Interning in the Creative Industry

Pushing Audio Restoration to its Limits

My first assignment at Boxland was to edit and mix The Great Tom Foolery, which was for Boxland’s Vox in the Box 102 Class. The students in this class spend a lot time learning about voice acting and how to develop a character vocally. To my knowledge, The Great Tom Foolery was a script the students read in studio for the purpose of being recorded and added to their portfolio. In general, the script, characters, and actors were all great. The only problem with this recording is that volume levels were all over the place. As well, all the dialog was recorded onto one stereo track.

For starters, there were some voice actors whose microphones were way too hot, which caused clipping every time they spoke. There were other actors whose microphones were way too soft, which made it extremely difficult to bring them up in volume without bringing up the noise floor. This is where Izotope’s RX Suite came in handy. While I was not able to fix everything in the recording, the De-Clip and Voice De-Noise plugins masked these mistakes as best as they could. One thing I noticed is that these plugins work very well with longer audio clips rather than short ones. For example, the character who was recorded at a low level had tons of noise in his dialog. On longer lines, I was able to remove most of the room noise from his dialog. However, when he performed one liners, the Voice De-Noise was not able to detect the noise fast enough and thus the noise is still audible. There were other times when the De-Clip could not interpret the audio as clipping because the source was too distorted, making it essentially a square wave. While these plugins are incredibly useful for small audio mistakes, they were never designed as a full production fail safe. Although when blended with some clever editing, they can turn a trashed recording into something usable.

Here you can listen to some of the audio completely unmixed, and then compare it to what was published online. While not perfect, it is far from where it was.

I think it is important to note that Boxland does not normally have audio recording issues like the ones here. In fact, during the same session the class recorded another show called Oregon Trailings, which had similar issues but not nearly as bad.

I was instructed to make Tom Foolery as cheesy as possible, the sillier the sound effects, the better. While I was familiar with sound designing more serious scripts, I had never done something that needed to be funny. Let alone something where my sound design was going to make or break the comedic factor of the show. The first thing I did was find sound effects for cues that were already in the script. As well as music that fits the scenes portrayed in the story. After that, I thought the show needed more SFX than the script called for. In my opinion, adding more music and SFX masked most of the audio errors mentioned in my last blog post.

I’m not too sure how to describe my methods of making Tom Foolery sound funny. If anything, I leaned into how ironic, and ego-driven the characters are. For example, Tom Foolery narrates his inner monologues out loud for all to hear. When giving these monologues, Foolery’s dialogue is encased with a dreamy reverb that is usually quickly cut off when someone brings him back into reality. The most ridiculous thing I had to edit was the “death” of an information broker who also happens to be a labradoodle. While this is supposed to be a sad moment in the story, the absurdity still needed to be connived. The music leading up to this event is very cartoonish and goofy, and then when the dog gets assassinated, it immediately shifts into what I pictured as a multi-angle slow-motion recap of him getting shot. I did this using a slap-delay with various feedback settings.

I’d highly recommend listening to Tom Foolery here, as well as Oregon Trailings, which is the script my co-intern and friend Cameron Howard edited.

For starters, I was in charge of finding us recording gear within a reasonable budget. I decided to go with two Shure 58s because they were cheap, sound great when eq’d, and would work for Femergy live events. I chose Rodes’ Rodecaster Pro mix board as our interface because it was cheaper and allowed up to four XLR inputs and a few specialty inputs like 3.5mm aux and Bluetooth. Finally, I grabbed us a pair of desk mic stands, (I’ve come to regret getting these over traditional mic stands, but they still work fine most of the time). For editing I use Logic Pro, and up until a few months ago I used all stock plugins to edit and mix. With enough time and practice, you can tune any stock plugin to sound the way you want it. I’d highly recommend any of this gear and software to anyone looking to start a professional sounding podcast on a budget.

I was also tasked with making a whole website for the show. This took me a few days to design but it came out great in the end. I don’t remember what service we used originally but the site has been condensed since then and moved to the main Femergy site here. To this day we still use Anchor (Now called Spotify for Podcaster) and YouTube to publish the show for free. Anchor has been super reliable at getting the show to Spotify, Apple Podcast, and every other service out there. It is kind of wild that it is still free to use in my opinion.

