The first thing I should write is, if you’re looking for good technical (i.e. computing, equipment, software) advice on setting up your podcast, then oh dear god, go somewhere else. This post is going to be about the lessons I have learned along the way whilst making en clair, and if any lesson has been felt most completely, it is that my knowledge of this field in general is paper thin. I’ll go through in a sort of generally chronological five-step how-to order since this tracks along with the way that my wisdom has deepened (ha!) with time, so let’s start with, well, starting…
The first steps are arguably the hardest. Before anything else, you need a good and sustainable idea, and to know how long you’re going to pursue this idea, and how deeply. At the risk of preempting the Format subsection below, there are roughly three kinds of podcast:
- The infinity, or episodic podcast: This tends to be a topic-to-an-episode affair. It might be informed by, e.g. current affairs like Brexit, or it might cover such a massive subject that it could just spin on into eternity. See, for instance, the History of English podcast that started in 2012 and has passed 120 episodes. See also, Stuff You Should Know, that started in 2008 and is past 300 episodes. Many news, politics, culture, and true-crime podcasts fall into this camp. So too does en clair. If you like this sort of format, you might want to start a document somewhere, and add new ideas to it as they occur to you. Otherwise you’re at risk of abruptly running out of material one day. The en clair future episodes spreadsheet currently has over 100 ideas on it, at different levels of development, most simply at the “this would be fun!” stage. Not all are equally as viable, and sometimes several get combined to form one episode (e.g. Messages in Music), but in short I won’t be at a loss this time next year for more content. The bonus of episodic podcasts is you only need enough material for one more episode to keep going. However, this can lead to podcasts meandering off into the hinterlands of irrelevant or lower-quality content as they start to struggle for ideas. This is quite different to…
- The finite, or serial podcast: This might be a very long story with a deep-dive into minute details, but that ultimately has a clear and intended end point. In other words, it’s very like an audio-book. It might cover one “book” per series, like, for instance, Serial or Slow Burn, or it might be dedicated entirely to one story, like Body on the Moor and Death in Ice Valley and 13 Minutes to the Moon. Plenty of people really like these “closed” types of podcasts, because, unlike episodics, there is a point where you’re done and can move onto something new. In some respects, the serial podcast is harder than the episodic one, because you have to know to a reasonable extent how the whole thing is going to run, from start to finish, so that you can divide up the episodes relatively equally to cover about the same amount of time or ground in each “chapter” until the thing is done. But, a bonus is that you know where you’re going, you know what your topic is, and there is no real risk of running out of ideas.
- The indefinite, indeterminate, who knows wtf is happening podcast: Sometimes people put out podcasts that are not clearly either episodic or serialisations. Instead they just seem to go wherever the breeze takes them. That can work, but I’ve not been a fan of the few I’ve run into and have tended to unsubscribe an episode or two in.
I’ll say more on all of this in the Format and Schedule sections below, but whatever type you go for, it will have an impact on other decisions that you make in both the long and short-term.
Once you have a topic and a general idea of what overall type of podcast you’re going for, you need a catchy title that encapsulates the whole thing. (I can do this. Literally I have Twitter threads dedicated to coming up with fun titles for podcasts I’m never going to record and books I’m never going to write. If there is a prize to be had for naming stuff – internships, conferences, bikes, whatever – I could be in with a shot of winning it.) But seriously, you need a good title because people judge pods by their names, and the percentage that will go on to read the description appears to be vanishingly small. This is Tindr for sound. Grab listeners fast or they go swiping off over the horizon.
Small aside: given all I just said about being so good at naming things, I like the name en clair very much, but in hindsight, and rather ironically given its meaning, it’s not a transparent title. I figured this out only minutes after committing myself to it when I realised that anyone searching for my podcast with terms like “forensic linguistics” would not find it, or at least, it wouldn’t rank high in the results. (The podcast name is usually the primary search field, followed by descriptions and episode titles.) To correct the error later, I had to add a lengthy, explanatory subtitle. Now my podcast is officially entitled en clair: forensic linguistics, literary detection, language mysteries, and more… In yet further hindsight, I don’t know what I would have called it instead, and I’ve decided to be happy with my choice, but still. It’s something to think about very seriously before you click “submit”.
