Making the Serial Issue 1: Episode Production Workflow
In one week, I am launching the first 2 episodes of my audio book and podcast serial, The Dauntless Gambit! I will have 11 episodes completed by then and ready to start the long, 64 episode journey over the next 8 months, producing 2 episodes a week (in 4 parallel phases of production!).
It took a bit to figure how how, or if, this was even possible. Turns out, with enough planning and time investment, yes, it is. I had to hire a freelance editor and a separate freelance narrator to pull it off, but we’re up and running now. Soon, you’ll be able to read each new episode or listen to it on any podcast app of your choice.
The chapters/episodes of the serial vary in length, but I try to keep them in the 5000 word range since the narration speed ends up being around 30 minutes. I want people to get around an hour of new content a week, so this is all relative to what it takes to get someone a single installment of 5000 words or around 30 minutes of narration.
At the end of the post there’s a 5 minute preview of episode one of The Dauntless Gambit for you to hear what the end result sounds like!
This is what I actively did today, Sunday May 24th 2020. While The Dauntless Gambit was written all at once over 8 months, August 2018 to April 2019, it sat for a year in hibernation. Now, I take the chapters sequentially and produce them 4 at a time in staggered parallel, here is an example of what the cycle looks like on a day like today (Sunday, May 24th 2020)
- Chapter 10 was completely rewritten, this took 4–5 hours.
- Chapter 9 was tuned-up after having been line-edited and sent off for proofing. This took about an hour.
- Chapter 8 was reviewed for recording and sent in its final form to the narrator. That also took maybe an hour.
- Chapter 7 was listened to and any audio fixes were noted and sent back to the narrator. This took a little less than an hour.
I keep 4 episodes in production. Now that I have sent chapter 10 down the pipeline, I will start in on chapter 11, and the narrator currently has chapter 8, and the editor has chapter 10 to edit and 9 to proof. Since there are 2 of them, they each have 2 episodes at a time, but since there is only one of me, I have 4. This is the only way to produce 2 episodes a week with a 14 day lead time. If I start the process on the 1st of the month, it is ready to be released to the public on the 14th.
Right now, my weeks look like this
- Sunday, I spend most of the day doing the entire workflow as above. Stage an episode for Monday release, rewrite an episode, listen to an episode, revise an episode, and proof an episode.
- Monday, the narrator and editor have these days to work. I don’t have to do anything for the workflow and can do other run-the-business stuff.
- Tuesday, the narrator and editor have these days to work. I don’t have to do anything for the workflow and can do other run-the-business stuff.
- Wednesday, I stage an episode and rewrite an episode. Since this is a weekday and I have a day job, I split some tasks to Thursday.
- Thursday I listen to an episode, proof an episode, and revise an episode, since it is also a day job weekday.
- Friday, the narrator and editor have these days to work. I don’t have to do anything for the workflow and can do other run-the-business stuff.
- Saturday, the narrator and editor have these days to work. I don’t have to do anything for the workflow and can do other run-the-business stuff.
You may be asking “Why not get ahead by doing more on the days you aren’t in the workflow?” Well, because I am also using those days to write an entirely separate book series and need that time to write. Monday, Tuesday, Friday, Saturday are “net new work” days for the other project which is planned for a traditionally-sized novel and audiobook release.
Now, let’s look at the process!
First draft: 7–10 hours
This is the oddball since the first draft of the entire story was already written. But, if I were try take the outline of a chapter and write it from scratch, I put out around 500–800 words an hour. That is very slow. I am a slow writer. I can sometimes go faster, but it’s just not how my brain works.
So, if I were to arrive at a 5000 word chapter, it would have taken me 7–10 hours of sitting and typing. Yeah… it is not fun. I am not a writer who gets a lot of pleasure out of this part. If I were an athlete and storytelling were my sport, the writing of the draft is my mindless hours on the treadmill doing cardio while the outlining, planning, revising, and producing into the final product are the actual sport I want to play. But, this is unavoidable and one of the few things I can’t outsource.
The entire first draft, 320,000 words, is all sitting in the Scrivener word processor like this:
Rewrite: 4–5 hours
As I produce the text and podcast episodes at the time of this writing (May 2020), I now take a chapter at a time and rewrite it and take it through the process with the editor and narrator.
