Pre-recorded for our live Comic-Con panel, The Track You Never Heard, with Colleen Connor and Anna Capezzera. Colleen is blind and runs Audio Description Training Retreats. She built it with a specific rule: the blind professional teaches, as an equal, instead of getting called in afterward to give feedback. Classes cap at eight. Notes go out ahead of time so class time is practice and feedback, nothing else. Anna oversees the Audio Description department at Deluxe. She started as a writer, then learned the audio side so she could talk to her engineers in their own language. She walks through the whole process, order to delivery, including the weekend her entire team wrote in parallel to hit a schedule that was cut to a quarter. Then the harder conversations. Distributors can drop full seasons and split them across writers, so nobody holds the whole story. Colleen makes the case for blind quality control catching what sighted QC won't. Anna makes the case for hiring blind voice performers and reading what the audience actually says. And Colleen's line, which is the whole episode in nine words: You can't be extremely fast and good at Audio Description.

Soud Habbas spent his career building systems for nonprofits working in some of the hardest environments in the world: Jordan, Egypt, Rwanda, and the West Bank. As chief operating officer, he made sure programs actually worked, stayed honest with funders, and didn't put people at risk. In this conversation, Soud tells the story of a $40,000 bet on a small women's dairy cooperative that outperformed programs ten times its size, and a million-dollar-a-year partnership with the John Deere Foundation that worked because of one thing: trust. He talks about what it takes to hold a partner accountable without losing the relationship, and how he's applying the same principle to his own life after a recent layoff sent him back to the bar exam he never took. If you've ever been told your idea is too small to matter, this one's for you.

Roy Samuelson sits down with Maria Miller, former Senior Private Sector Specialist with the International Finance Corporation. Maria spent 14 years traveling to more than 20 countries, including post-conflict zones, helping governments rebuild the economic rules that determine who can start a business, get a job, or build a future. She went to Liberia 3 months after a civil war ended. She monitored elections in Ethiopia. She worked in places where what she assumed was essential turned out not to be. Then she retired early and moved her work closer to home. She now sits on the board of Girls Global Academy in Washington, DC, the city's first all-girls International Baccalaureate charter school, as vice chair and head of community engagement. In this conversation: what Liberia taught her about what's essential, how you change a system slowly without claiming the credit, the personal cost of a career built across 20 countries, and what a teenage girl who said "I'm networking" revealed about why possibility has to be visible before it can be chosen.

Connor Scott-Gardner is a blind accessibility advocate, disability rights writer, and someone who has spent years thinking carefully about what it means to be helped, witnessed, and included. In this second half of our conversation, Connor names something the accessibility world doesn't say often enough: good intentions are not the same as ethical behavior. He talks about Be My Eyes volunteers who filmed blind users without consent, what that reveals about power and trust, and why intent is never the whole story. He also turns the mirror on himself, reflecting on a viral video he made years ago that still sits uncomfortably with him. Then the conversation opens up. Connor shares his complicated early relationship with Audio Description, how the field has changed in its treatment of race and identity, and why the expansion into live performance, including ballet, represents something bigger than access. It represents a field that is finally asking the right questions. This is a conversation about trust, craft, and what it looks like when an industry stops assuming it knows what its audience needs and starts listening instead.

Connor Scott-Gardner has presented at the European Parliament, Westminster, and SXSW. He also goes home, gets tired, and sometimes thinks he hasn't done anything. This podcast will prove that he has done much. In this first conversation, he talks about what genuine support for blind young people looks like up close, and what gets in the way of it. He names a trap he's fallen into himself: pointing out every barrier without offering a direction. And he explains why awareness, on its own, has a ceiling. This is a precise, generous conversation with someone who has thought hard about the difference between performing concern and actually creating change. Part 2 is coming.

Something went wrong during a live show. An aerial performer fell. The Audio Description professional described it as it happened. Then told the audience the performer was okay. No script covered that moment. No algorithm was ready for it. A human was paying attention, and that made all the difference. Dr. Cynthia Bennett is a blind researcher at Google who studies what happens when technology is built without the people it's supposed to serve. In this conversation, we get into why imperfect, unpredictable human moments are exactly what AI struggles with most. Why she thinks claiming to be objective is less honest than just telling you where you stand. And how the process of making research accessible to her, as a blind researcher, actually changed what her research found. This one stays with you.

Josh Miller came to accessibility through telecom, media consulting, and a business school project at MIT. That project ran headfirst into a quote so high it would have eaten up half the program's annual budget. That collision between intention and financial reality is where 3Play Media was born. 19 years later, Josh leads one of the largest media accessibility company in North America, serving over ten thousand customers across education, entertainment, and enterprise. He sits on the CTA's Content and Entertainment Council. He has watched captions go from a niche accommodation to something people without hearing loss prefer. And now, he's watching Audio Description continue on that same journey. In this conversation, Josh talks about what it actually takes to build quality at scale, why not all content is equal (and what that means), and what he believes filmmakers owe a blind audience as storytellers. He also says a sentence that carries a lot: Accessibility is about enabling people to consume content however they want or need to. We unpack what it means in practice.

Dr. Cynthia Bennett is a blind researcher at Google who studies the AI systems that are supposed to serve blind people. She found out blind people were making Audio Description, and she, a blind person, had no idea. That gap changed how she works. In this conversation, we get into what authentic representation actually requires, why "good enough" accessibility protects the wrong people, and the tool she helped build so blind professionals can finally create AD themselves.

Alex Howard and Lee Pugsley are back on The ADNA Presents, and a lot has changed since their first visit. The Darkroom podcast hosts and co-founders of the Blind Film Critic Society join Roy Samuelson for a conversation about what it means when blind critics evaluate Audio Description -- not just as consumers, but as people who can now name the performers, writers, and engineers responsible for the work. They talk about why crediting the full Audio Description team is gaining ground, how the back catalog movement is raising the bar for quality, and what it looks like when the field shifts from technical compliance to genuine artistic contribution. Alex also shares something personal: that bad Audio Description at a movie theater still makes him feel like he doesn't belong. It's a quiet, honest moment that says something important about what's actually at stake. The Blind Film Critic Society is growing. Darkroom is expanding its scope. And the conversation about who gets credited -- and who gets to evaluate that work -- is getting sharper.

Darian Rodriguez Heyman has spent thirty years helping mission-led organizations stop reinventing the wheel. He's the author of AI for Nonprofits, founder of Helping People Help, and someone who has thought seriously about why good intentions so often produce predictable, avoidable mistakes. What drew me to this conversation is that the gap he keeps closing, between what's available and what people actually reach for, lives in Audio Description too. We talk about the nonprofit assumption, what AI adoption actually looks like inside organizations right now, and why the accessibility sector may be solving important problems, and why you don't have to live in sacrifice to do the right thing. There's also a moment where he shares something he couldn't find while researching his book. It stuck with me. I think it'll stick with you too.