The next headache is graphics. If the title is not, or cannot be explanatory, your cover art had better take up the slack. And those graphics had better withstand the indignity of being shrunk to the size of a postage stamp on someone’s smartphone, or blown up to a gigantic full-screen image on a large desktop computer. The podcast name should arguably be in there, somewhere, legible and clean, just in case the text-version of the title has that second scrolled off the page to make way for an episode name exactly as someone is contemplating your show.
So how to make good cover artwork? Well, given the labyrinthine horror that is image copyright, and the fact that breaches of it can get your pod pulled faster than a dog can scoff a doughnut, the safest and cheapest option is to create something of your own. The best but costliest method, by contrast, is to pay someone to make something for you. Whatever the case, ensure you follow the podcast submission guidelines that Apple lay out about size, resolution, and shape. Why bother specifically with Apple’s requirements? Because they are the major player in the podcast field. If they won’t list your podcast in their catalogue, it might as well not exist. And if you fulfill their demands, then you fulfill pretty much everyone else’s too.
(Notice we haven’t even got near recording anything yet? That’s not going to happen for a while, so buckle in.)
You also need to come up with a tidy little description for the podcast in general. The blurb, as it were. An ideal metric is: can you fit it, or most of it, in a tweet? If yes, you’re probably on the right lines. The preview window on most devices is extremely short, so if your podcast isn’t selling itself within the first half-sentence, too late. Your would-be subscriber has found another show to spend their time on. Each episode will also need its own sensible description and that, too, should be short. The first five or ten words are key, so don’t waste space on fluff like, “In this episode we will discuss…” Just get stuck in. “Car chases! Murder!! Infamy!!!” Or, well, whatever your episode is about.
Next you need an episode-level format, or template, for each individual show. My podcast is a one-person stand-and-deliver, and I aim for 25-45 minutes. Longer than that, and I split into a multi-part miniseries. I know that I talk at an average rate of about 140 words per minute (you might want to figure out your own rate), so I aim to produce transcripts that don’t exceed 6,300 words in length. Within that allowance, I present the first half of a case with some jazzy music to give it atmosphere, I go into linguistic analysis with some meditative background ambiance in the middle, and then we conclude with the end of the case and any final remarks. Interleaved in all the appropriate places are an intro warning if required, an episode number and name, a short “welcome to this podcast” segment just in case the audio comes adrift from its source, and a closing acknowledgements section. Which no on listens to. Know this, and don’t stick anything vital there.
Other podcasts work on an interactive discussion or interview format, but you’ll still need some basics like an episode title and/or transitions between an overview of today’s subject matter, an introduction to your guests or co-presenters, and then a launch into the actual discussion. Some people have multi-section podcasts – part stand-and-deliver, part book review, part interview, part discussion, part clip from the news, part listener questions and comments, part something else… The crucial thing here is that the more segments you have, and the longer your episodes, the more audio engineering you’ll likely be doing later (see Editing below).
Luckily not a problem I have to deal with, but if you decide on a multi-person podcast – whether with guests or with co-presenters – you’re probably going to need to work out a system of hand-signs if you’re recording in person, or a method of managing the conversation if you’re recording from different locations. Natural chatter is all very nice but too much confusing overlap is likely to frustrate the listener. You’re also going to need to organise a lot of schedules well in advance of any publication day (see Schedule below), and have a system for how to cope if, say, your primary presenter is off. Or if you’re even going to have a primary presenter. And quite how you’re going to manage the tech and sharing audio and division of duties I do not know. My stress levels just thinking about this are off the charts. You can keep your interactive podcasts and I’ll stick with my just-me affair.
Anyway, finally, you need a podcast-level schedule. In a perfect world, especially for serialisations, you would record the whole thing, or a whole season, or at least a bunch of episodes well in advance, and then release them once per week/fortnight/month, and as that series is airing, you’d be working on the next episodes/series/podcasts. I did this in that I got four episodes lined up and airing, whilst I started work on the fifth onwards. Except somehow I blinked and there were three days left till the fifth episode was meant to be out. Moral of the story: I am an over-ambitious idiot. Don’t be like me. I even chose a monthly release date to keep the demands of the thing down to a rational amount and it sneaks up insanely quickly every single time. I am audio-engineering S01E10 at the last second right now and I do not recommend it. At all.