This is now it goes: I print the first draft to a PDF and read it on my iPad with an Apple Pencil stylus and mark it up, like this:
Then, I look at the PDF and the original draft in Scrivener and copy/paste into a new file and rewrite as needed. Some things are good, some are bad, some are totally new. The before/after diff in Scrivener at this phase looks like this:
I then take this new rewrite and run it through the ProWritingAid app which is like spell check and grammar check on steroids, and I send it to the editor. This process is about twice as fast as the original writing, which includes the re-read.
I experimented with trying to do the edit right in Scrivener without the PDF prinout and iPad, but the experience of marking it up with the tablet is just so much easier to think about. My mind processes text on a computer and hands on a keyboard very different than text on a tablet and hand on a stylus.
The iPad Pro 12.9″ and Pencil 2 with PDF Expert — took about 3 years to purchase, but I’ll never go without it again.
Content and structural edit: 2–3 hours
The block of 5000 words then goes to an editor (Hannah at The Storian) who does a read, comment, developmental critique, and line-edit. This takes her around 2–3 hours. She is not only line-editing and moving bad text around, but also leaving me comments in Google Docs about things that don’t make sense, appear out of nowhere, lack context, or somehow were lost in the shuffle.
This is one of the most critical steps because I can’t do her work myself. Even if I were to try, the author’s brain can’t see what the editor sees, just like you can’t smell your own bad breath. Even after all the time I put into the drafting and rewriting, she pulls out things that make me groan and roll my eyes at myself. This is what takes something from “decent” to “pretty good.” Here is the same page now in a comparison diff after it was edited:
Also, for the sake of the process, Hannah is line-editing a chapter and also proofing a different chapter in parallel. The way this workflow works (as you’ll see at the bottom of the article), the editor and narrator each have 2 episodes a week in parallel, and I being alone on the creator side juggle 4: a draft, a revision, a proof, and a script. Twice a week.
Fix and revise: 1 hour
Once I get the edited chapter back, I read and reply to the comments (so she has context for the proofreading) and make the changes. Most of the time, this involves adding content for more context, or re-explaining something I did a poor job at the first time. A lot of this is like a script consultant on a movie set, she’ll tell me “how did this character suddenly pick up a gun?” or “this made it sound like the characters were already on their way to the ship, but then you say they were seated at a booth.”
This is where it starts to feel more real. All the mistakes are being fixed, context is being added, and the narrative really starts to flow smoothly. I add all the changes and send it back to her for a proofreading round (along with the next chapter that needs editing).
Proofreading: 1 hour.
The proofread goes by a lot quicker since the task is not to do more edits. My changes are validated for coherence, but the rest of the text is simply proofed for error, grammar, structure. This is something that could go on and on for infinity, but as I want to release this as a semi-weekly serial and podcast, we can’t let perfect be the enemy of good.
Here’s what this part of the page looks like after a proof:
Narration pre-read: 1–1.5 hours
If you thought I sound like a slow writer, I am even a slower reader. My brain is hard-wired to read at speaking speeds, so I clock-in at around 175–200 words per minute. I’ve tried to learn to go faster, but it really feels like something neurologically constrained.
So, when I get the proofread copy back, I print it to PDF again and read it on the iPad with the Pencil, reading it in the narrator’s voice in my head and looking for anything that slipped past proofreading, but also anything that just won’t sound good when narrated. Narration is a whole different thing. Not only does narrative feel different when spoken vs. read, it’s also something that A) I can’t fix myself since my voice isn’t the narrator’s, and B) is sorta permanent since I can’t just sneak in a text edit once an audio chapter is released like I can with the text pages.
I go over it slowly, sometimes twice, and then just grit my teeth and send the script over with a summary of the feel and tone of the episode. At this point, we’re on a working ranch, if it’s not perfect, at least it’s worth what people are paying for it!
What the narrator gets at this point is what is posted as the text chapter. There’s no real photo to post since it is just a clean, edited Google Doc at this point.
Narration: 1–2 hours.
The is the black box where I just clench my buttocks and wait. As someone who wants to control the entire toolchain, relinquishing control to the narrator is the hardest part.
Thankfully, the narrator Chris Lynch ( here’s his website) somehow peered into my mind and delivers the finished episodes with hardly any corrections needed. 5000 words is about 30 minutes of audio, and usually I have to submit only around 4–5 single-word fixes.
The narration is what makes this a serious product. We’re using all the same tools, hardware, and techniques as any major audio studio. Chris is an experienced narrator with a quality setup, and I have 20 years of digital audio engineering under my belt to produce the finished product (more on that below).
