“how do you hold a square wave in your hand” — thoughts on “preserving” vgm and computer music

51 min readJul 6


many people on the internet talk about preserving things. however, very few of them have ~75% of a master’s degree in library science. I do! and my studies over the past year have enabled me to look at the discussions I see playing out often in internet computer-music circles from a totally new and hopefully kinda original angle. regular readers know I hate to claim authority on any topic, so I’m not really trying to do that here either. but I have been engaged in this one in multiple capacities for a number of years now… you can decide for yourself whether that counts for something. like, maybe if you don’t like the ideas in here you can just rest assured that it’s because I don’t actually know what I’m talking about! that’s a nice thought. anyway, please email me as always if you think I am wrong about something. (I guess I should be telling you to leave a comment for Engagement but I just think writing “please email me” is really funny so I’m gonna do it multiple times)

I have four-ish thoughts that are mostly unrelated. but they all have to go in here or else they’re never going anywhere. so I’m sorry if there turns out to be no Through Line, but it’s not like I’m getting paid for this or anything

(ok, jax from a couple months later checking in here: actually they are pretty related I think. so buckle up, you’re gettin throughlined. you’re getting lined through… you’re going to get through the line)

(anyway it’s pretty long as usual (and links a lot of supplemental reading) so you could still maybe take a little break between each of the parts if you want. when you get to the end of a section just imagine that wii sports “why not take a break?” screen popping up, and if you want you can just ignore it and keep playing tennis (i.e. reading).)

(ok while I’m here, here’s a classic jax disclaimer. my experience is mostly with “old video game music (aka vgm)” and “music made by hobbyists using trackers or sequencers”/“chiptune”/“demoscene”/whatever other term you might have for this. in fact, I think in a lot of cases I am using “computer” as a loose synonym for “internet”. there are all kinds of other computer musics that maybe don’t fit so well with what I’m writing about here. including Serious Music made by Composers that happens to use computers which I respect a lot but is probably in somewhat of a different world… I dunno I have been really beating myself up about scope and how well-read I am taking credit for being. so I am going to try to define the scope as precisely as I can at the beginning of each section but, uh, Sorry!!)


ptConserve — can “community archives” exist on the net?

(this section is about music made “on the computer”, i.e. by communicating with other people over the internet)

for some reason I’ve always hesitated to call internet spaces/groups “communities”. I guess it feels a little played out. a lot of things get called communities that don’t really seem robust or connected or friendly enough for me to use that word. when I see a post that starts with something like “as a member of the teleroboxer for virtual boy (1995) speedrunners fan meme edits community, I think…”, it makes me feel like a grouchy old person. what kind of community could possibly exist around a topic like this, I wonder. what kind of authority could a member of this community really possess. in theory I hold communities in a certain high esteem. I imagine that they develop as a form of resistance against oppressive forces of erasure and assimilation; that their members form unbreakable, lifelong bonds; that their stubborn persistence in spite of the mainstream ultimately leads to the invention of unique discourses, schools of thought, cultural products. can those characteristics really develop without proximity, hands-on collaboration, face-to-face coexistence?? maybe that’s an unreasonable target, anyway… everyone needs community and these days one of the easier ways to find it is by talking about a niche topic on the internet. maybe I need to allow for communities that are a little more humble. I’m usually all about humble things…

I “run” a small discord server dedicated to a particular free music sequencer, “pxtone”, though in many ways it sorta runs itself. for the sake of this post, I’m going to (begrudgingly) allow myself to call it a community. it has developed a relatively robust local culture and produced some interesting-enough, unique-enough works of art to qualify, I think. some particularities of the format have encouraged a sort of open-source-style collaborative climate. projects are bundled in a single “ptcop” file of relatively small size, easily shared, which often doubles as a playback file. (there is also a parallel format “pttune” restricted to playback only, in order to keep your samples/techniques secret I suppose, which very seldom sees use.) instruments can be exported out of these files for reuse. the interface and the features are simple enough that the basic mechanics of any innovative technique are pretty easily understood; no extensive scrutiny or reverse-engineering necessary, usually. all you can really do is load instruments, click notes in, manipulate a handful of parameters and apply a total of Two (2) effects.