Whatever your preference, it’s sensible to have a release schedule that you stick to. Listeners fit their podcasts in around their lives. I’ve had listeners tell me that they save new en clair episodes for their Saturday morning walk, or their Monday morning commute, or some other specific part of their life, and they expect it to arrive on a certain date. Of course, you can go with an ad hoc schedule of posting whenever you like. No one can force you to do otherwise. But that will likely affect your listenership.
You also need to bear in mind technical limitations. Lots of podcast apps will only download and hold three unlistened to episodes before kicking the podcast over into dormant mode, on the assumption that the listener has gotten bored of it, so if you suddenly spam out five episodes in a week and people don’t expect it and can’t keep up, the app will stop downloading after the third, and stop giving alerts. There’s a risk that the listener won’t realise more episodes have happened, and you might accidentally go entirely off their radar.
Also remember that less can absolutely be more. Plenty of people do not have infinite amounts of free time to listen to three two-hour episodes a week. I have unsubscribed from several weekly podcasts that “upgraded” to daily episodes simply because I could not keep up and kept falling hopelessly behind. In another case, they kept their weekly big episode and added shorter, daily episodes for Patreon subscribers, but then I found that all the daily material was being rehashed in the weekly episode anyway. With the same jokes. 😒
In short, pace yourself, give your listeners a clear schedule along with any breaks, and then live up to it as best as you can. Try to stay ahead enough that you aren’t scrambling to record, edit, publish, and promote an episode with only hours to go before the intended release date. Better yet, stay enough ahead that you can have a holiday without worrying about an impending episode, and return with time enough to record your next series or set of episodes in advance of the current ones all being aired.
Okay, okay, can we finally record some stuff now?
Yes! Finally, at long last, after creating a theme, a name, some artwork, a format, and a schedule, we can start recording! YAY! No wait… do you even mic tho?
In podcasting, as in all kinds of creative production, prevention is infinitely better than cure. Or essentially, lightly polishing a good recording will produce far better results than heavily polishing a bad recording. It is worth the investment of a little time, money, and effort at the start of the podcast and each episode. This will save you having to waste a heck of a lot more time, money, and effort afterwards fixing all the problems that you could have avoided.
Your shortlist of requirements are: good hardware and software that is properly set up, a good recording environment, and good recording hygiene (be still or move about silently, keep a consistent distance from the mic, etc.). After all, would you rather spend an hour setting up, followed by an hour of editing? Or a minute setting up, followed by ten hours of editing? Yeah. Thought so. Whatever the case, some polishing will be required, but we’ll get back to that in Editing, below. Mainly, remember that these sorts of efforts will also keep you all the listeners who would otherwise be annoyed by, and unsubscribe on the basis of poor audio. I am just such a person. I give each podcast two or three episodes to fix their sound, and then I’m out.
You might ask, “What’s decent equipment?” Well, there is a ton of stuff out there but my knowledge extends thus far: you can use something like a Zoom-type device (I have a H4N that I’m just starting to try out for a new podcast series) that is, in my simplistic little world, a glorified dictaphone. No need for computers, and highly portable, so excellent for interviews, especially at conferences. Or, as in my case, you could use something like a RODE-NT-USB mic with a homemade desktop “studio” that records directly to Audacity on my desktop computer in my office during quiet times. You could even try a smartphone. In the right environment, it could well work. It’s not something I’ve gone for but others have and they seem to have done quite well with it. Whatever the case, before actually starting to record and publish episodes for real, try out different formats and settings. You might find that some recordings sound great on the device and then surprisingly terrible through headphones, or vice versa. This last is the most crucial issue, by the way. You MUST try your proposed recording method out on (good) headphones. This is the way almost everyone will listen, and if it sounds terrible through headphones, it is dead in the water.
Once you’ve figured out your general sound, go record some shit! It’s fun and you’ll love it and very soon you will realise that the actual recording is by far the easiest part of the whole affair. Don’t believe me? I come back to this in the second Golden Rule, right at the end.
