Looking Back on Chapter Select Season 3 – Banjo-Kazooie

It is time to pack away your backpack and order a large pizza to get fat and lazy because the Banjo-Kazooie season of Chapter Select is done! We just barely got one season done within the first six months of the year. It wasn’t my plan to cut it that close, but we have been busy with other seasons of the show. 👀 I know I have said it in both previous Looking Back posts, but Chapter Select remains to be my creative highlight. I suspect it will continue to be for years to come.

Season 3 – Banjo-Kazooie was our toughest season yet. There were creative hurdles. The show’s structure was tested. As is now tradition, I want to take time to pull back the curtain on the production of the show. I’ve got more data, insight, and pre-production assets than ever before. I think a lot of this was bore out of the challenges we faced this go around. I also think it has led to a better product – both this season and in future ones.

Our biggest challenge in Season 3 was ourselves. Logan and I have been co-hosts for over seven years and friends longer than that. We ran a weekly show for three years and have had other shows and projects. For Banjo-Kazooie, we fell into a familiar complacency. Show notes were not as structured, giving conversation a looser flow. This directly lead to longer show times, bubbling well past our rather typical sixty to seventy-five minute show.

Half way through the season, I decided to implement some changes I had been mulling over for a bit. You may have noticed it. Logan is now the main host during the show. I intro and outro every episode, but Logan guides the conversation based off our show notes and structure that we write beforehand.

This change provides quite a few things for us. Before, we alternated hosting each episode. We each have different styles and this could create inconsistency in the cadence. Also, Logan is a natural host. He has a vision for the conversation and guides it well. I knew it was the best call for the show.

We also plan to stick to the seventy-five minute window we seem to have naturally settled into during the first two seasons. Some episodes will naturally be longer.1 This will give us, our guests, and you the listener appropriate expectations for each episode.

Old habits and familiarity are hard to shake. We realized that was happening to us this season. I think the changes we are implementing will help us stay on target. We want to make the best show we can. I believe these tweaks will let us do that.

Numbers and Performance

This season was an experiment for us in regards to the numbers. The show’s feed was quiet from the beginning of January to the middle of May. Numbers dove when the bi-weekly cadence of Season 2 was broken, especially coming off the 4,000+ download spike from Season 2. January 2022 had 3,889 downloads, February went down to 759 downloads. The lowest we got during the gap between seasons was 133 in April.

When were started publishing the Banjo-Kazooie season in May and the feed was active again, we numbers started climbing again. May capped out at 1,817 downloads and June was not too far behind with 1,579.

Since we were not publishing, we had the opportunity to see how it would affect performance. As you can see, four month pauses hit hard. I understand the God of War jump to the clouds meant are hard fall down. I also can see how these summer months have been performing better, not just due to activity in the feed, but swirling God of War: Ragnarök rumors and the announcement of The Last of Us
Part I.

Having Banjo-Kazooie be our focus also let us test the waters with a shorter season. Not only performance wise, but also listener reception. Honestly, the show gets little vocal feedback. This season did get a bit more engagement on Twitter—I tried using hashtags effectively. Some folks I told about the season had a positive, nostalgic response to Banjo and Kazooie. Their warmness affirms Logan and I’s thoughts on the series legacy.

Apple Podcasts’ data is, I think, quite positive. Show plays have grown by 229 since the end of Season 2, wrapping this season up with a total of 655. There are 10 followers2, 19 listeners, and 14 “engaged” listeners (whatever that means). Apple also reports a total of 57 hours listened. For Season 3 in particular, Apple tracks 183 plays across all the episodes, with S3E3 – Banjo-Tooie being the most popular at 61.

Spotify’s numbers are not as positive. The overall numbers have nearly doubled from last season. We are seeing 350 starts, 152 streams, 147 listeners, and 10 followers. Season 3 in particular though hasn’t done so hot. The Bear and the Bird only had 13 streams on the green app. That points to the growth occurring in the back catalog of seasons and episodes. That’s perfectly fine with me; just means Season 3 will have its time in the spotlight down the road for the Spotify audience.3

YouTube remains to be the lowest platform I can directly track. This doesn’t surprise me. The season’s total view count there is 56 across all five videos. Once again, the trailer is the top dog (most likely due to being published the longest), with the rest trickling down.

