The fastest way to lose a scientist's trust is to make them sound more certain than they are, right, and an editor can do that without changing a single word, just by trimming the half sentence where they said "in mice" or "so far" or "we think", or by cutting a clip that starts after the caveat and ends on the exciting bit, and suddenly a careful researcher is on the internet saying something they would never put in a paper.
That is the whole reason I think science communication shows need a different kind of edit, right, because the ordinary podcast instincts, tighten everything, chase the strong line, cut the qualifiers because they slow things down, are exactly the instincts that get a science show in trouble, so the job here is to keep the pace a general audience needs while protecting the precision the guest has spent a career building.
Why is accuracy an editing problem and not only a research problem?
Because most accuracy mistakes in a science episode are created in the edit, by cutting a caveat, merging two statements, or reordering an answer, so the editor has to understand what was said well enough not to distort it.
My editors are not fact checkers and I would never pretend they are, the science belongs to your guests and your team. What I can promise is that the edit itself does not introduce errors, right, and in my team that comes down to a few habits. We watch and listen to the whole episode before cutting, so we know which sentence carries the limitation. We never splice two half answers into one claim. We treat words like may, could, early, preliminary and in this sample as load bearing, so they stay. And when something sounds like it might be a slip, a number that seems off, a name that might be wrong, we flag it with a timestamp for your guest to check rather than quietly fixing it ourselves, because a well meaning fix by someone outside the field can be worse than the original slip.
If your show touches health or medicine, the review step matters even more. That is general information from the editing side and not medical advice, and anything a listener might act on should be checked by the qualified people on your team before it goes out.
My editors are not there to decide what is true, they are there to make sure the edit never says more than the scientist did.
How do you keep jargon in check without dumbing the science down?
You keep the technical term, make sure it gets explained in plain words the first time it appears, and use on screen text to reinforce it, so experts hear the right vocabulary and newcomers are never left behind.
The best science hosts already do half of this work in the conversation, right, they stop a guest and ask what a term means, and those follow up questions are gold in the edit, so we protect them even when they slow the pace a little. Where a guest uses a term that never gets explained, we look for the clearest moment later in the conversation where they describe it, and sometimes a short on screen definition card the first time the word appears does the job without interrupting anyone. We do not rewrite the science in captions or simplify it on screen in ways the guest did not say, the text always reflects what was actually spoken.
Captions do a lot of work for science shows, because specialist names, species, compounds and units are easy for automatic captions to mangle, and a caption that spells a molecule wrong undermines the whole episode for exactly the viewers you most want to impress. We check captions by hand against the audio, and the W3C's guidance on captions and media accessibility is a good reference for why accurate captions matter for deaf and hard of hearing viewers as well as for everyone watching on mute. You can read how we approach it on the page about captions and subtitles that match what was said.
If you have ever seen your own lab's work squeezed into a headline, you already know why I care so much about this part.
What visuals actually help explain a concept on a podcast?
Simple ones that the guest could sketch on a napkin, a labelled diagram, a scale comparison, a timeline or a figure the guest owns or is allowed to share, placed exactly when the concept is being explained and held long enough to read.
A podcast is a conversation, right, so I want visuals to support the explanation rather than turn the episode into a lecture. The ones that work best in our experience are small and specific, a cross section when someone describes a structure, a basic chart when someone talks about a trend, a map when location matters, and the guest's own photos from the field or the lab. The cutaway principles are in the guide on podcast B-roll editing, and the extra rule for science shows is that every visual has to be accurate, so we build diagrams from material you provide or approve, not from a quick image search.
Figures from published papers need a quick rights check before they go in, because the licence depends on the journal and the paper, and some are openly licensed with attribution while others are not. Your guest or your communications office usually knows the answer in a minute. To be very honest, the simplest route is often a clean redrawn version of the idea using the guest's own data, which we can design as a graphic once they confirm it says what they mean.
How should a science show handle short clips?
Carefully, because a clip strips away the context around a statement, so every clip has to contain its own caveat and should be approved by the guest before it is posted.
Short clips are how most people will meet your show, right, and I see them as where nuance goes to die if nobody is careful. The catch here is that the most clippable sentence in a science episode is almost always the most surprising one, and the most surprising sentence is usually the one that needs its qualifier the most. So when we cut clips for science shows, we start the clip early enough to include the context, we write on screen text that stays inside what the guest actually said, and we send the clips to you for approval alongside the episode. The broader thinking on cutting clips that work is on the shorts and clips page, but for a science show the guest's approval is the step that matters most.
I also like to agree a simple rule with the host before the first clip goes out, which is that no clip states a finding as settled unless the guest said it was settled in those words, and no clip text uses a stronger verb than the one in the audio. It sounds fussy, but it means I can let my editors move quickly on clips without anyone on your side worrying that a caption somewhere is promising a cure or a breakthrough the research never claimed.
What happens when a mistake is found after the episode is out?
You correct it openly, with a corrected cut where possible, a correction note in the description and show notes, and an on screen correction card if the error changes the meaning, because a visible correction protects your credibility far more than a silent one.
Every science show eventually finds something it wishes it had said differently, right, a figure misremembered, a study described slightly wrong, a new result that changes the picture. When that happens on a show we edit, we can prepare a corrected version of the episode, add a short card at the right moment that says what was wrong and what is correct, and update the clips that carry the same claim. Many podcast hosting platforms let you replace an episode's audio file, so the audio feed can carry the fix too, and the show notes are the natural place for a dated correction line. The page on show notes and chapters covers how we write them, and a correction line there is a small thing that tells your audience you take accuracy seriously.
Who is this kind of edit right for, and who should look elsewhere?
It suits researchers, labs, institutes, museums and independent science communicators who publish regularly and want a human team that protects nuance, while a one off recording or a show that needs on site filming may be better served by a university media office or a local crew.
A lot of science shows come out of institutions, right, and I will be straight with you, if your university already has a media team with editors on staff, using them may be cheaper and keeps everything inside the institution's review process, so it is worth asking before you hire anyone, and the post on podcast editing for universities and colleges goes into how those shows are usually organised. If you are an independent communicator or a lab that publishes every week or two and does not have that support, Pixel Samy Podcast is built for exactly that rhythm. We have also written about specific science communities, like the Tucson post on researchers and astronomers, which covers the kinds of shows that come out of a research city.
What budget does a science show need per episode?
A single camera science conversation of up to an hour on our Starter plan is 199 dollars an episode, with a full human edit, audio cleanup, intro and outro, and the first cut back within 48 hours of your upload.
If your show is shot on two to four cameras, or you want custom diagrams and graphics in most episodes, the Studio plan is the better fit, and every option, including monthly plans and episode packs, is on the pricing section. Episodes with heavy graphics can take a day longer, and we let you know that before starting so the guest review fits around it. The project files and everything we export belong to you, which matters for science shows more than most, because an institution may need to archive the material or hand it to another team years later.
Precision is the product
A science communication show is only as valuable as its trustworthiness, and that trust is built or lost in hundreds of small editing decisions that nobody in the audience will ever notice when they are done well, the caveat that stayed, the term that got explained, the caption spelled correctly, the clip that included its context, the correction made in public. If you send me a recent episode and a clip you were nervous about posting, I will show you how we would have handled both.
So yeah. That's my way of saying it.