In my case, and for my format of podcast, I write transcripts and largely read them as they stand, but sometimes elaborate where it feels right. For discussion podcasts you might want to have key talking points. For interviews you might have a list of questions that your interviewee has had a chance to read and prepare for, and then go through them as naturally as possible. Really it’s up to you. There’re a few things to be aware of however, that no amount of equipment and environment control can entirely prevent.
Your mouth will make weird noises. This will be especially pronounced if you have a good mic that picks up sound properly. Also you will probably make intrusive plosives. (That’s now the name of my next podcast. See? I can name… a n y t h i n g.) And you will inadvertently tap the desk or move or jig the table or brush the mic with your clothes. And you might talk too fast, or too long, or too loudly or quietly or incoherently. And you will make mistakes. And then your phone will ping. And then someone will loudly whack your office door on the way past with their bag. So here are some of the things I’ve learned.
Eating right before recording – especially sugary foods – is the worst idea. It makes your mouth extra noisy. Don’t do it. It’s going to be right in people’s ears. If you can’t stand those kinds of sounds (I really can’t) you can painstakingly edit the ever living shit out of your audio recording and kill all the little clicks and pops and crackles the way I do, and that way, madness lies. But it’s an option.
The easier option is to get the perfect balance between closeness to the mic and not eating sticky food beforehand, though strangely, some people recommend apple as a mouth-noise deadener. I don’t know that it works but it’s worth a try.
Also, sit still, or if you must move about, have your sleeves rolled up and loose clothing firmly tethered. Material brushing over a pop-guard will likely be inaudible to you, but it’s going to sound like a chainsaw in playback. Similarly, the bang of a coffee cup slamming down next to a mic will tear your ears clean off, so spare yourself that bloodshed. Put a cloth over your desk to stop the inevitable taps and clicks, make yourself a desktop studio to extra-kill the background noises (see my fabulous homemade affair below), don’t wear hand or wrist jewellery or have anything clinky like keys nearby, and try not to swirl back and forth in your chair or restlessly jiggle your knees or do anything else that produces tiny sounds that don’t really register to you, but that the mic faithfully and loudly records all over the top of your lovely episode.
Firstly, make sure you have a good pop-filter. That will help, but it won’t necessarily catch every plosive. After that, there’s something called a high-pass filter. You need to find out what your lowest pitch is (mine appears to be around 130 hertz) and then run the high-pass filter over the audio. This will kill off the worst of the plosives so that they don’t kill off your listeners.
Talking too X
If you know that you talk too fast, have signs in front of your face somewhere that say “SLOW DOWN!” and “ENUNCIATE!” and, if required, “BREATHE!” I do. Sometimes it even works. If you’re prone to dominate the conversation, maybe have a sign saying “SHUT UP!”, and if you tend to be too quiet, have one that says “SPEAK!” Whatever works, really.
Essentially, try not to forget your listener. If the topic is meant to be soothing, let the pace match, and if it’s meant to be tense, go with it, but if your podcast is a straight 90 minutes of breathless, rapid-fire, intense debate, your listener might be cognitively fried by the end. Assuming they even get that far. Think about ways to vary up the pace and texture and provide mental “breaks”. (See Format above for possible segment ideas.)
If you goof up or there’s an interruption and you want to restart that bit, do a loud double or triple click with your tongue close to the mic, or tap it twice with a pen, or whatever. It’ll put a big spike in the audio so that, visually, it’s easy to jump to and edit. I sometimes end up restarting three or four times in a row and I take those opportunities to take a deep breath, clear my throat, have a drink of water, and re-read my “SLOW DOWN!” sign.
Clicks are also useful to signal insertions, segment changes, chapter markers, the start or end of audio, and so on. This is because, if nothing else in the life of a podcaster, you are going to be doing some audio-engineering. Why? Well…
No matter how excellent your recording environment, unless you literally never make a mistake, or misspeak, or cough, or get interrupted by random noises, there will inevitably be some sound engineering to be done, even if it’s only to add some dead noise at the start and end. And I suspect you will be doing way more than just that.
All this said, however, aside from what little I’ve already explained above about high-pass filters and so on, I can give almost no advice in this section. One option, and an expensive one, is to throw money at this problem also, and hire in some sound engineering help, but this will bleed your bank account dry, and having done my own “audio-engineering” now for almost a year, I know why they charge so much. It can be a technical, demanding, and arduous job. It can also be very superficial and light, but it depends on how cleanly you record in the first place, and how intensively you want to pretty up your audio.
