Episode 9: Echoic Memory
Christa Mrgan: Hey, Siri, do you have memories? Alexa, what does it mean to remember?
In Echoic Memory, You play a factory technician tasked with validating the audio memories of a line of AI smart speakers called the ECM1010. As the work becomes increasingly challenging, the machines begin to speak to you. And what they have to say may be surprising.
Welcome to the Playdate Podcast, bringing you stories from designers and developers and the team behind Playdate, the little yellow game console with a crank! I’m Christa Mrgan. Today, I’m speaking with some of the folks who made Echoic Memory, a narrative, audio clip matching game.
Spoiler alert: we talk about the overall gist of the game and mention some of the ways in which gameplay becomes more challenging as the game progresses.
Okay. Let’s meet the team!
Carol Mertz: My name’s Carol Mertz and I was a level designer and the logo designer for Echoic Memory.
Rachelle Viola: So I’m Rochelle Viola or Rachel. I did some game art for Echoic Memory.
Samantha Kalman: And I’m Samantha Kalman. I was sort of the Project Lead for Echoic Memory. I pitched it to Panic and built a little team and did a lot of the coding, and other stuff. I started in software testing many years ago, transitioned into design and prototyping. And then I made a game called Centris. I worked on Astroneer after that. And I’m currently working on Apex Legends for Respawn and EA. And I actually made Echoic Memory in the time between working on Astroneer and Apex Legends.
Echoic Memory is a musical puzzle game. But it’s a memory game. So, it’s very much based on the old card game “Memory”, where you have pairs of cards, you lie them on the table, face down and you flip them over in pairs to try to find the match. Remember where the match is, until you clear off all the, the matched pairs from the board.
It’s that, except using your own echoic memory, which is the sense of how long can you keep a sort of audio clip if you will, or a sample in your mind and remember what it sounds like. So instead of flipping over cards to look at a picture, you are, you know, flipping these buttons to hear a sound and you’re looking for the two matching sounds.
Christa Mrgan: Yes, that’s the main gameplay mechanic, but there’s another twist As the game progresses, it becomes more challenging, not just because the sound grid keeps expanding, but also the sound clips become distorted in different ways. And you have to use the crank to sort of decode the sound so you can match them before time runs out.
Oh, and there’s also two different game play modes, but we’ll get to that. First: how did the team hear about Playdate and how did they decide to make a game for it?.
Samantha Kalman: I guess I heard about it when it was first revealed. There was a little bit of website info and there was plenty of buzz on Twitter. I took a look at the device and I just thought it was really cool. I saw that they had a list of a handful of developers that they had already announced that they were working with.
And I basically just looked at that list and said, I should be on that list. They should work with me. I wanna make a game for this. So I pursued the opportunity pretty hard. And yeah, somehow talked them into letting me do it.
Rachelle Viola: I actually saw a Playdate on Instagram and Twitter, 'cuz like, I just love video games. So like all these like little handhelds and stuff, I think they’re really cool. And I saw the Playdate on there and I was like, oh, this is really awesome. I’m gonna pre-order one. And then luckily since I do a lot of, one-bit art, Sam was looking for some artists for the Playdate and I didn’t realize it was gonna be for the play date.
And she contacted me and I was just like, oh yeah, let’s do it.
Christa Mrgan: So, they did! and pretty quickly the idea expanded from Samantha’s original concept.
Samantha Kalman: Yeah, the original pitch for the game was just the pair match mode. That was in my mind, the core of the game. And when we got into it, it seemed like there was actually room for a story mode. I wanted to try to knock it outta the park.
So I was thinking a lot, you know, about these concepts of music and memory. And I was starting to do some research about the theories about how music helped human memory develop you know, hundreds, thousands of years ago. And I just had an idea for a story. You know, I was thinking about sort of the rise of smart devices and what they are, how people perceive them.
We put these things in our homes. We speak to them, they speak back, but you know, what, if there was something more there? And what could be a slightly fantastical evolution of this audio technology? And what kinds of, you know, interesting conflict or narrative drama could, could come out from that?