given all of this, when someone posts music on the pxtone server — often via the bespoke sharing website “ptweb”, developed and hosted by one of our own (thanks sidedishes!!) — it is generally understood that their sounds and techniques will be adapted, modified, and evolved by other users. pxtone folks are eager to participate in this cycle, generally unafraid of plagiarism or theft or exploitation of their work, offering up their brain-products to the communal well of inspiration and drinking freely of it in return. this may sound naive or too good to be true. let’s return to that thought in a little bit…

screenshot of “teleroboxer” for virtual boy taken from JagOfTroy’s longplay on world of longplays. hmmm this blog really doesn’t have as many jokes and anecdotes as the last few. sorry about that. I can’t even think of something funny to put in this caption. you could say I’m a new member of the “not being able to think of something funny to put in jaxcheese blog captions” community.

archivists vs. communities

about a year ago a computer music archivist posted a tweet asking the public about methods for downloading a complete copy or “site rip” of ptweb so he could store the project files in his online repository. he suggested that were he not to make a copy, the site and its contents would likely be lost to time. this is not really misaligned with the basic philosophy of most internet preservation practices: stuff on the net is super-ephemeral, and it fails/disappears suddenly rather than gradually, so we should make copies while we can and provide access to those copies independently from the original infrastructure (access is an important part of preservation, after all. or maybe it isn’t. let’s revisit that in “exhibit b” or c or something). but it kicked up a wave of worried discussion on our server. sure, we had uploaded our music to a public website, knowing well that any visitor could readily download it. but we did that in a certain context: knowing the volume and general makeup of the site’s visitors; using the site’s features to prescribe a certain presentation and add commentary which might include important requests or attributions, if not just valuable context. of special concern was the likelihood that files marked as “not for download” by the author would in fact be collected by any automated site rip tool, which would download the tunes not through the site’s user interface that disables the download button, but in some more direct computery way that doesn’t care about a download button.

ptweb’s developer, looking to preserve the site’s users’ privacy but also to work productively with the archivist if possible, discussed some possible measures to prevent — or at least warn in good faith against — the downloading of those private files. others of us discussed the implications of our tunes traveling beyond their original context and audience. would their new listeners revere and respect them in the way we did? would our works plant seeds of new creativity around the net and bring new collaborators into our space? or might they be absorbed into some more thoughtless (or even exploitative) practice of reuse — download en masse, rip off, throw away (or reupload somewhere else with even less context).

one summary feeling that emerged: these concerns would be alleviated if the archivist was part of our community or at least took some time to get acquainted with it. we knew this firsthand, because ptweb was our community-led archival effort, designed for and by us (mainly by one of us, but also refined frequently via suggestions and collective consensus), aligned with our particular needs. maybe any internet-archivist’s instinct here is to funnel it into a larger centralized repository, but it’s worth thinking deeper than instincts. is our localized archival effort really more vulnerable to the ravages of internet-time than any other? in any case, can we really not be trusted to be the best stewards of our own culture? what inspires this mistrust; maybe the small size and apparent insularity of our niche?

community archives

these questions and ideas are connected to a trend (which is not to call it a fad, rather a recent-ish, still-developing breakthrough!) in archives that was introduced to me as “community archives”. though there seems to be a whole web of interconnected theories and practices under various names covering similar ground. the crux of it, to my understanding, is that archival efforts should be performed in coordination with, or better yet led by, the creators and/or original stewards of the items in question. their knowledge of the items — the circumstances of creation and use, the connections between the items and the community and its activities, and more — are too important to dismiss. this is opposed to methods that simply bring items in to established, resourced archives and into the care of trained archivists, valuing most highly the archivists’ training in the traditional methods of their profession and the superior technical capacities of the institution.

in “opening archives: respectful repatriation”, kimberly christen writes about the prevailing archival obligation towards “public” or “open access” in the context of working with the cultural and historical products of indigenous peoples, where it conflicts with their localized information practices:

…Collecting institutions’ core commitment to access predicated upon openness to the public severely limits the possibility of seeing indigenous claims as alternative types of openness (access differently conceived). Oftentimes in these situations, indigenous systems of information management are defined as “cultural values” or “tradition.” In either case, while collecting institutions may be sympathetic to these “concerns,” they do not see them in the same semantic light as the assumed universal claims on which their assertions of a uniform typology for access are based. Paul Dourish and Johanna Brewer suggest that viewing information as a “natural category” rather than as a “cultural category” limits one’s ability to see the processes and relationships that ground information systems within larger cultural logics and historic events. A tacit naturalization of information as a universal category, then, works against recognizing indigenous systems as on par with accepted Western institutional models. At the same time, popular notions of censorship and information lockdown delay imaginative responses to other modes of circulation. When access as openness is taken for granted as a de facto public good, systems based on limiting access often get defined as oppressive.