Over painstaking weeks of getting thoroughly acquainted with Audacity, I have learned how to apply various effects (e.g. I use reverb for quotes), how to amplify quiet bits and quieten down noisy ones, how to seamlessly trim out pops and crackles, how to have multiple tracks work together so that I can have the talking and music going on at the right place and volume, how to fade stuff in and out, and so on. For proper advice on this there are worlds of excellent manuals and guides out there which I have picked through as I’ve required each new skill. If you really want to do this long term and have no budget for hiring someone to edit the audio for you, I suggest you bite the bullet and get learning. Audacity is free, intuitive, and the three or four major skills you need – deleting mistakes, adding in silence, moving audio around, fading in and out – are very easy to pick up.
Whatever you do to create the finished sound, there is a general consensus that it’s wise to save your final product in a high quality format like .wav. You may go on to upload an .mp3 to the podcast hosting site later, but for your own records and files, it’s good to have a better one available that lower quality ones can be copied from. I save the Audacity .aup file so that I can always go back and make further changes if needed. Once finished, I export a .wav using all the standard settings in Audacity, with all the metadata fields filled in properly, consistently, and carefully (these are vital so don’t gloss over them), and that combination has worked out nicely so far.
Another thing I should add here is that there is a world of amazing free music out there. I mostly use Free Music Archive, BUT you have to be careful about the restrictions. Non-commercial academic ventures are largely okay but you have to be clear that there’s no commercial aspect to what you’re doing, such as advertising or Patreons or whatever. And some don’t even want you modifying their music at all. (I sometimes loop tracks to get the right amount of time out of them.) However, don’t underestimate how much time it takes to sample different tracks, find the right sounds and durations, interleave the music and talking to hit peak points at the right moments, and so on. One can lose a world of time here in making the audio perfectly fit the tone and tempo of the piece. It’s an effort I consider worthwhile and I think it pays off for en clair, but not all podcasts need this sort of dramatic emphasis in their lives, so contemplate just how much you really need to add music before going down this route. It’s a fairly stout time commitment on top of what is already a big job.
Finally, at least for your first handful of episodes, you might want to consider “peer-review”. It is essential to get the pace and tone and format and so on right as soon as you can, but with your nose stuck to the glass, as it were, it’s far too easy to lose perspective. This is an awkward one, though. The perfect reviewer will like you enough to do you the favour of listening to advance episodes of your podcast, they will be honest enough to tell you the issues with it (there will be issues; no podcast is perfect), and they will be insightful enough to suggest possible fixes. But you, in turn, must be professional enough to graciously accept whatever feedback they can give, and respectful of their time. Of course, you can simply chance it with The World At Large and hope for kindness and constructive insights from online strangers, but I can say that the episodes I had “peer-reviewed” by two separate, objective listeners, and that I then modified on the basis of their feedback, have gone on to be some of my highest rated shows.
One of your final decisions, and a big one that it’s difficult to get to grips with at first before you even know how successful your podcast is going to be, is to decide where you’re going to publish your podcast. You can host it on cloud servers like Box, but note that this can be an unreliable source for others to access or stream from, which can lose you a lot of potential listeners. You will also get no real information about download numbers or worldwide listeners, and it will make getting your podcast onto Apple’s iTunes catalogue very difficult without some serious tech help.
Another, easier option, and the one I’ve gone for, is to throw money at the problem by going with some sort of commercial service like SoundCloud or Blubrry or Podbean. Some charge by the quantity of audio you upload, some by the duration, some by the number of downloads. I chose SoundCloud because it’s a flat yearly fee for as much as I can be bothered to host, and I get good stats about listenership around the world.
They also provide a feed that is compatible for submitting to Apple and other catalogues, I can schedule episodes to appear without me having to manage it manually, and I can re-upload any episode even after its release without losing listener stats if I spot an error in the recording.