The mystery of podcast analytics continues to stump me. Perhaps true, actionable data is blocked behind paywalls and creepy trackers. I’m not down for paying to track listeners. If I personally don’t like it when company’s track me, I would never do that to my audience. But I would love to know the answers! Why are the Libsyn numbers so high? Is this growth good, natural, or some wonky server somewhere spiking numbers?

Despite the fog of bar graphs and stats, I continue to be encouraged and find the positive in our upward trends. Even with a four month gap, we maintained growth across all measurable platforms. The gap to Season 4 will be much shorter, which I hope and suspect will help keep this trend going.


There is another set of numbers that I wanted to break down from this season. This time around, I did my best to time track playtime and production. I missed a few editing sessions here and there, but I did get a full account of my playtime thanks to the in-game clocks.

Playing through the four Banjo-Kazooie games, three for the first time, took me 56:41:46. Here’s the breakdown per game:

  • Banjo-Kazooie – 12:17:08
  • Banjo-Tooie – 17:13:53
  • Banjo-Kazooie: Nuts & Bolts – 12:35:13
  • Banjo-Kazooie: Grunty’s Revenge – 3:28:03
  • Conker’s Bad Fur Day: 11:07:29

I was surprised to see Kazooie and Nuts & Bolts be so close to each other, especially considering that the Xbox 360 game felt longer. Which leans into the five hour jump for Tooie and why we complained about that game’s length. If I thought Nuts & Bolts felt more like fifteen-ish hours, than Tooie felt well into the twenties. Thank goodness for the briskness of the Game Boy Advance game.

As for logging production time, I have just shy of 21 hours clocked. This does not include the first episode. Some episodes I clocked the time it took the video version to export, others I didn’t. This was not the most consistent effort of the season. Heck, I’m not even logging the time I am spending writing this piece. Here’s a total breakdown by episode:

  • Banjo-Tooie – 3:58:59
    • Audio Editing – 3:58:59
    • Video – N/A
    • Export – N/A
  • Banjo-Kazooie: Nuts & Bolts – 11:47:09
    • Audio Editing – 6:18:52
    • Video – 3:33:02
    • Export – 1:55:15
  • Banjo-Kazooie: Grunty’s Revenge – 3:48:16
    • Audio Editing – 3:25:44
    • Video – N/A
    • Export – 23:32
  • Conker’s Bad Fur Day – 6:04:20
    • Audio Editing – 3:23:58
    • Video – 7:22
    • Export – 2:33:00

I am by no means a fast editor: I think these times showcase that. I can become easily distracted while editing the show; looking up research for show notes, rabbit trails of adjacent show tasks that need doing, etc. My best time to edit is super early in the morning with a cup of coffee. I find myself laser-focused in the first few hours of my day.

Nuts & Bolts obviously has the most logged time here. I know for sure that just the edit did not take six hours. I wager I was running the clock while prepping the show’s publication on Libsyn and Max Frequency. I suspect the show’s edit was similar to Tooie and Grunty’s Revenge. Take the show time, double it, and add some miscellaneous time for my distracted, but adjacent work.

We can also use Nuts & Bolts to see the time I poured into the video edits. It’s essentially the same formula of taking the show length, doubling it, and adding some extra time for whipping up graphics or finding just the right clip. Now that the video production has changed, I am gaining those three-ish hours back. It helps cement my decision as the right one for the time being. Seeing my time written out helps put the effort into perspective.

The video quality of the episodes is dependent on the highest quality I can play and capture in. That peak right now is 4K60 with HDR.4 Now let’s compare the video export between Nuts & Bolts and Grunty’s Revenge.

The Xbox game was captured at the highest quality, while I only captured the Game Boy Advance title at 1080p.5 The difference in sheer image quality on the export time is nearly six times. It’s no skin off my back, since I just let the iMac plug away, but dang, 4K60 is resource hungry.

Between all the playing, recording, editing, creating, writing, and publishing, Season 3 wound up probably being close to 100 hours worth of work. All for four episodes! I can only imagine the time spent on Paper Mario and God of War, both with more games that are the same length or longer!

Since I track my time for work, I have been trying to do so for Chapter Select as well. My time logging should be more clear and segmented for more accurate numbers. I am happy to have some data and hope to continue to get better for future seasons.

Guests

What a jam-packed lineup of guests! As I mentioned in the season announcement post, Season 3 has the highest ratio of guests to episodes so far. I think we found a knowledgable group of folks to help us dig into Rare’s most iconic characters.