Editing wise, the podcast has not evolved too much from where it started. Most recently I have started to use Izotope plugins to handle a lot of the de-noise and de-reverb issues (we recorded some episodes in a large public co-working space). I also use Neutron just as a general channel strip, I really like the active EQ on it. However, if you’re looking to edit a podcast yourself you don’t need anything fancy, you just need to know the general tips. In my opinion, removing silence from your tracks and layering compression in key. I try to avoid layering all the recorded tracks as much as possible because you tend to hear a lot of room tone this way. If you look at the image to the right, you can see I use strip silence to remove the areas where Christina, Maylin, and the guest weren’t talking. This process takes up a majority of the time used for editing as I have to go back and listen to the whole podcast and manually remove or trim some of the mistakes made by strip silence, but it is definitely worth it.

You also don’t want to use just one compressor for the whole mix. I tend to have one, or maybe two compressors per individual speaker. I usually set the ratio at 2:1 or 3:1 and look for anywhere between 2-5dB reductions (I also use a compressor on the board while recording so that nothing clips). Then I use a summing bus for some general compression and EQ to make the show have an NPR texture to it. The compressor on the bus hits a tad hard, but it is only used to cut down on some of the louder moments in the recording. Finally, I use a master compressor and limiter to bring the show up to -16 LUFS. At this point, the show is leveled, and everyone should be at the same speaking volume.

To be transparent, this blog post has made the whole process sound easier than it is. I have learned a lot throughout my two years of working with Femergy. Audio recording and mixing is a skill and talent that comes with time, and I am very thankful for Femergy taking me on and giving me the opportunity to use True Print as a way to grow my knowledge of audio engineering. If you’d like, here is a recent episode of True Print that really emphasizes the importance of growth.

Creating a Proposal Plan

Think Big! But not Too Big! This may as well be the slogan for creating a proposal plan on a budget. For context, Boxland is currently moving to a new location and with a new team. While this is very exciting and opens a lot of potential sources of income for Boxland, there are still resources and gear that Boxland needs in order to achieve these big ideas and events they want to do. I can’t get into specifics, but I can talk about my experience coming up with a proposal plan, and then revising it into a realistic proposal plan.

As an audio engineer, I know what tools and software are optimal and which are not. However, I have come to realize that explaining why one microphone or interface is better than another may not be the easiest thing for someone else to understand, especially when they’re the ones buying it. Long story short, my initial proposal plan had a lot of expensive gear on it. In my mind, everything that was on the list was still on a budget, but what I neglected to think about was what needed to be focused on and what could wait.

While I can’t get into specifics, there are things at Boxland that need to be upgraded for the new space, and there are ideas and business plans that, while amazing, will have to wait because of budget reasons. So, for my second proposal plan, I solely focused on stuff that needed to be upgraded and gear that would assist in small-scale events. This plan and list of equipment was still expensive, but nowhere near the size of the first list.

Creating these proposal plans expanded my understanding of running a business and the concept of prioritization. You cannot explore every great idea all at once, you need to focus on what pays the bills first, and then start expanding the business.

While I did not set up the cameras themselves, I was taught the importance of their angles. The room we recorded the shoot in is quite small, and in reality, it is just a decorated nook. Though, you would never know this by looking at the footage. The three-camera set-up allows each person to have a close-up angle and one wide-angle shot for both of them and the set. Lighting is also important. For this shoot, we had one very large production light right behind the cameras to light up the individuals and set as a whole. There was also a light above the set to remove any shadows on the faces of the individuals. Finally, built into the set itself were a few lamps in the background that removed any shadows caused by the large light behind the cameras. 

While recording, I was shown how to operate the cameras as well. The hardware and software that Boxland uses allowed me to switch camera angles live while recording, meaning any camera switching is encoded into the final timeline of the video. None of these switches are hard-baked into the timeline, however, and can be manually adjusted later. This style of switching live saves an incredible amount of time later during post-production.

Audio wise… I was not happy with the final recording. The main issue with the audio is that there is a lot of room tone and noise heard. I did not edit the audio, but I did place the microphones. Originally, they were placed in a way so that the individuals could speak clearly into them, about a few inches away from their mouths. Despite that, I was instructed to move them out of the way because they blocked faces in the camera shots. For the future, I want to create a better mic/mic stand placement that captures the individuals optimally and does not obstruct the view of the cameras.