That is: archive institutions often marginalize the information access values of the communities they “collect”, in favor of an emphasis on open access that many archivists see as a universal ideal. I fear a little that what follows is a comparison that trivializes christen’s work or its profoundly important subject (so, so far beyond the importance of niche computer music). but bear with me for a little while I nonetheless try to extract something useful from the comparison.

computer- and internet-oriented archivists, more than any other sort, frame information in this “natural”, “universal” way, and are raised on “free and open” as an ideal of the highest order. this has many advantages in many cases. it is probably the only way to approach internet-scale projects, to make even a feeble attempt at preserving the unthinkably vast volume of information that appears on and disappears off the internet every day. no time for worry over case-by-case nuances when a whole library could evaporate in the time it takes to evaluate one book, especially when you have a magic copier that can reproduce verbatim that entire library in a matter of seconds. but who and what is left behind by this way of thinking? what opportunities are missed? how useful is an item archived so thoughtlessly, if the people who need it are even able to find it? how crucial is “context”, that information that is not strictly part of the item but inherent to its original home or stored in the minds and interactions of its creators and users?

scraping the complete contents of ptweb doesn’t just steamroll its privacy restrictions and abandon its oft-vital author commentary, but in fact removes the tunes from an even more vital social/community context; engagement with which is the only way of really understanding them (/engagement with which is also our “alternative type of openness”, I think)

screenshot of one of my tunes on ptweb. chiptune archivist, if you’re reading this, I’m sorry I wrote three thousand words “about you”, you actually didn’t do anything That bad. I just have to seize an opportunity when I see it

online chat as community archives

with all of this in mind, a certain popular internet refrain has begun to trouble me the last few times it has crawled back into my feed: “it sucks that so much important information is only kept on discord servers these days. instead of being searchable you have to join and then talk to people to find what you’re looking for. this stuff should be on wikis or forums instead.” I rarely see dissenting comments on these posts. it leads one to believe that the only reason that stuff is Not on wikis is that everyone is too lazy or careless to do proper documentation. (for my part, I love to do documentation and would do it all day given the opportunity.) I would like to propose an alternative (if not universally applicable) point of view here.

a whole lot of pxtone knowledge and resources exist not just exclusively within our discord, but within and “between” us, its members. this demands that people seeking deeper understanding do more than search a wiki, more than join a server and search a message log, more even than ask a question and await an answer. they must build rapport, integrate themselves, make their own contributions (which we aim to always meet with enthusiasm and support). the answers to their questions can’t be documented on a wiki because they depend on our knowledge of the asker, the time of asking (i.e. the state of the art), and the people who are around to freshly discuss and respond. instead, we preserve that information by keeping the conversation going, by producing new works, and by inviting new people into our community.

here’s an example. one technique that is re-discussed and creatively iterated on with some frequency is “multisampling”, which by our definition specifically involves dynamically switching between multiple samples of the same instrument in different registers to reduce the artifacts that result from re-pitching a sample too far from its original pitch. (oops, I’m using “artifacts” in a very different sense here than I will for the rest of the post. for this paragraph read it in the sense of “artifacting”, i.e. the emergence of distortions or anomalies as a result of digital processing.) the ideal parameters of this depend on several contextual factors, and our perspective on the technique and how best to apply it varies over time and from person to person. if damifortune, who pushed the technique to its then-limits in his 2020 tune “I’ll never bid you farewell”, posed a question about applying multisampling in a new work, I and others might launch into a reflection on our most recent usages of it, comparing notes and evaluating the best approach for some new boundary-pushing iteration. this conversation itself may well advance the state of the art. it will be open-ended and somewhat saturated with jargon. some people will witness the conversation, attempt the technique themselves, and then might be involved in the next iteration of the conversation. on the other hand, if a relatively new member asked about reducing those re-pitching artifacts, someone would likely be around to evaluate the context and give the advice that will be the most helpful to them. they might link a more straightforward example that expresses the fundamentals of the technique. they might recommend that the musician use a ptvoice (patch for the built-in synthesizer, which does not have the same re-pitching issues) instead, if they know that one is available that matches the desired sound. does this make any sense — the best answer for any question can vary a lot, but we make and keep it available by continuing the conversation and putting it to practice. recording this kind of thing in a wiki would be unwieldy and would probably leave a lot of people with answers that don’t really help them.

