Moneybird sent Frank a cheerful anniversary email: ten years as a customer, and, by the way, the price is doubling in two weeks. That single email opened up a broader frustration with what writer Cory Doctorow calls enshittification, the gradual degradation of digital products as the pressure to extract more revenue overrides the original reason the thing was built. Dropbox is the clearest example: a few years ago it was rock-solid for moving large video files; now it quietly offloads everything to the cloud, prompts you to back up every external drive you plug in, and nudges you toward a higher storage tier. The product has drifted, and the price has not.

The SaaS math compounds fast. Dropbox, invoicing software, Adobe, Microsoft 365 with a terabyte of OneDrive nobody uses, Netflix. Each feels manageable month to month, but Frank's rough estimate puts the total around a thousand euros a year flowing toward tools that have either gotten worse or are barely used. The response he's started taking is replacing paid subscriptions with open-source or self-hosted alternatives wherever possible. He rebuilt his own website and Sana's photography portfolio using Claude, cutting several hundred euros a year down to about fifty, with a setup where new galleries can be added by handing Claude a batch of photos and a single instruction.

Frank and Sana also get into how they actually use AI day to day: Frank uses Claude Code to run multiple tasks simultaneously across sessions and built a small app to control his Blackmagic cameras remotely over Bluetooth; Sana throws festival lineup flyers into ChatGPT to get clean formatted text and uses it to soften emails when Dutch directness would land wrong with contacts in Japan. They close on a live test of whether Apple Pages and Microsoft Word can genuinely collaborate through the .docx format (they can), and whether that's enough to justify dropping a Microsoft 365 subscription you're only keeping for Word and a storage plan you never fill.

Published on Subwave

https://subwave.app/@femous/post/enshittification-and-the-saas-reckoning

Two Blackmagic cinema cameras, a laptop's built-in FaceTime HD camera, one XLR microphone, and a Thunderbolt hub. That's the entire multicam studio setup, and it costs a fraction of what a single Sony rig would run once you add capture cards and accessories. The Blackmagic UltraStudio recorders are Thunderbolt devices, so they daisy-chain into a single hub and effectively become a multicam switcher on the cheap. The microphone connects directly to the Blackmagic camera via XLR, which means studio-quality audio with no separate audio interface. The trade-off is manual focus, since these cinema cameras have no autofocus, but autofocus hunting mid-shot has its own problems.

The recording side runs through Detail Studio, the Mac multicam recorder. Detail iOS runs on iPhone and iPad and, on a 2025 iPad with an M3 chip, tears through auto-editing and transcription in seconds. The gap is that these two apps don't natively talk to each other, so I built a converter tool to bridge them. The test: record a five-minute, three-camera session in Detail Studio, export it, AirDrop it to Detail iOS via the converter, and see if the project loads cleanly and edits properly.

Routing footage through Detail iOS is worth the extra export step because of raw speed. For unscripted footage with no teleprompter, plenty of rambling, and the usual filler words, the auto-edit strips out the umms in seconds. Five minutes of three-camera footage goes through the converter, into Detail iOS, and the result is a clean, editable project ready to cut.

Published on Subwave

https://subwave.app/@femous/post/testing-a-three-camera-multicam-export-on-a-budget

A full minute of filler words and dead silence can hide inside a single unscripted recording, and most people don't notice until they see the numbers. I walk through how Detail surfaces both: every "um" and "uh" that slipped through, and every stretch of dead air where nothing was said. In one test recording, silences added up to 41 seconds and fillers added another 31, together accounting for more than a minute of content that contributed nothing to the final video.

Using a teleprompter cuts filler words significantly because there's something to follow. Without a script, the ums and uhs stack up fast. For those who record off-the-cuff, Detail's transcript panel does the work of finding everything: tap Silences, tap Fillers, and the app flags every instance automatically. Then it's two taps to delete both categories and let the timeline compress around the cuts.

The pacing difference after those deletions is immediate. A minute removed from a five-minute video makes it 20% tighter, and viewers register that as confident delivery rather than just faster speech. If your videos are losing viewers early, dead air and filler words are the first place to check.

Published on Subwave

https://subwave.app/@femous/post/cut-a-minute-of-dead-air-in-seconds

A $50 teleprompter rig from Amazon, an iPad running Detail, and a Sony A7 IV connected over USB is all it takes to get pro-camera eye contact on every take. I walk through how to connect an external camera as the recording source inside Detail, pull in a second angle from an iPhone on the same multicam session, and get both feeds live within a few taps.

One setting trips people up the first time: when you place the iPad in the teleprompter clamp and switch to full-screen mode, the script appears flipped because the one-way mirror reverses it. The fix is opening the teleprompter menu in Detail and tapping the mirror button. Font size up, script readable, and you are ready to record.

Detail has a built-in teleprompter that works with your iPhone's selfie camera and requires zero extra hardware. It is the right starting point for most people. But routing an external pro camera through Detail and placing the iPad in a physical rig produces a noticeably sharper, better-framed image while keeping the eye-contact advantage of reading through the mirror. The hardware cost stays low and the setup is faster than it looks.

