There is a funny assumption that I run into a lot with audio only shows, right, which is that because there is no camera, there is basically nothing to edit, you trim the start and the end, you run it through a plugin, you upload it, and that is the job done, and I understand why people think that, but to be very honest an audio only show is the format where the edit has the least to hide behind, because the listener has nothing to look at, so every sound is the whole experience.
A video podcast can get away with a slightly awkward moment because the viewer sees someone smile, or nod, or reach for a coffee, and the picture explains it. On an audio show, that same moment is just two seconds of somebody breathing near a microphone, and the listener's thumb is already moving towards the skip button. So when someone asks me what an audio podcast editing service should actually include, my answer always starts with the same thing, a human being who listens to the entire episode before cutting a single second of it.
What does a proper audio only podcast edit include?
A proper audio edit includes a full listen-through, content cuts, filler and mistake removal, levelling between voices, noise cleanup, music and sponsor placement, and a final master at the right loudness for podcast apps.
Let me go through that loosely, right, in the order we actually do it. That is one, the listen-through, where the editor hears the full conversation and marks what is working and what is not, the tangent that goes nowhere, the story told twice, the question the guest misunderstood. Secondly, the cuts themselves, which are decisions about meaning, not about silence. Then the cleanup, removing the ums that pile up, the false starts and the coughs, without making the host sound like a robot, and I wrote a whole piece on cutting filler words without robotic audio because it is easy to overdo. And on top of that, levelling and noise, music and transitions, sponsor reads placed where they belong, and finally the master.
With us, a person on the team hears your episode end to end, and software only ever helps with the technical cleanup, so the decisions stay human. That matters on audio more than anywhere, because there are no camera angles to cut between, so the only edit left is the one that decides what gets said and how it flows.
Why does an audio only show still need a full listen-through?
Because the decisions that make an episode good are about meaning, and you cannot judge meaning from a waveform, so somebody has to hear the conversation the way the listener will.
This is where auto-editing goes wrong, right, and why auto-cut podcasts feel so choppy. A silence detection tool sees a gap and removes it, but a gap after a hard question is a person thinking, and a gap after a joke is the laugh landing, and a gap before a big point is the host building up to it. Take those out and the conversation turns into a rapid series of sentences that technically contain everything and somehow feel exhausting. We do not use silence detection for automatic cuts at all, for exactly that reason.
The other half of the listen-through is structure. What I notice over and over on audio shows is that the strongest stretch of an episode tends to turn up well after the start, once everyone has relaxed, and part of the editor's job is spotting that and making sure the first few minutes hold the listener long enough to reach it. Sometimes that is trimming a long warm-up, sometimes it is a short cold open, and you can only make that call after hearing the whole thing.
Nobody can see your face on an audio show, so every um, every breath and every bump in the level is doing all the talking for you.
How loud should an audio podcast be for Apple, Spotify and the other apps?
For Apple the published target is an overall loudness of roughly minus 16 LKFS, give or take a decibel, with true peaks kept at or under minus 1 dB FS and AAC as the preferred format for the feed, and I treat that as a sensible target for the other apps as well.
You can read the detail on Apple's audio requirements and its broader podcast requirements for the feed itself. In plain words, loudness is about consistency, right, so your episode does not sound much quieter than whatever your listener played before it, and true peak is about avoiding distortion when the file gets encoded. Hitting a number is the easy part, any tool can do that in a few seconds. The harder part is getting there without squashing the voices so they sound flat and tiring, which is the difference between a master and a limiter turned up too far.
Levelling between voices comes first, and it is the thing listeners notice most without knowing what to call it. When a host is close to a good microphone and a guest is on a laptop across the world, one is loud and warm and the other is quiet and thin, and the listener spends the episode adjusting the volume. Getting both voices to sit together is most of what our podcast audio cleanup work is about, along with hum, hiss, room echo and the odd keyboard clatter.
