focusing on visual fidelity, narrative coherence, and technical execution

Master complex video prompts in Sora AI to create more dynamic, detailed, and professional-looking videos tailored to your needs

Discover how to use image-based prompts in Sora to generate visually stunning videos that align with your creative vision

Learn how to incorporate motifs, MacGuffins, and themes into your generative AI projects to create compelling narratives that resonate with audiences.

• Purpose of lecture
  Introduce Suno as a fast, high‑quality AI engine for mood music and soundtrack creation.

• The take away
  You can generate multi‑minute songs from a single prompt, audition different model versions, refine prompts for style and emotion, and download usable audio for your projects.

• Why it matters technically
  Working with Suno’s newer models (v5, v5.5) shows how modern AI handles structure, arrangement, and timbre to produce convincing, non‑generic music suitable for real production workflows.

• Who it is for
  This lecture is ideal for creators, educators, and developers who need quick, customizable background music or songs for videos, courses, podcasts, and prototypes without traditional music production skills.

A brief explanation of the issues with Udio at present and why recommendation has shifted to using Suno.

Note this demonstration uses Udio - currently not the recommended service due to terms of service changes... Explore how musical motifs can enhance storytelling in generative AI projects by adding emotional depth and reinforcing themes.

Learn how to combine music and video generated by AI tools in Adobe Premiere Pro for seamless post-production workflows

Get hands-on with Adobe Firefly Video, Adobe’s generative AI tool designed for creating stunning visuals and videos from simple prompts. Part 1 covers the basics of Firefly’s features.

Dive deeper into Adobe Firefly Video’s advanced features, including fine-tuning visuals, integrating assets, and creating revised outputs.

This lecture provides a structured walkthrough of Nano Banana 2 (Gemini 3.1 Flash) within Adobe Firefly, focusing on real production workflows.

You will learn:

• How to configure resolution and aspect ratio for different use cases
• How to use reference images to guide composition and colour balance
• How to test and refine text rendering inside generated images
• How Google Search integration augments factual and architectural prompts
• How Nano Banana 2 compares with the Pro tier in terms of mood and detail
• Where Firefly’s interface limits differ from node-based systems such as ComfyUI

By the end of the lecture, you will understand how to strategically use Flash-tier generation for editorial layouts, architectural concepts, structured typography, and data-driven visuals.

Suno AI Explained: From Prompt to Complete Song in Minutes

Finding the right music for a video scene can take hours. Suno AI solves that problem by letting you generate a custom track built around the exact mood you need.

This lecture walks you through building a sentimental pop track from a simple description. You'll see how the AI handles verse structure, instrumentation choices, and vocal tone automatically.

You'll leave with a clear process you can repeat for any project. Whether it's a sad scene, an action sequence, or anything in between.

Not every scene needs words. Sometimes what your video really needs is a clean, atmospheric instrumental that sits perfectly in the background.

This lecture shows you how to build exactly that inside Suno AI. You'll go through the process of leaving lyrics blank, crafting a detailed style prompt, and using the magic button to push the result further.

You'll also get a practical tip for testing any AI-generated track before using it commercially. A quick YouTube upload test that could save you a serious headache down the line.

By the end of this lecture, you'll know exactly how to get the most out of Epidemic Sound — one of the OG royalty-free audio brands available.

You'll discover how to browse music and sound effects, understand the practical difference between diegetic and non-diegetic sound, and learn how the platform's AI tools can generate custom-length versions of any track.

You'll also get a clear breakdown of the subscription plans, so you can choose the right one for your needs.

The AI features are coming fast.

This information changes fast, so always read up on the t&c for platforms before using them.

Content Aware Fill allows the use of AI to fix minor problems and to remove artifacts from AI generated videos