Narration quality assurance and fixes: 1 hour
This is by far the hardest part for me.
Listening back to the narration and reading along to check it for errors floods my body with adrenaline, I am completely on edge and overwhelmed with anxiety. I pull up the text in Scrivener and follow along in Logic, listening for each word as I read.
You know that feeling when you’re watching a child or friend do a performance and you’re just on the edge of your seat? Or you’re watching your favorite Olympic figure skater do a routing and every jump makes you hold your breath? That’s what it feels like. It’s such a strange reaction. I listen in 5 minute bursts and take a minute or 2 break. If a word comes out wrong, I mark down the time and make the note for the revision, and just plow ahead.
The text on top, and the audio file below, listening to every word in real time.
Typically the fixes are just single words, things like adding or removing a pluralization, or reading a word wrong that probably felt correct at the time. Usually I get the punched-in file with the corrects in just a few hours.
Episode processing: 0.5 hour
Thanks to a consistent source file, actually creating the episode is not a lot of work now that I have done the first one. The intro music and outro sounds are all the exact same, all the episodes are in the same Logic Project with each previous episode muted.
Each narration track goes to the same effects bus so the compression, limiting, noise processing, and equalization are all basically the same, and I just tweak what is needed, add my outro, and bounce it all to disk. I purchased music and the bumper sounds I use, so all I do now is drag and drop the narration, tweak it, watch the final output RMS levels, and it’s good to go.
Every episode lined up with the intro music, busses, and repeated intro narration as a single reusable track at the bottom in red.
This is the fun part, getting to finally see something that was once just a miserable task of slogging away at prose so many labor-hours ago now feel like a real-life, premium production.
All together now
Overall, each episode ends up being around 30 minutes of audio and 5000 words, or about 10 pages in a paperback. It takes around around 18–20 hours of labor, meaning the entire book will take around 1200 hours to produce a 30 hour audio story of 320,000 written words.
Of course, half of the upfront time has already been spent in the manuscript writing in 2018, I could never do this at this pace if I had to write the story. My personal weekly time investment per episode is around 10 hours per episode, so I spend 20 hours a week to deliver 2 episodes of the Dauntless Gambit to the readers/listeners, and the editor and narrator contribute the other 6 per episode, or 12 total.
I suppose the current weekly labor investment to deliver a pair of 30 minute audiobook podcast episodes is around 32 hours. Wow.
Want to hear the first 5 minutes of episode one right now, the very words used in the example photos? Here you go!
Hopefully you’re asking both how and why I would do this. Well, hopefully before that you’ve sampled some chapters/episodes and liked them and THEN asked how and why.
The “how” answer is: it’s just a total marathon slog. I work my day job, I tend to my family, and I do nothing else but work on being a fiction creator. The COVID-19 quarantine helps make that easy, but even if it weren’t a thing, this is what I’d be doing. Thankfully my spouse and kids support this quest, and hopefully one day it can contribute to the financial support as well. But even if it doesn’t, I am not going to stop, this is what keeps me going.
99% of people would hate doing this, and I would not blame them. Most of the time, I hate it too. But, there’s no other way to get to the end-result I want, which is to put mainstream-quality fiction products into the hands of people all over the world that they will enjoy. If I have to sacrifice rest, relaxation, recreation, friends, a social life, exercise, self-care, and sanity to do so, that’s worth it.
I spent a lot of years, decades really, wondering when I’d “have the time” to be an author and creator of high-effort fiction products like all the authors and storytellers I aspire to be like. It took me until I was 37 years old to find the answer: never. I will never have the time. I will never have the energy. I will never have the freedom to be an artist.
So instead, I take from all the scant resources I have around me and make of it something more and do it without time, energy, or freedom. It’s not a fun process, but it’s the only one I have. But, I don’t have to have fun making my creative products for people to have fun experiencing them, and that’s what makes it all worth it. And in the end, I think needing the process to feel good is what stops people from creating things.
Once I stopped needing it to feel good, it became so much easier. I’ve learned the most important lesson I could learn: I don’t need it to feel good, I need the result to make you feel good.
Check out the first results!
So head on over to the Dauntless Gambit! If you’re reading this after June 1st, 2020, then episodes are already posted. If not, go to the main page and subscribe to the email list and follow me on the podcast apps and Patreon!
Originally published at https://www.erikflowers.com.