I knew as soon as we landed on Banjo-Kazooie for the season that we needed Michael Koczwara on the show. He is the biggest Banjo fan I know and he, ya know, mods the game to make his own. He has a deep appreciation and understanding of the first game. I thought his voice would be great to balance out my own, given that this was my first time playing Banjo-Kazooie.

Keeping the Michael train alive, having our pal and long-time podcast co-host Michael Ruiz play Nuts & Bolts was obvious to us. Why? Because he is the only person we know that says they like Nuts & Bolts. Combine that with the years the three of us have worked together — that episode pairing was a no-brainer.

For Banjo-Tooie, Brian Henken was our man. As a game designer, a composer, and a staunch defender of Donkey Kong 64, I knew he was the guest for the sprawling sequel. What I wasn’t aware beforehand was how much Brian played Banjo-Tooie. This guy out here dabbles in speedrunning and collecting everything in the game.

This season’s line up felt natural. Each guest brought richness and context to these games that Logan and I could not have done on our own. I learned so much and I hope you did as well.

Audio

The audio production this season remained largely the same as last season. As for the pre-recorded credits experiment, I continued it this season. I think it will stay a part of the show for now. I did straight-up forget to add them to one, maybe two, episodes. I would get to the end of a long edit and blank on dragging the file in. By the Grunty’s Revenge episode, I was dragging the credits in at the start of the edit, so I wouldn’t forget. Perhaps I need a check list for the edit process.

There was one close call with some low volume audio. I needed to resort to a portion of the backup recording for a chunk of one episode, but the audio coming into the backup was low for one person. Audio Hijack 4 had launched not too long before and included a new feature dubbed “Magic Boost.” As described by Rogue Amoeba, Magic Boost let’s you “crank up quiet audio without increasing already-loud sounds.” So I ran my backup recording through a new Audio Hijack session with Magic Boost. It saved the track and the chunk of the show. It truly worked like magic. Audio Hijack continues to prove how essential it is to my workflow.

Video Production

The video this season changed quite a bit. The video version is still produced, but I simplified the edit on my end. Now the gameplay is long, uninterrupted chunks. I made a full post about why before the Banjo-Tooie episode debuted. I don’t want to rehash that all here, but the change has been relieving for these past few episodes.

By far the best video from this season was the season announcement trailer. We had the entire opening cutscene of Banjo-Kazooie rewritten to be Chapter Select themed. I still can’t believe it worked so well.

It started out as just an idea to adopt Grunty’s rhyming scheme and whip up a script for Logan and I to perform. I just pasted the script in Craft and wrote a first draft. Using the power of Rhyme Zone, I was able to create a story about Logan and myself picking Banjo-Kazooie as the season theme and pitting the N64 against the Xbox Series X. It was a good first pass that proved the concept had potential. The immediate issue I had after writing was that the primary focus was this made-up twenty-year-old console war, not the games themselves.

The other major “issue” is/was that Logan and I are not actors. We could have read the rhyme and it would have been fun, but super cheesy. So I approached the second draft with the goal of keeping all the characters in their roles. This led to what you see today.

It was then that I started thinking about putting the script in the game itself. Modding Banjo-Kazooie is fairly common in the N64 mod scene. I thought I could perhaps learn how to tweak a few lines and apply it to a ROM. I was wrong about the learning curve. There were some applications I found that let me fiddle with the game, but finding and modifying the cutscenes was not as surface level in the applications as I had hoped.

Then I thought about our first guest Michael Koczwara. He is a Banjo modder and knows the game far better than I ever could. I described what I wanted to do and asked if A) it even was doable and B) if he’d be willing to help us out. Thankfully, the answer to both questions was a yes.

It didn’t take long to see test footage of my rhymes in the game’s opening cutscene. Michael and I quickly learned though that changing these cutscenes wasn’t so cut and dry. For example, the first three lines could not be changed. That issue was easy enough to work around given the first editable line was Dingpot bringing up Tooty: A perfect opportunity to mention a certain podcast instead.

The biggest issue was Banjo and Kazooie staying silent. For whatever reason, changing the dialogue rendered the stars mute. Fortunately, they don’t talk too much to begin with. That did create some slightly too long scenes of their character models waving wings for no discernible reason. I tried to trim those down in the video editing process, but finding a clean frame to cut off of was tricky.

Turns out that modding N64 games is hard! It introduces bugs and creates issues that seem unexplainable. I am forever indebted to Michael for his patience and assistance in turning this goofy idea into a reality.