Published on Subwave

https://subwave.app/@femous/post/using-details-teleprompter-with-a-pro-camera

Frank from Detail recorded a demo in his studio in the Netherlands with a mobile air conditioner running at its loudest fan setting, using the internal iPhone microphone on purpose to stress-test the new audio enhancement feature. The result is clean, present voice audio with the AC noise gone, processed entirely on-device with no server upload and no network connection required.

The workflow is a single tap. After recording, you open the timeline, scroll to the audio section, and tap Enhance. For clips under a minute the processing is near-instant, and a slider lets you compare the original and processed audio side by side so you can hear exactly what changed. Detail trained the underlying models on their own audio data, which is why the results feel calibrated rather than generic.

Audio enhancement is a pro feature in Detail. For anyone recording on an iPhone in less than ideal conditions, it removes the need to re-record in a quieter space, open a separate audio editor, or accept noisy footage. You get clean audio without changing how or where you record.

Published on Subwave

https://subwave.app/@femous/post/details-on-device-audio-enhancement-2

Recording near a concrete truck means fighting wind, engine noise, and general chaos. The simplest fix is also free: move further from the noise source. Even while keeping the truck in frame, adding physical distance changes what the microphone actually picks up.

When distance isn't enough, getting the mic closer to your mouth does a lot of work. Your voice becomes louder relative to everything else, which pushes background noise further down in the mix. A windshield helps with wind specifically, and a quieter location, when accessible, remains the cleanest option of all.

For situations where there's no external mic at all, Detail's on-device audio enhancement handles the kind of background noise that makes a clip otherwise unusable. It runs entirely on your phone, processes near-instantly, and your audio never leaves the device. It's a pro feature, but for anyone shooting in unpredictable environments, it turns a salvageable clip into a usable one.

Published on Subwave

https://subwave.app/@femous/post/fix-bad-audio-fast

Filler words like "um" and "uh" are a normal part of unscripted speech, but they cause real problems on video. They break pacing, pull viewers out of the moment, and make otherwise solid takes feel unpolished.

Detail's latest update adds automatic filler word removal, cutting those hesitations out of your footage without any manual hunting or re-recording. You get cleaner audio and a more watchable result without changing how you record.

The filler word removal feature is available now on the App Store.

Published on Subwave

https://subwave.app/@femous/post/detail-now-removes-filler-words-automatically

Subwave offers a mobile-first video experience designed around intentional watching instead of algorithmic feeds. Every piece of content appears as a clean, full-screen card showing what it is, how long it takes, and who made it. Speed controls let you watch long-form content at 1.5x or 1.8x depending on how fast someone talks.

The platform supports multiple formats for the same content. Tap to switch from video to article view with proper formatting and typography, perfect for situations without sound. Audio mode lets content play in the background like a podcast. Everything can be downloaded for offline access with no premium tier or paywall required.

The app works across iPhone, Android, and web, with shortcuts to short-form content from followed creators. The goal is a viewing experience that feels calm, intentional, and genuinely enjoyable rather than designed to maximize screen time.

Published on Subwave

https://subwave.app/@femous/post/how-subwave-makes-video-content-actually-enjoyable

Frank explores the creator experience in Subwave, focusing on how the platform consolidates what typically requires multiple tools into a single workflow. Format freedom means choosing between video, audio, or written articles without platform constraints.

The AI features adapt to individual creator voices, generating outlines, show notes, and full articles that match chosen tone and style rather than defaulting to generic output. Content becomes accessible through formatted articles with pull quotes, clean typography, and full transcripts.

Distribution includes automated short-form content creation for social platforms, eliminating the need for separate tools for recording, transcription, AI writing assistance, newsletters, and blogging.

Published on Subwave

https://subwave.app/@femous/post/one-channel-one-workflow--the-creator-side-of-subwave

Frank and Jacco launch their podcast by diving into one of the most unsettling developments in AI security: Anthropic's Mythos model, which identified 271 previously unknown vulnerabilities in Firefox in a single night. The company made the unprecedented decision not to release the model publicly, raising urgent questions about who controls these capabilities and what happens when they fall into the wrong hands. Jesse Klaver, former AIVD intelligence officer and member of parliament, joins to explain how AI-powered vulnerability detection works, why it's like a defensive crypto mining race, and how state actors from Russia, China, and Iran are changing the cybersecurity landscape.

The conversation expands to cover the Solvinity controversy in Dutch politics, the balance between accessing cutting-edge AI tools and maintaining digital sovereignty, and what democratic control over critical infrastructure actually means in practice. Frank shares an unsettling realization from the Fair Plus robotics conference in Shenzhen: if robots can already perform precision tasks, what happens to manual labor in 10 to 15 years when humanoid robots can be deployed by subscription?

Digital independence, the future of work, and the fundamental question of whether technology serves democracy or undermines it emerge as central themes. Frank and Jacco are creating space for substantive conversations about where technology is taking us, without the usual marketing fluff, one episode at a time.

Published on Subwave

https://subwave.app/@femous/post/technology-democracy-and-the-digital-arms-race