What sort of questions could that raise about, you know, what does it mean to be a person? What does it mean to be a human? What does it mean to have memory. To remember things at all? And so this story just sort of came out. I wrote the draft of the outline. I wrote a lot of the dialogue.
The conceit is that there are these robots that are smart devices. They play music for you. You speak to them and you are working at the factory, that’s making them. And so you’re, you’re sort of doing the final quality checks, if you will. Maybe that’s a throwback to my time in testing.
And you’re verifying that that the memory on the device that’s used to store and playback audio is valid. And when you do, they speak to you and they speak to you at first incoherently and then as you keep playing, as you keep repairing more of these devices or validating more of them you know, there’s an aspect of restoring memory that that could be happening.
There’s a question of like how intelligent, like, how smart as an intelligence are these devices actually? And, and it sort of goes from there. And so the way that the robot characters, the ECM 10 10 is what they’re called the, the way the ECMs talk to you, and what they say and and how it’s presented that was all Everest. I think Everest is one of the, one of the best people in the world for procedural narrative like this, and sort of, you know, storytelling and just what a, what a wonderful moment to work with work with them. I had sent them an overview of the story. It’s like, here’s a sort of vibe that we’re going for. You know, can we do something really cool? And they had ideas about how to present these concepts really gracefully that, you know, those approaches shipped in the game.
Christa Mrgan: I unfortunately was not able to speak with Everest Pipkin for this episode, but procedural narrative refers to a narrative with variables that have been randomized in some way. So there are lots of story elements and fragments. It’s not quite the same as a branching narrative, where there are pre-written story chunks that can connect in different ways.
The story partially assembles itself as you play, with different names, expressions, and memories in different combinations. So even if you play through all five acts of Echoic Memory multiple times, the game won’t be exactly the same each time.
The story mode is really immersive and flows well with the audio matching gameplay, and if you’ve ever played Pick Pack Pup or listened to the podcast episode about it, you’ll notice that a Koch memory is similar in that you can play either in the story mode, in which gameplay is tied into an unfolding narrative, and becomes more challenging as acts of the story are unlocked, or you can choose a pure gameplay mode called “pair match” mode in which you can set parameters like grid size, music, genre, and time limit, and just play the game as an audio puzzle only.
But a key element of difficulty that can make the audio pairs, especially challenging to match is the audio distortion parameter that I mentioned earlier.
Samantha Kalman: The very first idea was actually, what if I used the crank to change the playback speed? So what would that even sound like? And that was the very first prototype I made, in the pair match mode of the game. It’s that classical memory concentration rule set. And I tried, you know, well, how difficult would it be? How interesting would it be to be trying to match these pairs, but they’re all playing back at different speeds normally? So even if you hear the two sounds that match, they might at first be playing back differently so you don’t recognize them. And then, you know, the crank, you know, seemed like a perfect opportunity to make that change. It’s, you know, it’s sort of like twisting a knob on a musical instrument or an effect pedal or a turntable. And I just wanted to try that out. And when I did that first prototype, I thought it worked really well in a way that was really exciting to me.
Christa Mrgan: But changing the pitch turned out to immediately make the audio matching extremely difficult. So Samantha realized she would need to lead players through a series of levels of difficulty with other types of distortion before introducing pitch shifting.
Samantha Kalman: A lot of that came through experimentation. So when things get really fast or really slow, it really changes and transforms the sound. And then it’s like, okay, what are some, what are some filters that are easier to approach? So the, the crank on the device controls a filter, controls an audio filter that can change throughout the game. And there are many of these filters that are sort of built into the Playdate SDK. Um, And of course you can make your own if you want to. And I just started by playing with the different kinds of filters.
What are some filters that are harder to approach? And how can we use a progression of filters to make a progression of difficulty, to keep the gameplay interesting, to keep the puzzles interesting and challenging? So just sort of like, went in there and played around with it. It was kind of a more organic process.
