Hands-On Tech 238: Organizing Home Videos with Facial Recognition
Host: Micah Sargent
Date: October 19, 2025
Podcast: All TWiT.tv Shows (Audio)
Episode Overview
In this episode of Hands-On Tech, host Micah Sargent answers a listener question about how to organize decades’ worth of digitized family videos—specifically, how to identify and group clips by the people appearing in them using facial recognition technology. Micah explores available options across macOS and Windows, highlighting both simple and more technical solutions for viewers looking to categorize home movies efficiently.
Listener Question & Main Theme
- Listener: Mike, a long-time TWiT podcast fan, has digitized his late father's collection of 8mm family films into MP4 files and wishes to group video shots by individual family members.
- Main Question: Is there any facial recognition software for movies that can help automatically sort these clips by person, ideally on Windows, but potentially also using a Mac mini M4?
Key Discussion Points & Solutions
1. Facial Recognition in Video: The Challenge (02:10)
- Micah notes most facial recognition tools are photo-only:
- Quote: "There are so many, many, many tools out there that have facial recognition built in, but they only work for photos...every single one would be like 'the most powerful tool for organizing your media'...and then as you got into it...it was always about photos."
— Micah Sargent (03:05)
- Quote: "There are so many, many, many tools out there that have facial recognition built in, but they only work for photos...every single one would be like 'the most powerful tool for organizing your media'...and then as you got into it...it was always about photos."
Windows-Specific Search Result
- No suitable video facial recognition found for Windows, despite exhaustive search.
2. Solution 1: Apple Photos App (Mac) for Video Face Recognition
Timestamp: [03:40]
-
Recommendation: Use your Mac mini M4’s built-in Photos app, which supports facial recognition for both photos and videos.
-
How it Works:
- Add all videos to the Mac Photos app.
- Allow the Mac to stay on or asleep but running, so the background process can analyze files.
- Face detection happens passively; you need patience as it may take from a day up to a week.
-
Steps:
- Import video files to Photos.
- In an open video, click the menu bar > View > Show Face Names. Assign names to circles around detected faces.
- The “People and Pets” section will populate as faces are classified.
- You can then filter videos by individual people and export the grouped clips.
- Pro-tip: Export categorized footage to your preferred editing machine (e.g., Windows PC).
-
Quote: "Let the Mac sit. Let it be alive, let it be running. I would say at least a day...this is a background process."
— Micah Sargent (04:11)
3. Limitations of Photos App Face Recognition
Timestamp: [08:43]
- The facial recognition analysis is entirely background and beyond user control.
- Processing duration is unpredictable and may not meet urgent needs.
- Quote: "The problem with that tool is that it is a background process that Apple doesn't give you control over...it could be that it takes a day...it could be that it takes a week. It's hard to say because you don't have direct access to that background process, which is very annoying."
— Micah Sargent (08:48)
4. Solution 2: "Immich" – Open Source Self-hosted Video/Photo Organizer
Timestamp: [09:30]
- Micah introduces Immich (https://immich.app), a free, open-source alternative that features robust facial recognition—including for videos.
- About Immich:
- Requires self-hosting—run locally, no network-attached storage (NAS) or paid server necessary for a one-time use.
- Recommended: Use Docker Desktop (free for personal use) to quickly deploy on a Mac.
- Once set up, import your videos, let Immich process facial data, and view/export groupings by identified person.
- Comprehensive and active documentation available for setup and troubleshooting.
- Optional: Support the project with a $25 individual contribution.
- Quote: "Image—IM M I C H—Image is a tool that you have to self-host...specifically for this. It literally has a whole category about face facial recognition in photos and videos..."
— Micah Sargent (09:35) - On Usefulness: "So while for a short period of time it might be taking...it's going to take up a lot of space...but as this tool to get this job done. And then once you're done...no big deal."
— Micah Sargent (12:28)
5. Flow & Structure Advice
- For Windows users, no built-in or mainstream solution at the moment.
- Use Mac Photos for “out of the box” functionality, understanding its patience requirement.
- Use Immich for more control, efficiency, and for advanced users comfortable with running local servers or Docker containers.
Notable Quotes and Memorable Moments
- Micah’s Understanding: "I always say this anytime it's a question that I feel there are at least five other people out there who are also asking...love it when it's even more! I'm going, yes, yes, yes. Let's answer this question." (02:00)
- Encouragement for Feedback: "I would love, love, love to hear back from you on if you decided to go the Apple Photos route and that was enough for you, or if you decided to go the Immich route and how this tool worked for you." (14:29)
Timestamps for Important Segments
- [01:57] — Show introduction and listener question
- [03:40] — Overview of the challenges (photo vs. video face recognition)
- [04:11] — Mac Photos solution explained; step-by-step
- [08:43] — Limitations of Mac Photos’ facial recognition
- [09:30] — Immich open-source solution described in detail
- [12:28] — Advice on using Immich as a one-time tool
- [14:29] — Micah invites feedback from the listener
Tone & Style
- Friendly and encouraging: Micah strikes a helpful, “walk-you-through” tone, addressing both beginners and tech-savvy users.
- Engaging and empathetic: Micah relates to the listener’s challenge, validating their frustration and curiosity.
- Thorough and honest: Points out not only what works but also the pain points and limitations involved.
Final Thoughts
Micah Sargent’s in-depth response offers practical, actionable advice for a common problem among those digitizing large video archives. Both beginner and advanced paths are presented: use macOS’s Photos app for a “set it and wait” workflow, or dive into the technical but powerful open-source Immich if you need more speed, flexibility, or feel frustrated by Apple’s lack of control and transparency. The episode closes with an invitation for listeners to share their real-life results—a nice touch reinforcing the podcast’s community-driven nature.
Contact:
Questions for Micah? Email hot@twit.tv
Next Episode Teaser:
Micah will be back next week tackling another practical tech conundrum!