there’s another advantage to keeping our information local to this trusted social context: in fact, it is the very foundation of that “too good to be true” thriving collaborative culture of our scene. it allows us to share our work and knowledge freely without worrying about who might use it and how, and to expand and iterate on each other’s knowledge and work without fear of stepping on toes. put another way, it is the most effective method of insulating us from bad actors/action — from “drama” and rudeness and hate, from malicious plagiarism, and, these days, from nft-minting or ai-training or whatever future technology may crop up to exploit our work without our consent. I think many of the people who sign on to those “keeping information on discord sucks” posts would agree these are real concerns!

screenshot from the pxtone discord. if you want to hear what else this endless iterative chatter amounts to, check out the latest PXTUNES compilation at pxtunes.bandcamp.com — if it’s not obvious from everything I wrote here, I am really quite proud and fond of my fellow pxtoners and the stuff we’re making

too good to be true?

a few objections to this line of thinking stick out to me as worth exploring in particular.

A: though I have tried my best to frame our culture as free and open in its own way, I can see how some people would see it as closed-off, inaccessible, even “paranoid” or “xenophobic” by the most uncharitable reading. this is fair. I (and we) have gone to certain efforts to make our community open and welcoming to all newcomers, even those who might often feel unwelcome in internet spaces or struggle to do the kind of socializing that we “demand”. these efforts have not always been successful but I like to think that as I learn and gain experience, they are improving. perhaps the greatest ongoing challenge is distinguishing those bad actors from the rest. I have taken some pages from the “just ban bigots” book, which I find very insightful and quite aligned with many of my experiences communicating with internet troublemakers. but I also find that it is in some ways designed for a larger, more unwieldy group than ours, and that it does not quite encompass every troublesome type. mostly we have benefited by keeping the population small and “focused”, in the sense that almost everyone who joins is legitimately interested in the topic. I have tried to keep this under control by advertising the server in only a select few adjacent spaces, not treating growth as a priority. I get the feeling that this is counterintuitive for some people but I really do feel that it has produced good, healthy, slow growth as opposed to quicker but out-of-control bloating, which is liable to attract exploiters or, at best, lots of people who care so little about the space that they are perfectly happy to make a mess of it and then get kicked out.

B: yes — all of this preservation activity is dependent in some ways on a for-profit company continuing to exist and not make decisions that will interfere with it. “but what happens to you if discord goes down someday” is a pretty easy argument and one with some merit. I believe that the community and much of its resources exist independently from any specific technologies we use, and that it could flourish somewhere else if necessary. it’s really all about the people. a migration would be a little painful though and there might be some little losses and setbacks. but putting all that aside, ANY place on the internet is subject to such failures in one way or another. all we can do is gamble on choosing the ones that will be the most usable and durable. sometimes that will leave us vulnerable to the whims of big companies. oh well! big companies run a lot of the internet now, just like the real world. (maybe someday that will change??)

C: it must be conceded that much great computer music throughout history has only survived to the present day via the work of those internet-oriented archivists whom I hope I have not too severely offended or libeled in previous paragraphs. the scrapers, compilers, and rehosters do make a contribution, and it is not an insignificant one. I wouldn’t dream of sullying the names of our most famous mega-repositories of tracker modules, game music data rips and the like, many of my friends being among their contributors. (nonetheless I may do exactly that in “exhibit c”. look forward to that) are we insulating ourselves not just from bad actors but from our only real chance of long-term survival? or can the conversation go on forever? I don’t have the answers to this one yet either. or if I do they’re naively optimistic ones that I’d better not publish.

ok, enough about pxtone.