I worked a lot with Carol, too. She did a lot of the, the groundwork on level design to help me understand what was easier or more difficult from her perspective and what could be the best sort of roadmap for the players from the beginning to the end of the game?
Carol Mertz: The level progression was really thoughtful actually with this game. So we started out with the like hundreds of clips that Samantha had put together, audio clips of little, like song nuggets, essentially, that captured the vibe that she was going for with the game.
So there were all sorts of different genres and like overlaps with each other things that sounded similar and things that sounded wildly different from each other. And that was my starting point as you know, a level designer. We also knew that we wanted to incorporate a handful of different effects.
And so what, the first thing that I really did was analyze each of these effects, you know, the pitch, the low pass, the high pass. And I essentially ranked them. You know, I’m not an audio person, so I had a chance to go through and listen to these effects and say, as a lay person to these styles and to these effects, what feels most difficult for me to process?
What, what feels like it adds a level of substantial difficulty and like, what’s the gradation of that, right? So we determined that low pass, high pass and band pass had the least level of difficulty as far as recognizing what the original sound was, where you know bit crush and ring mud were sort of, you know, medium difficulty and then affecting the pitch was actually incredibly difficult.
And so we incorporated that into the level progression. So I did a sort of an overlap of sound that sounded very similar. And slowly applying these effects from a, a difficulty standpoint. And that’s how we fell into the level progression and the level of difficulty. So you’ll notice it starts out very easy, 'cause we’re trying to introduce you to a lot of these concepts fairly slowly, but you know, the game is, you know, it’s only a few acts and it needs to crank up, so to speak, it needs to ramp up quite quickly. And so, you know, by the fifth act, it needs to have every single filter incorporated. And we had to introduce that in the best way that we could.
So I started out by doing a first pass on all of the level designs. I made recommendations as far as which filters to incorporate into which act and how to advance from level to level in a meaningful way that felt like it might be accessible to the player. And so I essentially passed that off to Samantha, who is the, you know, creative lead on the project. And she went through and took a look at everything and decided what was gonna actually make it into the final project and what needed some adjustments. But we were really careful about trying to incorporate, you know, the larger panels, you know, with, going from the eight grid to the 10 grid, to the 16 grid.
And then finally the 20 grid. This was all part of that process and, you know, she would provide feedback and I would update a little bit, but ultimately essentially, I, I made my initial recommendations and Samantha worked off of that, which was really cool to see.
Like I’ve always admired Samantha’s work in audio specific games. Centris was awesome and just kind of a mind bending experience for me to play. And so I knew that this game was going to be similarly amazing, you know, as far as just a musical auditory experience. But me, as not necessarily an audiophile, not somebody who works frequently with nuanced tones and sounds and music and things like that, it was really interesting to need to translate that into a level design and a difficulty arc, and it was exciting to be able to kind of switch my brain from game mechanics, which is, you know, where my background is, into incorporating difficulty and incorporating level design specifically with audio.
Samantha Kalman: Carol is just incredible designer. Jill of all trades and wanted her to help out, so, it was very much a, a kind of a hub and spoke model. I was sort of the hub and I worked with all the different people, basically one on one. There was a little bit of knowledge sharing, but everyone’s tasks were pretty compartmentalized. So I was lucky to be able to bounce around from person to person. And I think that was because the structure of the project, everyone was sort of working in the time that they had.
There were no, you know, set office hours. There was no office at all. Everyone was working from home. I made most of the game in London when I was studying actually I was studying electronic music at Goldsmith’s. So it was a good project for me to be working on at the same time as my studies.
And I guess just about half the game was made during the pandemic. So a lot of bedroom development and in my case, working from the university library.
And so I was just trying to keep the, the vision cohesive, if you will, about, how could the, the procedural narrative be? What should the logo look like? What should the art be? What is the design for these, you know, these robotic audio device characters? And so I was really lucky to be able to work with Rochelle on all the art. The art’s incredible.
