Blog

How to turn a podcast into a video without uploading it

By Nicholas Johnson, who makes Podwobble · 16 October 2026

The short answer: use software that runs on your own computer instead of a website. A desktop app such as Podwobble, a video editor, or one FFmpeg command turns the audio into a video without sending it anywhere. Captions are the step that most often goes online, so check how each tool makes them, or test it with Wi‑Fi off.

Almost every guide to turning a podcast into a video starts with the same step: upload your audio. For a finished, public episode that is usually fine. For an episode that isn’t out yet, an interview your guest hasn’t approved, or a show you make for a client or your own company, it may not be. This guide covers the routes that keep the file on your own Mac or Windows PC, how to check whether a tool really does, and how to get captions without the audio leaving the room. One of the routes is Podwobble, the app I make, so read that part knowing who wrote it. Facts about other tools come from their own pages, linked as I go, as they stood on 16 October 2026.

Why keep a podcast off someone else’s server?

Because the audio isn’t public yet, or isn’t only yours to share. That isn’t a judgement on any web tool, but once you upload, the copy sits on their servers under their policy, for as long as that policy says. The usual reasons people want to avoid that:

• An episode under embargo. Clips made before release day are exactly the files you’d rather not hand around.
• A guest’s raw interview. The unedited recording may hold things the guest asked you to cut.
• Someone else’s show. Client work, or an internal company podcast, can come with rules about where the audio may go.
• Big files on a slow connection. An hour of 48 kHz, 24-bit stereo WAV is about 1 GB. Uploading that before you can start, every week, adds up.
• No connection at all. A train, a plane, a studio with bad Wi‑Fi.

How do you tell whether a tool uploads your audio?

Read what it says, then test it, because “in your browser” doesn’t always mean uploaded and “desktop app” doesn’t always mean local.

1. Read the page. Look for the words upload, cloud, our servers or locally on the tool’s own site and in its privacy policy, and check each feature separately: a tool can build the video locally and still send the audio off for captions.
2. Test with Wi‑Fi off. Open the tool, turn off Wi‑Fi (and unplug the network cable), then take a short test file all the way through: import, captions, export. Whatever still works is happening on your computer.
3. For a browser tool, watch the network. Open the browser’s developer tools, go to the Network tab, and add your file. A request roughly the size of your audio file means it went up.

Two examples of why the test matters. Cleanvoice’s free Audio Visualizer runs in a web page, yet its page says it works “locally in your browser” and doesn’t keep your recording; it takes clips of up to 5 minutes. Microsoft’s Clipchamp editor (a Windows app, or Edge or Chrome on a Mac) goes the other way: with a personal account it accesses your media directly from your computer and processes it locally, unless you turn on content backup, but its autocaptions use Microsoft’s Azure services to process the audio.

Which tools make the video on your own computer?

Desktop software and a few in-browser tools. The difference between them is mostly captions and how much you build by hand.

FFmpeg is the free, no-watermark route. One command puts a picture and a moving waveform over the audio and saves an MP4; I’ve explained it line by line in adding a waveform to audio, and the still-picture version is in converting an MP3 to an MP4 on Windows (the command is the same on a Mac).

Video editors such as the free version of DaVinci Resolve or Apple’s free iMovie give you full control, and you build the motion, the captions and each aspect ratio yourself. Fine for one video, slow for a weekly show.

Why are captions the step that usually goes online?

Because turning speech into text takes a speech recognition model, and many tools run that model on their own servers rather than ship it to you. It is the part of the job to check first.

Some tools do run it on your computer, with limits:

• Podwobble transcribes with a speech model that ships inside the app, so Captions work with no connection. It is English only.
• Final Cut Pro has a Generate Captions feature that needs a Mac with Apple silicon and is U.S. English only. Apple says it downloads the language model the first time you use it, once, and that download needs an internet connection.
• Clipchamp’s autocaptions cover many languages, but as above, the audio is processed by Microsoft’s Azure services. Microsoft says no person has access to it.

Offline captions for a show that isn’t in English

This is the gap most guides skip. whisper.cpp is a free, open-source speech recogniser that runs on Mac and Windows, and its command-line tool can write an .srt caption file in the language you name. It is for people comfortable with a terminal: you build it from source with CMake, following its README, and download a model. Models are multilingual unless the name includes .en, so for a Spanish or German show use one like small (466 MiB on disk) rather than small.en.

The tool reads 16-bit WAV files, so convert the episode with FFmpeg first, then transcribe (each is one line):

ffmpeg -i episode.mp3 -ar 16000 -ac 1 -c:a pcm_s16le episode.wav
./build/bin/whisper-cli -m models/ggml-small.bin -l es -osrt -f episode.wav

-l es sets the spoken language (Spanish here; auto tells it to detect the language) and -osrt writes the result to an .srt file. Read it through before you use it, since names and jargon are where any speech model slips. In Podwobble, bring the file in with Import SRT/VTT… in the Captions step and style it like any other captions.

How do you do the whole thing offline in Podwobble?

Podwobble is a desktop app for Mac and Windows made for this one job. It needs no account, has no analytics, and audio, images, projects and transcripts stay on your computer. Each project is one .podwobble document, saved wherever you choose, that holds copies of all its media, so an embargoed episode lives in one file you control.

1. Voices. Drop in the episode (MP3, WAV, M4A, FLAC, AIFF or Ogg) and give each host a photo, a name and a colour. Solo recordings, if you have them, let it work out who is talking when.
2. Look. Pick one of six templates or build your own: a ring around the host photo that reacts to the voice, or EQ Bars, Rise, Wave or Dots.
3. Captions. Transcribe on your computer, in English, or use Import SRT/VTT… for another language.
4. Export. Tick 16:9, 9:16, 4:5 and 1:1 and they render in one go as H.264 MP4 at 30fps, for the whole episode or between an in and an out point.

To be exact about the network: the app works without an internet connection, but the version you download from this site checks for updates when it opens, and checks a Pro licence key if you have one. Neither involves your audio. The full walkthrough is in turning a podcast clip into a video, and it works the same on Windows. Free exports carry a small Podwobble mark in the corner; Pro removes it. On a Mac it needs Apple silicon (M1 or later) and macOS 13 or later; on Windows, 64-bit Windows 10 or 11 on an Intel or AMD processor.

Which route fits which job?

• One short clip, nothing to install. Cleanvoice’s in-browser visualiser, for clips up to 5 minutes.
• Whole episodes, no watermark, no budget. FFmpeg, and whisper.cpp if you want a caption file too.
• A one-off video with full control. A video editor, with its caption feature checked using the Wi‑Fi test.
• A clip every week, in every shape, with captions, all local. That is what Podwobble is for. I’ve compared it with the other free ways to turn a podcast into a video too.

Questions

Does a browser-based tool always upload my audio?

No. Some, like Cleanvoice’s Audio Visualizer, process the file inside the browser. Most podcast video sites do upload it, though, so check the tool’s own page or watch the browser’s Network tab.

Can I get captions without uploading the audio?

Yes. Podwobble transcribes English on your computer, and whisper.cpp can write an .srt file in other languages offline. Clipchamp’s autocaptions are the exception to its local editing: they send the audio to Microsoft’s Azure services.

Does Podwobble need an internet connection?

No. It works offline. The direct-download version checks for updates when it opens, and checks a Pro licence key if you have one; your audio stays on your computer.

Can I make a full-length episode video offline, not just a clip?

Yes. FFmpeg, a video editor and Podwobble all handle whole episodes. Cleanvoice’s in-browser tool stops at 5 minutes.

Written with the help of AI, from my notes on how Podwobble works. Facts are checked against the linked sources.