screenshot of pxtone collage. I have never once uttered the words “enough about pxtone” before. PLEASE email me about pxtone


instructions for a performance

(this section is about music made “for computers”, i.e. the computer is the instrument that plays the music)

for me, thinking about preserving computer music has caused a lot of tricky questions to unravel; questions about what an “item” of music is, what artifacts it consists of, which of them should be preserved, and how this preservation can be judged as maintaining (or not) “authenticity”. different kinds of music are exposed to these questions in different ways. there are cases where preserving a recording seems adequate, and others where additional items of documentation or context are vital, and yet others where the live-, present-ness of the performance is so critical that some fixed artifact can never really preserve its essence. let’s talk a little about classical music first to explore this premise.

musical artifacts

the composition of classical music, at least up through the beginning of the 20th century, produces one central artifact whose preservation represents a preservation of the work as a whole (though there could be other items of much interest) — that is the “manuscript”, the final and definitive inscription of the musical contents of the work, direct from the composer’s hand to the page, probably signed and dated by the composer on its final page. any published editions that follow are, in a sense, derivative of the manuscript and represent a dilution of the composer’s engraved intent, vulnerable to external meddling by editors and publishers, etc. (is this a radical way to frame this? I’m truly not sure. if there’s literature on this topic I’m not really up on it. please email me if you are).

a manuscript, of course, does not make sound, and music is mainly accessed through the ears. well, you could wave a manuscript around in the air and make swooshy sounds, but they would be the wrong sounds (or would they???). it has to be translated into sounds via performance on instruments, and these translations have typically been regarded as, at best, “interpretations”. there is no single definitive auditory derivation of a manuscript. I think it is useful here to frame manuscripts (and musical scores in general) as “instructions” for a performance, and assume that these instructions are bound to be read and interpreted differently in different contexts. after all, some of the most important instructions are just italian words meaning “agitated” or “sweet” or “lively”, words which can vary widely in meaning. performance is just subject to too many dimensions of individual bias/expression (and, perhaps, some uncontrollable random factors). performers who dedicate their lives to the study of manuscripts and historical performance practices can still land on pretty widely divergent interpretations of the same piece. occasionally this is the cause of some controversy!

so if we are unable to produce the sound Precisely As The Composer Originally Intended, can the work really be considered “preserved” “authentically”?? we really can’t treat that as an ideal because it is simply impossible, even more so the further the work is removed from us in time. now things are getting interesting…

sometimes the instructions in a manuscript might even be “wrong”, as read in the present day — check out this lovely article about beethoven’s oddly fast metronome markings and the conspiracy theories that have developed around them. pictured, mostly unrelated: the “DONKEY METRONOME” feature on the nintendo dsi metronome app, taken from footage by Tiny Cartridge on yotube.

beyond the artifact

good news: some people believe, in various contexts, that preservation is not really about preserving a certain artifact in precisely its original form forever. sarah norris (after m. jadzinksa), in “toward an ontology of audio preservation”, proposes that (and I really don’t know how culturally reductive this is) preservation methods can be contrasted between “east” and “west”, where the west is chiefly concerned with that physical conservation of the original item, that hopeless battle against entropy; and the east instead prefers to preserve the content or “spirit” of the original in ways that, holistically speaking, may be more durable, though the neglect of the physical artifact may well result in irrevocable loss of a certain type of information. in establishing the eastern approach she produces the interesting and oft-cited example of the ise jingū shrine in japan that is traditionally rebuilt every twenty years. the original materials of the shrine, as shaped by the hands of its original builders, were disposed of long ago. but through this periodic refreshing, the original methods of construction are passed down through successive generations of artisans, and the shrine is kept “alive” more profoundly than if the original structure was instead carefully conserved over time, growing more fragile and slipping into history. norris also examines walter benjamin’s early 20c work concerning reproduction of art, which asserts that the original, physical work possesses an authentic “aura” that simply cannot be reproduced; I think this exemplifies the western approach. I have found this paper and this model very instructive in defining and interrogating authenticity (which I am about to do!)

let’s keep the scope limited to classical music a little longer. it’s not hard to argue that continued performance of a piece of music, even if its precise form becomes a little more and more distant from the “original intent”, keeps the music alive in a way that is accessible and durable. it keeps the music in people’s ears and minds, cements the composer’s legacy. a model of performance as preservation accepts that the parameters of these performances will change and evolve, insisting that the preservation lies in the simple act of performance.