After seeing some concept promotional art for the different parts of Boxland (Academy, BPN, Improv, etc.), I could tell they wanted something retro and video game related. The promotional art had all the logos for these silos on classic NES styled cartridges. From that, Cameron and I thought it would be a cool idea if we created some sort of fighting game "character select" video to showcase the new team at Boxland. The only issue was that neither of us knew how to animate something like that, let alone do it in a day.

We decided to split up the work. Cameron would design all the audio for the video and gather any voice-over we needed. Meanwhile, I would work on the animation and compile all the art and pictures that were needed for the video. I ended up using DaVinci Resolve to create the animation. The video needed to be done in a day, so I did not have time to learn a whole new software. Instead, I decided to utilize as many features as I could in DaVinci. Inherently, this software is not used for animation, but you can automate the movement of objects within a video. So that's exactly what I did, but with a lot of objects.

While the video would never win an award, I think it turned out amazing given the time it was made. Making it in such a tight schedule made me realize the importance of balancing what is reasonable and what is feasible. Working in the creative industry, sometimes my eyes can become much larger than what I can stomach. Meaning sometimes my vision of a project can only go so far due to limitations. However, this has never stopped me from creating something I can be proud of, and this video serves as a prime example of that.

If you'd like, you can watch some of the other goofy videos Cameron and I made on Boxland's social media. I will link to their Instagram below.

During our last weekend at Boxland, Cameron and I were told we were going to record a choir group. We knew going in that it was going to be a large group and that, in total, it was going to be around twenty or so songs. While this is a large project, we, at the time, were not stressed about it up until the night before when we were notified that this was no normal choir group. Instead, we were scheduled to record the official cast recording for the musical Second to Sherlock. A project that was crowdfunded on Kickstarter and was going to be sold on physical media to promote the show. The pressure was on, not only for Boxland but for Cameron and me as professional audio engineers.

So, here we are, the night before the recording, realizing that we were going to need to take this project much more seriously. We immediately started preparing and brainstorming. We needed to figure out the best way to record soloists, make sure everyone could hear the accompaniment, figure out how the accompaniment was going to be recorded, how the larger group songs were going to be recorded, and how the room itself was going to be heard in the recording. We also had to plan around our limitations of the space and equipment. The morning of, Cameron and I came in very early to set up and triple-checked that everything was going to work the way we planned it.

Luckily, the director of Second to Sherlock wanted the recording to sound as it would live, which is the way we planned on mimicking (why they didn't just record the actual show performance is beyond me). As you can see in the photos, we set up three solo microphones using Audix i5s. These were used to capture soloists or the lead vocal performers. Then, for large groups, we set up an XY pair of Samson c02s with an MXL 990 and Blue Yeti set in cardioid below. Then, to capture the accompaniment, which turned out just to be a keyboardist playing live, we used a simple DI box. All in all, everything turned out amazing. We were scared the keyboard, which was being sent to a monitor for the actors to hear, would bleed too much into the microphones. However, there was minimal bleed, and anything that could be picked up made the recording as a whole sound like it was being performed in a small theatre. The Samsons and MXL when paired sounded great. The Samsons provided great stereo depth while the MXL made the choir sound less like individuals and more like an ensemble. As well, the Audix i5s did an amazing job at capturing the soloists with minimal bleed from the large group.

At the end of day two, Cameron and I were very impressed with what we were able to capture. With minimal mixing, the recording was already sounding like a professional theater recording. Of course, it also helps that everyone involved with Second to Sherlock are amazing performers, actors, and musicians. I am incredibly proud of the work we did for this recording and incredibly thankful to be a part of the project. I will be the first to admit I was stressed when I realized the scope of this production, but as a team, Boxland, Cameron, and I prevailed. As of now, this is the largest recording I have been a part of and knowing that it will be enjoyed by hundreds of people fills me with immense joy. The Kickstarter backers and the supporters of Second to Sherlock will certainly be getting their money's worth.

Below are some photos of my time at Boxland