Basted In Blood Tutorial
In a couple of hours, I covered this 1997 SNL Sketch where “Cinder Calhoun” introduces her special guest, Sarah McLachlan, to perform a "Holiday Anthem" protesting the "Turkey Holocaust" of Thanksgiving.
This is my cover of that song and music video and then the tutorial of how I did it below.
Feel free to download it and introduce your friends and family with a rare Thanksgiving holiday song :)
How I Created a Thanksgiving Turkey Music Video Using AI Tools
A complete walkthrough of transforming an SNL sketch into a fully animated AI-generated music video in just a few hours.
The Inspiration: Finding the Perfect Source Material
The creative process began with rediscovering a 1997 SNL Weekend Update sketch featuring Ana Gasteyer as Cinder Calhoun alongside Sarah McLachlin. This comedic gem from November 22, 1997, provided the perfect foundation for a Thanksgiving-themed parody project. The original sketch's humor and melody sparked the idea to create a modern AI-powered reimagining.
Creating the Music in Suno
Uploading the Audio Reference
The first step involved getting the original SNL sketch recording and uploading it to Suno as a reference track. Rather than starting from scratch with lyrics, Suno's cover feature allowed the audio file to serve as the foundation for a new version.
Defining the Musical Style
After uploading the audio, Suno automatically detected and transcribed the lyrics. The key to achieving the desired sound was specifying the style parameters: "acoustic guitar, two-part harmony, female vocalists." This combination created the warm, folk-inspired tone needed for the Thanksgiving theme.
Selecting the Best Version
Suno generates two versions of each creation, allowing for comparison and selection. After reviewing multiple iterations with slight variations in the style instructions, the version with the most compelling vocal harmonies and instrumentation was chosen to move forward.
Professional Audio Mastering with BandLab
With the basic track complete, the next challenge was achieving professional audio quality. BandLab proved invaluable for this stage of production.
Splitting Stems and Adjusting Mix
BandLab's stem separation feature allowed for individual control over vocals, guitars, and other instrumental elements. While the platform offers extensive remixing capabilities—including adding new instruments and making detailed modifications—the primary goal for this project was focused mastering.
Applying Professional Mastering
The mastering process in BandLab took just one minute to apply but made a significant difference in the final sound quality. The result was a punchier, more polished audio track with better balance and clarity across all frequencies.
Planning the Visual Concept with ChatGPT
The vision for the music video was ambitious: two attractive female turkeys playing acoustic guitars and singing. ChatGPT became the strategic planning partner for bringing this unusual concept to life.
Initial Prompting for Video Strategy
The first conversation with ChatGPT outlined the goal and asked for the best approach to creating such a video. The AI provided several options, including Midjourney prompts for character creation and recommendations for different lip-sync animation tools.
Discovering the Right Animation Tool
ChatGPT suggested several platforms including DID and Runway ML for lip-syncing animation. However, after noting that Runway wouldn't lip-sync perfectly, the search expanded. This is where a crucial discovery happened through content creator Matt Farmer's video, which introduced Wan Video as a superior solution for lip-sync animation.
Creating Character Images in Midjourney
Generating Initial Turkey Characters
Using the prompts suggested by ChatGPT, the character creation process began in Midjourney. Multiple variations were generated to find just the right look—turkeys with personality and visual appeal that could carry a music video.
The Breakthrough: Using Google Nano Banana Pro for Multiple Views
The real challenge emerged when trying to create different camera angles and perspectives of the same characters. After numerous attempts and experiments, Claude Sonnet (specifically the thinking model) proved to be the perfect solution.
By uploading the original turkey character images to Claude and requesting specific modifications—"change this to a closeup," "put them on stage," "make them face each other in profile," "zoom closer"—it became possible to generate consistent characters from multiple angles. This capability was essential for creating a dynamic, engaging music video rather than a static performance.
Preparing Audio Segments for Animation
Cutting the Track into 10-Second Clips
Wan Video's animation feature works with 10-second audio segments. Following Matt Farmer's recommendation, the tool mp3cut.net was used to divide the mastered audio track into individual 10-second clips. Each clip would be paired with a corresponding animated character image.