let me take a couple of steps further, perhaps over the line. first I would propose that this evolution might even be desirable, preservation-wise. think about “the composer’s original intent” more broadly: what if we assume that x long-dead composer’s intent for their symphony was “to move an audience to tears”. audiences these days aren’t so impressed by the same things that moved them to tears hundreds of years ago. so is it really a problem if a piece is performed by a larger orchestra than would have been standard at that time? if some elements of interpretation are anachronistic? via these changes, we might come closer to “preserving” the effect that the piece originally had on its original audience, as its composer intended. that sounds like an authentic performance to me.

second, I would like to consider that musical activities beyond performance could also be modeled as preservation methods (here’s where we’re kinda tying into “exhibit a”!). a professor of mine often liked to quote the composer luciano berio: “The best possible commentary on a symphony is another symphony.” I would suggest that the composition of new music that is inspired by, iterates on, or replies to existing music has the same preservation properties that I have assigned to performance. it facilitates a vibrant and active discourse around the older music that allows it to persist through time more durably and completely than if it had remained unsucceeded and unchallenged. that is, “the best possible preservation of a symphony is another symphony.”

(my current boss(-kinda; I have been contracted to catalog and arrange his manuscripts and other historical items (!)), America’s Most Wired Composer tod machover, recently revamped his ~30-year-old boundary-pushing piece/experience “brain opera”. in 1996, its incorporation of music generated by the audience playing on bespoke instruments and submitted over the internet was awing; in 2022, a preview of “brain opera 2.0” saw performers instead interacting with an audio-generating AI in an equally awing live call and response. having witnessed the recent iteration and skimmed much of the critical response to the original, I can say with some authority: brain opera preserved! nicely done (why then did I spend all that time putting things in folders?? the real preservation work was happening right under my nose…))

but I’m getting ahead of myself here. the point is: music that takes the original form of “instructions” is inevitably subject to variation in performance. this stands in the way of the ideal of preserving a definitively “correct” or “authentic” audio representation of that music, so we have to think about other methods and ideals for preserving this music (like some of the things I’ve just suggested. but maybe there are other ways too!)

some “classical” music is actually meant to exploit the fact that its instructions will result in a different sound every time they are performed/interpreted. john cage is famous for pioneering this sort of thinking. his “imaginary landscape no. 4” is performed on 12 radios (pictured here: a performance by hmkv at dortmunder u centre for art and creativity). the performers change their volumes and tunings at defined intervals, but the actual audio-content of the piece depends entirely on what is being broadcast at the time and place of the performance.

computer instructions

check this out: computer music also takes the original form of “instructions”. they’re just encoded instructions to a technological performer, rather than notated instructions to a human performer (I owe this metaphor partially to the national archive of australia’s “performance model” of digital records, which describes the display/playback of any type of digital record, not just audio, as a “performance” of a “source”). one might assume this much more readily allows for a definitive correct performance according to the composer’s intent (which makes preservation of the music “easy”, right. you just have to keep the instructions and/or the recording), because computer software and hardware are engineered according to the principle that the same inputs or instructions will always produce the same outputs. as I gestured at in “exhibit a”, I have seen that assumptions of absoluteness like this one often drive conversations and practices around preserving computer items, especially in the niches I’m focusing on here. they’re risky assumptions! as we’ll explore next…

(disclaimer reprise: major focus on old game music in this section. just because I know a lot about it and it opens up some interesting philosophical dimensions (you’ll see))

first of all, the path from instructions to our ears can be a little more unpredictable than one would think. there are multiple intervening translations: computer software/hardware “follows” the instructions to generate an electric signal, which might go through an amplifier or some other post-processing stages, and then is translated into air vibrations by a speaker. if I wanted to get really detailed I might say these vibrations are further “processed” by the particular properties of the vibrating medium and its context (e.g. the room the speaker is playing in) before it reaches the ears. all of these extra steps are in fact pretty tough to standardize and maintain.