There was a moment in the first sort of half of the game’s development that I was looking for an artist to work with and not having an artist was starting to put the schedule at risk and it was freaking me out a little bit. I reached out to I reached out to Panic. I was like, Hey, here’s the situation. I had someone lined up. It’s not working out. I need to find a new person like ASAP. And yeah, Cabel, I think specifically was tweeting about it.
I tweeted and he was helping spread the word and rallied a bunch of a bunch of candidates to be the artist. And um, ended up finding Rachelle through that process. And I had never met, Rachelle before. I had, you know, previously known Everest and Carol, but I guess, you know, meeting and getting to work with her for, for the project that was like an unexpected moment.
I’m really, really glad it happened. ’ Cuz the art is so good in the game. And trust me y’all would not have wanted me to ship this game with my crappy art. It would’ve not been a good thing.
Rachelle Viola: Since it was the first game I ever worked on, it was just like a really cool experience. And like, especially with Sam. So she’s like so awesome at it. It was like a really great introduction to the industry.
I’ve been doing pixel art since 2010. It started out in art school, I was really into the Scott Pilgrim video game. I was stealing sprites from the game for like a project I was working on. And then my professor was like, why don’t you just try making your own?
And I was like, okay, I can try that. So I just like opened up Photoshop and I did it, like, the hard way by like drawing out squares. And then I realized you can just make the canvas really small and actually draw with the pixels. So, ever since then I got really involved with the chiptune community online and they’re all super into pixel art.
So I did a lot of work with them. And I did a lot of live visuals with my pixel art. And now I’m doing games finally.
It was a fun challenge. ’ Cuz I really enjoyed working with dithering and all the texture stuff. At first it was like a fun exercise to like try to create different values without having to use different colors or shades. And like well a lot of stuff that’s like further away. If it’s in color, it’d be more muted. So for backgrounds, like I try to like use not too many like complicated textures. And then foreground is gonna be a lot more darker textures.
So the background would have like more dots rather than patterns. But sometimes it changes depending on if I feel like experimenting.
I, I like drawing like old mechanical things, things that are like kind of square are really aesthetically pleasing to me.
For like, a speaker that I drew, I took inspiration from like those old school speakers that you would hear like announcements on. It has like wood paneling and the beige fabric over it. And then it makes like a “womp wah womp womp” sound. But a lot of it was like 1960s, seventies. I’m not sure I’m not that old. But just like an older inspired type of mechanical stuff.
I would say for me, since I have a, more of a design background, anything that was animated, I was struggling with a little bit. I took one animation class in school and it was fun. It’s just like, it’s very time consuming and I wasn’t used to that type of workload. So it was just trying to. Change the pace that I work in, try to speed up and work with a different element that I’m not used to.
I actually use Aseprite. That seemed to be one of the most popular pixel art programs. and it’s really great for making games. It has a whole animation part on the bottom of the screen or wherever you decide to move the UI. It has everything you need to animate, like it has onion skins. You can easily flip through all the different frames. And then you can just like easily export the sprites as a Sprite sheet. I’ve been using it for two years now, I have like 2000 hours locked into it. But I really love that program. It’s helped me a lot with game design and game development.
I think Sam and I mostly communicated over email and I would just like give her weekly updates and then she would tell me what she wants, like different, or if I need to add anything else. It was very exciting to see everything like come together. She’s really great to work with. I like her a lot.
Christa Mrgan: I definitely think fans of mid-century industrial design will appreciate the speakers in Echoic Memory. And the art is really crisp and detailed, which is especially important when it comes to the grid interface with its knobs and lights, and the fuses that indicate how many audio pair mismatches you have left to make before the unit you’re validating is sent to the scrap heap
There’s also a clear battery level indicator. And draining a unit’s battery, which happens when you run out of time to make your matches became increasingly frequent for me as the game progressed.
And then something terrible happened: I accidentally deleted my save data! I had been playing Echoic Memory while helping to test one of the beta versions of Playdate’s operating system. And when it was time to update to the next version of the beta, I reformatted my Playdate without backing anything up and completely lost all of my game, save data. So this exact scenario is unlikely to happen to you, but since the game we’re talking about deals with restoring memories and I happened to accidentally delete my progress for the game, I thought it’d be a great time for a little aside about how saves work on Playdate, generally. And someone asked about it on Reddit too.
