It’s almost Christmas, and I just filled up this year’s two-terabyte external hard drive. Or rather, I nearly filled it. A little quick maintenance and I got the space issue sorted out. The drive will get me through the rest of the year, and then it is on to the next one.
I’ve never filled an entire hard drive before, but for the past couple of years, I have been very blessed to have a steady and diverse array of client work—including some video projects, which I suspect to be the things that put me close to maxing out the memory space on the drive.
In this post, I want to talk a little bit about the methods I use to manage the process of going from raw files to finished projects with a priority on sanity and family time. I am not trying to suggest that I have a magic formula for doing this, but rather that I have an evolving, organic process that helps me get things done while also maintaining full-time work and a family.
Some of this may work for you, and some may give you ideas about how to put things together that will work for you.
For purposes of clarity, I’m going to focus on my audio workflow. If folks are interested, I’d be happy to do a similar post on show-running, and on writing. Just let me know if you’d like to see that in the comments.
My hard drive’s top level is arranged by clients (for them) and projects (for me). That is, if it’s something I’m running, like Things Not Seen or The Francis Effect podcast, that I run in-house, that’s the name I use. Otherwise, I have a standard client reference (in the business we call this a ‘slug’) that governs all storage, whether on the external drive, or a DropBox shared drive or a local drive. That slug also gets used on physical file folders, if I have those related to a project.
This means that I have had to make some decisions about nomenclature. For example, over the last couple of years, I have worked on a video project with my friend Rahuldeep Gill. The project has gone through a couple of working titles—so I had to decide early on, am I filing everything by his first name, his last name, or one of the working titles? (I ended up using his last name).
This might seem like needless overthinking, but the nightmare scenario is when you have critical files scattered across different folders that are named differently but all refer to the same cluster of projects, and you can’t remember which name you happened to be thinking of that day you filed it. At one point early on I did that, and it about drove me crazy. Choose your top-level slugs, and stick with them.
If a project has repetition, or even if it might conceivably have repetition, design your next level down from the top-level for repetition. This means that you need to take a bit of time to think through how you are going to make each file at this next level to be a uniform iteration. Is the production episodic? Figure out your episode naming structure.
(For my shows, it’s often year+episode week, so this year’s weekly TNS episodes are numbered 2001, 2002 etc up through 2052, while my client work is often Ep1, Ep2 etc.)
In the same way that a top-level has a standard slug that is used everywhere, I try to use the episodic naming conventions consistently. That means that the episode number in my production files will match the published episode number that faces the audiences. It just makes everything easier for those times when I need to go back and adjust something months later (which happens).
This is a good time to say this: I have a mindset when I am in the thick of working on something, and everything makes sense in that context. But I find that as soon as I am done with a project, I flush the flow and context out of my brain. That means if I have to go back at some later time to get under the hood and work on it again, I can’t access any of that linearity of flow or context that I had. So my only hope is to structure that context as much as I can into the physical placement and flow there in the file structures.
That leads us to the practice where I have a standard set of folders that I use in every project. Every project has a file labeled _1001. The underscore makes the file structure throw it right up at the top, so it floats above anything else I am working on in the project. In that file, in each project, I have anything that I know I am going to be accessing with regularity whenever I work on new iterations within the project.
So, for example, in every project, I have a file named _1001_FOLDERS-TO_COPY. In that file, I keep a set of empty folders, usually with the following structure:
_ADOBE
_EDITED-AUDIO
_IMAGES
_NOTES
_OUTPUTS
_RAW-AUDIO
_SUBMIXES
_TEXTUAL
Every time I am working on a new iteration of the project (a new episode, for example) I go to the _1001 folder and copy the empty file folders and paste them into the project.
So a sample workflow would go like this.
First, I get audio from a client or record it myself. Often, that means multiple files of individual tracks or one big multichannel track. No matter what the particulars, this goes into _RAW-AUDIO and it stays there in its original form like that, forever.
No matter how the audio arrives, the task is to get it into a form that I can use. That means that I have each voice as isolated (aka ISO’d) into an individual channel file (as much as possible). Having each voice isolated gives you as much control as possible over making the overall sound clear and strong.
Once I have all the voices ISO’d, I open up Adobe Audition and create a submix. If there are two voices, this becomes a stereo track with one ISO panned full left and the other panned full right. If it is three or more voices, I set it up in Audition as a POLYWAV file.
In all cases, the goal of the submix is to have all the ISO tracks ganged together, so that I can be editing all the voices in real-time with each other, while also having individual control over the sound of each track.