Organizing the Production Workflow
With multiple audio segments and various character angles prepared, the production process became systematic: match each 10-second audio clip with the appropriate character view, upload both to Wan Video, and generate the lip-synced animation.
Animating with Wan Video
The Sync-to-Audio Feature
Wan Video's sync-to-audio feature proved to be remarkably effective. The process was straightforward: upload a character image, upload a 10-second audio clip, and let the platform generate the lip-synced animation. The first test produced nearly perfect results, confirming this was the right tool for the project.
Optimizing for Different Shot Types
Through experimentation, certain patterns emerged about what worked best:
Closeup shots delivered the most accurate lip-syncing. When focused on a single turkey's face, the mouth movements matched the audio almost perfectly.
Wide shots with multiple characters were acceptable but less precise. The lip-syncing was still functional but not as convincing as the closeup perspectives.
Profile and mid-range shots struck a good balance, offering variety while maintaining decent lip-sync quality.
Video Editing and Assembly in Kapwing
Building the Timeline
With all the individual 10-second animated clips generated, the final assembly took place in CapCut. The mastered audio track from BandLab was dropped into the timeline first, providing the backbone for the entire video.
Each animated clip was then placed sequentially, building out the complete performance piece by piece. The 10-second segments aligned perfectly with the pre-cut audio segments, making the assembly process smooth and efficient.
Creating Visual Variety
To keep the video engaging throughout, different techniques were employed:
Split-screen sequences featured both turkeys simultaneously, requiring individual animation of each side before combining them in the editor.
Guitar-playing shots didn't require lip-syncing, so these were created through standard Midjourney animation, slightly sped up to match the rhythm of the music.
Strategic camera movement included pans and zooms to add dynamism even within the 10-second clip limitations.
The Strategic Use of Lyrics
On-screen lyrics served a dual purpose in the final video. Primarily, they enhanced viewer engagement by allowing people to sing along. But they also provided a subtle visual distraction during moments when the lip-syncing wasn't absolutely perfect—giving viewers something else to focus on when the mouth movements might be slightly off.
Project Timeline and Efficiency
The entire project, from concept to completion, was accomplished with remarkable speed. Most of the work was completed in just a couple of hours on the first day, with only the final ending sequence finished the following morning. This rapid turnaround demonstrates the power of modern AI tools when used strategically and in combination.
The process included:
Experimenting with multiple AI platforms
Learning new tools like Wan Video on the fly
Generating dozens of character variations
Creating all animations and lip-sync sequences
Editing the complete video with effects and lyrics
Key Takeaways and Lessons Learned
Tool Synergy Matters
Success came from combining multiple specialized tools rather than trying to force one platform to do everything. Suno for music creation, BandLab for mastering, Midjourney for character design, Google Gemini for consistent character variations, Wan Video for lip-syncing, and Kapwing for final editing—each played a specific role.
The Importance of Problem-Solving
When initial approaches didn't work (like Google Gemini creating videos instead of providing instructions, or early lip-sync tools not meeting quality standards), persistence and willingness to explore alternatives led to better solutions.
Following Creator Recommendations
Matt Farmer's video recommendations for both Wan Video and mp3cut.net proved invaluable. Tapping into the knowledge of others who are experimenting with AI video creation can dramatically accelerate the learning curve.
Strategic Imperfection Management
Rather than pursuing absolute perfection in every frame, smart compromises (like using lyrics as visual distraction or mixing closeups with wider shots) allowed for a professional-looking result without getting stuck on technical limitations.
The Final Question
Will this AI-generated Thanksgiving turkey duet become a new holiday classic? Only time will tell. But the process of creating it demonstrates that professional-quality multimedia content is now accessible to individual creators willing to learn and experiment with the rapidly evolving landscape of AI creative tools.
The combination of humor, technical skill, and creative vision resulted in a project that's both entertaining and a showcase for what's possible when multiple AI tools work together in service of a single creative vision.
