How it works
How LevelCast evens out the host and the guest
Getting a whole episode to a sensible volume is something editors already know how to do. What bothers you on a drive is the difference between the two people talking in it.
Why the guest sounds quieter
The host talks into their own microphone, in a studio they know. The guest brings a different microphone, sits in a different room, and sometimes joins over a remote call. So each voice enters the episode at its own level.
Editing makes sure the episode as a whole plays at a sensible level, but the gap between the two people talking in it usually stays. At home it's a little annoying. In a car, with road noise, it means your hand is on the knob the whole time.
What the app does
LevelCast listens to the episode as it plays and fixes the gap as it goes, so the quiet voice comes up and the loud one comes down a little. All of this happens on the phone. The producer's file stays as it is, and no audio gets sent anywhere for levelling.
There are three strengths. Normal is the lighter touch. Strong closes bigger gaps and suits a noisy car, and Off plays the episode exactly as it arrived. You can hear the difference in the demo.
Where the episodes come from
Podcasts are published in an open feed, and the audio itself sits on the producer's server. The app finds shows in Apple's podcast directory, and you can also paste in a feed address or an episode link by hand. The episode plays straight from the producer's server, and the download is counted there.
Everything inside
- Speaker levelling at normal or strong, and you can set a different strength for each show.
- Intro skip. The app spots where the music ends and the conversation starts, and offers to jump there.
- Pace matching. Whoever talks slowly is sped up a little, and whoever rushes is slowed down.
- Playback speed from 0.8× to 3×.
- Capture a moment. Hold the back button on your steering wheel, or tap the button in the player, to save what was just said along with its transcript. Moments wait in your library, and you send them once you've stopped.
- Live transcript that scrolls like song lyrics, search inside the episode, and key points built from sentences actually said in it. Transcription uses iOS speech recognition, on devices that support the episode's language.
- The links from the episode notes, all in one place.
- A short automatic rewind after an interruption, so the end of the sentence doesn't slip past you.
- Downloads by hand, automatically or ahead of time, and a background refresh that tells you when a new episode is out.
- Sleep timer that lowers the sound gradually at the end, instead of cutting a sentence in half.
What's free and what's in a subscription
Free
Searching and following shows, strong levelling, a queue of up to three episodes, picking up where you stopped, manual download, intro skip, moment capture and the episode links.
Pro
Per-show levelling, an unlimited queue, automatic and advance downloads, speeds above 1.5×, pace matching, background refresh and alerts, and the sleep timer.
Pro AI
Everything in Pro, plus live transcript, search inside the episode and key points.
Both paid plans will be available through Apple, billed monthly or yearly, and the price will appear on the purchase screen before you pay.
Privacy
There's no account and no sign-up. Your library, queue, moments and settings are kept only on your phone. The app contacts Apple when you search for a show, and the producer's server to fetch the feed and the audio. There is one exception, which is transcription on a device that can't recognise speech locally in the episode's language. In that case iOS sends the clip to Apple's service, and it never reaches us. All the details are in the privacy policy.
Which devices
You need an iPhone with iOS 17 or later. The interface is in Hebrew and English.