Once you have your feed, and you’ve tested and validated it (there are plenty of free podcast feed validators out there) your final task – aside from recording episodes, of course – is to submit your podcast to as many registries and catalogues as possible. The main one is Apple, noted above, but you should also consider Blubrry, Podbean, Google Play, Spotify, Stitcher, and plenty of others besides. You can find extensive, up-to-date lists of these online. Submission is usually very quick and its worth the effort to ensure that you’re showing up in as many different apps and on as many sites as possible.
After this, to really maximise your reach, you should consider having a blog and/or at least a Twitter feed to go with your podcast. Why do this? Well, sometimes you might want to share images, or documents, or other content alongside the episode in question, and it’s just nice to have a transcript so that the content is more widely accessible. Also you’ll probably need a place for acknowledgements, and it gives listeners the chance to leave comments and get in touch. For my purposes, an accompanying blog and Twitter account work together perfectly – the blog for the full transcript and acknowledgements and references and so on, the Twitter account to advertise when new episodes are afoot, get feedback, answer questions, and flag up related, amusing, or interesting stories in the media.
6. Golden rules
Overall, I’ve come up with three golden rules for podcasting that I think are the major takeaways of this whole post:
- You are not entitled to an audience, nor their time, and even successful podcasts take a while to establish. Don’t expect 10,000 downloads on day one. Or even year one. Expect, instead, to work at it for months and build up an audience slowly and to occasionally wonder if it’s even worth it. Remember also that as your audience does start to grow, these people are not faceless drones. They could include politicians, taxi drivers, CEOs, retail workers, brain surgeons, nurses, master criminals, petty thieves, judges, police officers, professors, students, and many besides. You just won’t know, but have it in mind when you turn out your next finished episode that it could end up being piped into the ears of a Prime Minister or a President. It could get aired on a major channel or feature in a breaking news story. It could land you a book or TV deal. The potential for all these things is out there, so just how good do you want it to sound now? See also 2. and 3.
- Doing this properly is going to take a lot of dedication, time, and energy. Plenty of people have been podcasting (and broadcasting more generally) for plenty of time now. If there was some method that was supremely easy and took almost no effort, it would have been discovered and we’d all be doing it. You are not going to chance upon that effort-free method because it does not exist. Similarly, you are very unlikely to possess that rarest of personalities that makes people delight in everything you do no matter what sort of meaningless dross you turn out. And if you think you do, you might want to take some online personality tests. Expect to pour a lot of blood, sweat, and tears into this if you’re going to make it work. A practical example: one episode of en clair at the moment takes around 50 hours to produce from start to finish. How does that time break down? Research: 12h. Transcript-writing: 12h. Recording: 1h. Editing: 24h. Publishing and promoting: 1h. Note how the recording bit is one of the shortest? If you do the maths, The Yorkshire Ripper five-part miniseries took an investment of 250 hours to create. That’s nearly seven full-time weeks. You can cut corners, of course – no music, minimal editing, no prep work, and so forth, but every decrease in quality will correlate with a decrease in listeners. Also, refer back to 1. (would you want a nation’s leader to listen to this and see it as a typical example of your work?) and then go on to 3.
- You can’t please all the listeners all the time. You won’t please some of them at all. If you start to build up any kind of audience, then inevitably, at some point, someone is going to whatabout your podcast. (“Why haven’t you talked about X yet? You should do Y! I INSIST THAT YOU DO Z!!”) Someone is going to take offense at something you say. (I once thanked the country of Hong Kong for listening. That made some people Very Angry.) Someone is going to tell you how much they hate your voice, accent, vocal fry, pitch, speed, mannerisms, name, face, attitude, whatever. You’re putting out public content for literally anyone with an internet connection, of any background or politics or motivation, anywhere in the world, to consume. They may even rate and review your podcast, and some will go as far as getting in touch with you directly, in public or in private, to tell you their thoughts and feelings. Some of the feedback will be positive, or at least constructive. Some of it will be rude, unfair, petty, personal, and vindictive. In short, there are assholes out there, at large, and some of them may decide to be assholes directly at you about this thing that you’ve laboured over long and hard. If, on balance, that prospect outweighs all the positives of doing a podcast, then you might want to rethink whether this is the project for you.
So, there you have it. The surprisingly enormous world of podcasting. Assuming this post hasn’t put you off, go forth, plan, prepare, ponder, and then finally, cast some pods and have fun!