Art

Creating the art for the season is always a good way to stretch my visual creativity, especially in a audio-focused medium. This season’s art was, by far, the toughest to realize. I struggled to come up with the gradient + emoji + custom font fusion that ended up being used.

I started out digging around for concept art. While neat to find N64 era concepts, it wasn’t fruitful for my goal. Plus, I did the concept art bit for Season 0: The Last Of Us.

Then I was drawn to the art on the original run of the Banjo-Kazooie soundtrack from iam8bit. I’ve always liked the artistic rendition of a real bear and bird with a backpack. Should I use pictures of real animals? No. I just did real photographs for Season 2: God of War.

Was I creatively tapped out three seasons in? It felt that way sometimes. I then went into concept mode.

Having an epiphany that my Wii could load up the N64 games, I thought perhaps I could buy the Fangamer Banjo-Kazooie plush and take photos of it in front of the various games on their original displays. It ended up being artsy, but not good podcast art.6

It wouldn’t be until I was editing the first episode that I landed on the emoji concept. While breaking up the chapters for each world as we discussed/ranked them, I thought it’d be a nice touch to add emoji to the chapter titles that paired with the level’s theme. A small touch that adds character to the chapters and would be repeatable for each episode. Why not tie the emoji into the art as well?

This kicked off an iterative phase. I took the Chapter Select template, threw in four emoji that captured key elements or characters in the game and color matched the background to one color from the original box art. It looked…flat. Even after changing the color of the bird, witch, and puzzle piece emoji to match the stars of Banjo-Kazooie. Logan suggested adding the logo to more clearly communicate the game, but that made the art too wordy.

That led Logan to suggest changing the color of the font to the iconic blue and red. I decided to do that and experiment with the font itself. So far, I’ve been particular about keeping the corners consistent across season’s for brand identity. I dug around to find the font that was used by Rare and landed on Lithos Black. We tried that, but it lacked the charm.

During all my searching, I stumbled across a reddit post and tweet of a customized blue and red Banjo-Tooie styled font by Oakland Elliff. It’s the perfect font, but all tied up in a single image. I decided to go through the effort of isolating the letters in “Chapter Select” and sizing them for the art. It immediately sold both Logan and myself. The font clearly communicates the franchise, while the emoji denote the specific game.

The background color was still flat, so I switched one color out for two in a gradient. The diagonal gives some direction, while the colors help elicit the specific title.

This season’s art was a real hurdle for myself. I’m no artist artist. The best work I’ve done is doodles in the margins. I am, however, dedicated to making something unique for each season of Chapter Select. I think it’s helping me become a tiny bit of an artist.

And chronicling these, for a lack of a better phrase, creative exercises helps me remember where it all started. I would like to think it helps you too. Transparency about the creative process has always been built into the production of Chapter Select. Sharing all of this provides context for decisions on the show. I hope it enriches the show. Perhaps it helps you make something too.

The Future

We have the next three seasons planned out. We’ve never had this many locked in at once. It’s ambitious of us, but the slate of episodes will be incredible. And the future of Chapter Select is a whole lot closer than you may think.

For the first time, we are announcing the next season’s focus well ahead of release. In fact, we are announcing the next two seasons’ themes — today.

At 12:00 PM EST, I’ll make the official announcement here and on Twitter @ChapterSelect. Season 4 is an idea I have had to do with Logan for years. Season 5 is going to be scratch a tasty itch Logan and I both have.

Season 6 is being kept under wraps for now, but we have been working on it in one way or another all year long.

The future of Chapter Select is bright. I am stoked to be making the upcoming seasons. When we plan and talk about the show, we both get so jazzed about where we are headed. I think you’ll enjoy what we have on the way. See you soon.


Update 7/13/2022: Surprise! We have added a fifth episode to the season all about Conker’s Bad Fur Day! We decided to take this opportunity of playing Rare 3D platformers to try out their more infamous N64 game. I’ve updated this Look Back with stats for this episode.


1. I’m looking at you Ragnarök.

2. I miss the term “subscriber.”

3. Speaking of said audience, according to Spotify, the musical artist they listen to are Ice Cube, Nas, Rick Ross, Lil Wayne, and Dr. Dre.

4. No episodes have been released in HDR simply because my iMac cannot export one to two hour 4K60 HDR videos in any sort of timely manner.

5. I suppose I could have played on my MiSTer at 1440p, but I wanted to use my Analogue Pocket.

6. Plush not featured. That was my Banjo and Kazooie amiibo standing in.