Okay, so! Unlike, say, a smartphone, Playdate doesn’t have multitasking a thing that most people take for granted with modern computers in general. So when you’re running a Playdate game, the game is pretty much the only thing that’s running, apart from some firmware that’s handling button presses or crank turns or things like that. But the operating system itself is not running at the same time that a game is running, for example. Playdate is kind of old school in that way.
So, say you’re playing a game on your smartphone and you switch away to answer a phone call or something. Your phone will probably save the exact state that your game was in when you switched away, because it can store that game’s whole current state in its RAM memory, at least for short periods of time. But with Playdate, if you’re playing a game and you hit the menu button,
Which serves as the game pause feature, and also the system wide options menu, this freezes, the game’s run loop in place right away. If you were in the middle of, say, using the crank to turn a stereo knob, the game just stops before the next operation it was going to perform and Playdate hands control over to the code for the system menu. The menu is small and lean enough that play date can load that at any time without messing with the game that’s paused in the background.
If you then hit the menu button again to dismiss the menu, the game resumes its functioning exactly as before it was interrupted from that last turn of the stereo knob.
But if after you hit the menu button, you then choose home, instead of resuming your game, Playdate stops running the game completely, and it starts running the launcher application, which is part of the operating system. The game is fully quit and unloaded.
So what happens to your game state in that case? Well, the answer kind of depends on the game’s developer, but one thing that won’t happen is a full state save, l ike you might get on a smartphone. So if you exit a game, return to the launcher, and immediately restart the game, you won’t be in the exact spot where you left off, but you’ll probably be pretty close, because most developers are saving your progress as you play, automatically keeping track based on certain checkpoints, inventory, and achievements. They save that to the disc, where it’s nice and safe across time. Unless you happen to reformat your Playdate! The save file will usually contain this kind of information: what level you’re on, how many points you have, and what all you’ve unlocked or discovered so far. When you return to the game, it checks for this data. And if it’s present, it starts you there, instead of all the way back at the beginning.
So for most Playdate games, you don’t have to worry about saving. If you exit a game, you’ll pick up pretty close to where you left off, whenever you return to it. Though, the game won’t be in precisely the same state it was when you exited. Luckily you’re unlikely to have your game interrupted by a phone call on the Playdate.
And while you can explicitly hit “save” on a lot of games on modern consoles, auto saves are pretty standard these days, too. Like in the game Legend of Zelda: Breath of the Wild, you can always hit save to later pick up pretty much right where you left off, but the game is also automatically saving in the background, every time you collect a new item, or sit by a campfire, et cetera. And if you think about what’s saving on an older console, like the Super Nintendo was like, you’ll appreciate how much faster these current saves are. Even on Playdate, they’re mostly unnoticeable. But if you’ve ever saved a drawing or a song on Mario Paint for super Nintendo, you might remember that it took a while. While Playdate may not have the processing power of a modern smartphone, developments in CPUs and hard drives have come a long way since the early nineties!
And as for my lost progress in Echoic Memory, I was able to dig into my Playdate’s disc and adjust my progress via JSON file. So I could pick up where I left off. This is not cheating in my opinion. And I was lucky that Echoic Memory happens to be a game whose save data files, aren’t encrypted, which meant that I could mess around with them.
Some developers encrypt their game save files. So you can’t do this kind of thing. Sas-cough-quatchers-cough Oh, excuse me.
Anyway, the moral of the story is to back up your game data before fully reformatting your Playdate, and don’t worry too much about saving your game progress. Developers have you covered for the most part.
And as a developer, Samantha had a pretty good experience with the Playdate software development kit.
Samantha Kalman: So the Playdate SDK really surprised me. Like honestly, when I first looked at it, it blew me away how comprehensive it was. And this was years ago, this, this was within the first year after they had announced.