The Adobe .sesx session gets saved in the _ADOBE folder, and the submix output itself gets saved in the _SUBMIX folder with a _SUBMIX suffix on the filename.
I then close out the Adobe session, and open a new blank session, into which I open the submix file I just made. I then immediately re-save that file into the _EDITS folder, with an _EDIT suffix added to the filename (e.g., filename_SUBMIX_EDIT).
What should be becoming clear is that I am creating a workflow where, at each point, I can protect myself against catastrophic failure. If something goes wrong at a certain stage of the process, I can revert to the earlier stage of the process and work my way forward again.
The _EDIT file gets adjusted in a number of ways. The stereo and polywav setups allow me to go in channel by channel and perform fine-grained cleanups on each track (scrubbing out noise, pulling out filler words, tightening up the flow) while also maintaining a common time flow for all the tracks.
Once I have the ganged tracks all cleaned up, I then split the ganged track back out into individual tracks, and save each individually in the _EDITS folder.
Now, if I’m building an episode, I actually start that process. I open a new Adobe .sesx file (sometimes from scratch, often using a template I have set up for the particular show) and save the session in the _ADOBE folder.
I then import the individual files I saved in the _EDITS folder and bring them into the right places in the session.
This gives me the best of both worlds—I have the files all synched properly (so lining them up at their starting points preserves the time flow) but I now can apply individual processing (noise reduction, EQ, compression, etc) as needed to each individual track.
Once I have things the way I want them, I will render the session outputs into the _OUTPUTS folder. If I am working with a client who needs to approve drafts and finals, I often add some version control nomenclature to the draft filenames (I have a step process that works up sequentially from v1, with v5 being the approved final).
I know a lot of folks are eager to find a one-shot solution to audio. That is, they leave all the audio mostly in the form they found it, and just apply a lot of brute processing. I have never been satisfied with that approach.
That said, I know a lot of editors may look at this and find my process needlessly fussy. Nevertheless, I have heard a difference in the quality of audio I can produce as I have adopted more of these processes.
I’m mentioning all this not to suggest that folks copy my workflow, but to encourage you to think deeply about your own workflow. If nothing else, I think the following goals are good ones to which to aspire:
It needs to be fast, flexible, and iterable - If you have to spend a lot of time thinking about where to put something, that steals brain power from creative or client work. Your time is better spent on cooler things than deciding on a file name each time you open a new project.
It needs to be as non-destructive as possible - Mistakes happen. I do my best to make sure that I am saving often, and that the saved file is going to the right folder at the right time. That way, if something blows up with my edit, I can go back to the submix and have that be in the same state I left it before the edits began.
It needs to save non-removable processes for close to the end - audio processing is a delicate process, and it can save the sound or it can introduce a whole new raft of artifacts. Do your best to leave the irreversible stuff to the very end, as close to the final outputs as possible. This is different from things like by-hand noise removal—get that stuff out of your files as early in the process as you can.
You need to be able to go back to the session later and not be lost—It’s a beautiful thing when a client wants to go back and revise some audio (and pay you again), or when it’s the end of the year and you are putting together a best-of clip show. Whatever the reason, if you are going back to earlier work, you want it to be in a familiar state. Keeping your work processes consistent increases the chances you will be able to recreate the session with the least amount of effort and time. Again, the less time you have to spend worrying about where you put things, the more time you can be spending on the work.
Creativity is often a lot of details. Those details can either work for you, or work against you. Creating a workflow takes time and thought on the front end, and it requires you to reassess aspects as you go.
For me, for example, these two weeks before the end of the year become the time I set up the new workflows for the coming year. I prep my external drives and revise the file structures and incorporate other lessons learned over the previous year.
I wish I was as good about this with my writing projects as I am with my audio work, so I also want to emphasize that wherever you are in this process, start there. Don’t wait to get things perfect. Do what you can, and revise as you go. But starting the process pays dividends almost immediately.
I know this post has been an information dump. Thanks for sticking with me. If you have specific questions, please ask them in the comments below, and I will be glad to talk more about these ideas generally or offer more specifics from my own experience.
Also, I’d love to hear some details about your own workflows, and how they work for you. Thank you for reading, and courage!
If you like these posts, please share them with others who might benefit. Also, if the post made you think about something, please offer a comment. Finally, all this content is free, but you are also welcome to support it if you find it valuable. All those options are available through the buttons below. Thank you again for reading!
Really interesting stuff, David, and I appreciate the time you took to explain it. If you decide to do something similar related to your workflow for all things written, I'd love to read it! This year has been a lot of trial and error for me (meaning, I would love to learn from you).