What happens to music, intros and sponsor reads in an audio edit?
They get placed and mixed by ear, so the intro hands over cleanly to the conversation, music sits under voices rather than on top of them, and sponsor reads land at natural breaks.
Music is where audio only shows often sound cheap without meaning to, right, the theme starts loud, the host starts talking over it and the two fight for a few seconds. A good mix ducks the music under the voice and brings it back on the transitions, so it feels like part of the show and not a separate track. Sponsor reads need the same care, they should sound like the host talking, placed where the conversation naturally pauses, not dropped in the middle of a guest's best story. If you want to go further on the whole mixing side, I wrote about what podcast mixing and mastering should actually deliver.
Can software do an audio edit instead of a person?
Software can do some of it well, and for a clean solo show recorded in a quiet room an automatic levelling and loudness tool may honestly be all you need.
I want to be fair here, right, because automatic tools are good at the technical layer, they level, they hit loudness targets and they remove steady background noise quickly and cheaply, and if your show is one person on a decent microphone with no guests and no music, paying a team might be more than you need. Where software falls short is everything that involves judgment, which tangent to cut, which pause to keep, where the sponsor read belongs, whether the guest's answer actually made sense. Our comparison of Auphonic vs human audio mixing goes through what each one fixes, and the same logic applies to the wider categories, a solo freelancer can be great but has one calendar, an AI auto-editing tool is fast but cuts by silence, and a large production company is usually more studio than a weekly audio show needs.
When should an audio only show add video?
Add video when you want to grow on YouTube, want clips for social, or when your guests and topics are visual enough that people would genuinely want to watch them talk.
Audio only is a perfectly good format, right, and plenty of shows should stay that way, especially intimate conversation shows where the audience listens on a commute or a walk and would never sit and watch anyway. But there is a ceiling on discovery when the only place your show exists is inside podcast apps, which is what I wrote about in why audio only growth stalls. Video gives you YouTube as a second home for the show, and it gives you vertical clips for short-form platforms. If you are not ready to film, YouTube also lets you deliver a podcast through an RSS feed, which can be a gentle first step. The full trade-off is laid out in our page on audio only vs video podcast.
The nice part, if you do switch, is that the editing habits carry straight over. The listen-through, the cuts, the levels and the master are the same work, and the video just adds sync, angles, colour and graphics on top. So nothing you build as an audio show is wasted if a camera turns up later, the sound you already have simply becomes the backbone of the video edit.
How much does audio only podcast editing cost with us?
Audio only shows are edited by our team too, and they are priced on the call, because the length, the number of voices and the state of the raw audio change the work far more than on a video edit.
To be very honest, a clean forty minute solo episode and a noisy ninety minute panel with four remote guests are completely different jobs, so I would rather hear a recent episode and quote properly than publish a number that fits nobody. So send me a recent episode, or book fifteen minutes and play me one, and you get the price before we finish talking. There is no contract to sign up to either, each episode is paid for on its own and you can stop whenever the show does, and the session files and finished masters stay yours to keep. If you want show notes and chapters written by the person who edited the episode, that is on our page about podcast show notes and chapters, and the general plan details are on the pricing section. Does that feel like a fair way to price it, given how different two audio shows can be?
What should I send for an audio edit?
Send the separate audio tracks for each voice if you have them, plus the mixed recording as a backup, your intro, outro and music files, any sponsor reads, and a few recent episodes if the show already has a sound you like.
Separate tracks make a very very big difference, right, because they let us level and clean each voice on its own, so if your recorder or remote platform offers them, turn them on. Drop the lot into the shared folder, however big the tracks are, and the first edit is back with you within two days, with your notes applied a working day after you send them. Across the more than 400 episodes that pass through Pixel Samy Podcast each month, the audio only ones are, at the end of the day, where you hear most clearly whether somebody listened before they cut, and that listening is really the whole service. So yeah. That's my way of saying it.