And before I think before they even started taking it to to conferences to show off the device. So, so there was very little information available until I met with Panic and got a copy of the SDK and the API. It was really, really comprehensive. I had previously worked on Unity and the Unity documentation was very comprehensive.
And so looking at this, I was like, wow, this reminds me of an early version of Unity. Like it’s, it’s just complete, it’s informative. It’s helpful. Like, I can go to this documentation with my questions and just like read through it and learn what all I can do with this. So I used Lua to make the game and it was my first time using Lua and they had support for Lua in there.
The simulator tool was awesome to be able to, you know, test things on the computer, just change, build, test and then publish to device when I needed to. But it kept me in a tight sort of iteration loop that I had really gotten used to working with other tools like unity. It was really, I was really, really pleasantly surprised about it. So tons of props to the team for having great documentation and, and support there.
I would be happy to make another Playdate game. I’ve got a bunch of ideas for more Playdate games. I don’t have a problem with ideas. I have a problem with having too many ideas. So if I did that, it would probably need to be like a hobby project.
I’m pretty devoted to my day job right now, and that’s taken a lot of my creative energy and making a game, you know, is, is a commitment. Even a small one, I think. Echoic Memory’s pretty small, but it still took you know, a year of hard work. What would make me most happy until I get a chance to make a new Playdate game would just be to, to be involved in the community, to be seeing other people’s games, you know, people can ask me questions and I will be happy to sort of point point folks in the right directions about how to like design things or code things or work with Lua or something about the API.
I’d like very much to be part of the community of, of Playdate and developer world. Hobbyist, pro, whatever, doesn’t matter to me, just participate. And so that probably means that I need to get on the forum more often and, and actually participate!
Christa Mrgan: Yeah! The Playdate SDK is free and available for anyone to download, and there’s a pretty active developer forum where you can talk to other developers and find answers to your questions. I put a link in the show notes.
So what kind of experience does the Echoic Memory team hope that people have when they play the game?
Samantha Kalman: I hope that they are charmed above all, at first, I hope they are charmed. I hope that they’re curious about what even is Echoic Memory and what is gonna happen here? I hope that the story mode delights them.
I hope that they stick with it. And if they do, I, I hope it unravels in a way that really grips people. And makes people think, or makes people feel something about having memories, I guess I’ll leave it at that.
Carol Mertz: What I think is really cool about Echoic Memory is that it sort of teaches you how you listen to things and what you recognize. Which I think is beautiful.
Christa Mrgan: It is! I hope you enjoy the close listening and fine tuning it takes to restore the decade spanning stories in Echoic Memory. You can find out more about Samantha Kalman, Carol Mertz, Rachelle Viola, and Everest Pipkin via the links in the show notes. Thanks so much for listening and stay tuned for more episodes, coming soon to the Playdate Podcast feed.
Carol Mertz: Thank you!
Christa Mrgan: The Playdate Podcast was written, produced and edited by me, Christa Mrgan. Cabel Sasser and Simon Panrucker composed the theme song. Additional music and sound effects were composed by Samantha Kalman and come from Echoic Memory.
Huge thanks to Tim Coulter and Ashur Cabrera for wrangling the podcast feed and working on the website, James Moore for making me an awesome Playdate audio extraction app, and Neven Mrgan, who created the podcast artwork and site design. Thanks also this week to Neven Mrgan and Steven Frank for reading over that part about how Playdate saves work, 'cuz I was a little worried about getting it right.
And thanks as always to everyone at Panic! Playdate is shipping now and available for pre-order at play.Date.
So did anything especially interesting or surprising happen during development?
Samantha Kalman: Interesting or surprising… I think it was all pretty like boringly straightforward. I think maybe, the situation I was in is that I was starting up this this degree program while I was working on the game. So I was really juggling you know, one moving and living in a different country and being in full- time academic program and trying to make this game . So I had to establish a routine to get it done. And there wasn’t, there wasn’t really room for surprises. A surprise probably would’ve been bad.