the instructions-to-signal step requires a precise configuration of hardware and software that serves as translator, and over time that hardware will be vulnerable to physical decay. emulation is an alternate option, by which the software is recreated to run on newer, more readily available hardware, but emulation that perfectly reproduces all the quirks of the original hardware is elusive, maybe even impossible in some cases. I remember hearing rumblings, though I can’t find documentation of them now, that every 2a03 (the computer “soundchip” in the nintendo entertainment system) has slightly different behavior as a result of their cheap production. at the very least there are many subtly distinct revisions of the hardware. probably no way to definitively emulate that! this kind of thing must be why enthusiasts of such hardware prefer to make “hardware recordings” of tunes, capturing the audio signal directly from the original machine even when reasonably accurate emulators are available for free and produce “basically” the same sounds with much less work. there is just some uncaptureable aura in the original sound, hiding in minute details that we may not even consciously notice (hmm, that sounds familiar!).

the speaker-to-ear step is no less tricky. consider this: how is a game boy soundtrack “meant” to be heard? through the little built in speaker, or through headphones? what kind of headphones — the kind a child might have had in the early 90s? (should you be in a car on the highway while listening?) these options will have different frequency responses, among other differences, which might mute some details and accentuate others. the following is pure speculation: the “kirby’s dream land ii” and “mega man ii” (game boy) soundtracks have always sounded kinda harsh to me, listening to them via emulation on nice headphones. in kirby everything is super articulated, instruments that fade away in an instant and leave constant little silences (a little tiring for the ears), and a lot of the timbre is kinda buzzy and uncomfortable. mega man (among some less subjective issues with weird dissonances and mysteriously poor intonation) stays in a really high register almost constantly. lately I have wondered if these elements were meant to compensate for shortcomings of the speakers/headphones that the music was likely to be played on; extremes of articulation and register to ensure clarity of rhythm and pitch.

in college I attended a guest lecture by alexander overington, who was in charge of sound for some pretty significant podcasts. he told us that in sound-engineering his podcasts he tried to target common listener playback scenarios, including something like “iphone in a cup on the bathroom sink while the listener is showering” (I might be embellishing this a little). this philosophy is far from universally applicable. I think much commercial music is engineered to sound good under “ideal” playback conditions, and the listener left to approximate those conditions as well as they can afford, though this seems to be subject to a complex and shifting relationship between “artistry” and “industry” that I will not attempt to explore here. but! in a case like this where probable playback conditions are known, it makes sense to me that the music may be optimized for them.

my good pxtone pal willthur sent this message to me totally unprompted as I was in the midst of drafting this section. he’s talking here about the “metroid hatchling theme” from metroid ii on the gb (screenshot here from Tarosan’s longplay on world of longplays). I take this as independent confirmation of some the above speculation! and it also happens to be quite poetic and affecting. subtleties of computer “interpretation” can alter the expression of this music in ways that interact meaningfully with its original context…wow

alternate performances

the whole “crt pixel art” phenomenon of the last couple years has also really opened my eyes to this idea. (for the uninitiated: some internet posters, foremost among them the @crtpixels twitter account, brought to greater attention the notion that pixel art in old video games was probably actually meant to be viewed on a blurry crt screen, not a new sharper monitor where individual pixels are distinctly visible. many pixel artists quite clearly used lines and colors in ways that the crt would distort into images of apparently greater depth and detail than the console’s video hardware was really capable of. wild…) some official soundtrack releases of old game music add reverb to the original audio, perhaps most memorably the 1997 cd release of the music from pokemon red & blue. in 2010 konami also did this to only the chiptune soundtracks (that is, the nes ones, the orig game boy ones, the mega drive one, etc., all the ones made with a synthesizer “soundchip”) in the “akumajo dracula best music collections BOX” set of castlevania soundtracks on cd. I used to scoff at this kind of thing and at “smooth pixel” filters in emulators. why would I want an ahistoric distortion of the original work, rather than one that presents all its details in total clarity? is it not poor preservation to obfuscate the original artifact in this way? these days I still don’t like listening to chiptunes with added reverb, but I’m developing some new feelings about it. maybe the reverb softens some of those harsh edges that were never really meant to be heard in the first place (as much as I have come to love them). I think we have to start thinking of these not as obfuscations but as alternate performances, and investigating their unique strengths and weaknesses. again, iterative and interpretive “performance” could be a key means of preservation here! (I would love to philosophize at length here about popular youtube uploads of nes game soundtracks with added stereo panning, reverb, or bass boosts but I should be moving on…)

maybe certain alternate performances are a bridge too far, or can be more definitively classified as “inauthentic”. but some of these wrong performances might represent the authentic experience of certain listeners. many pc games of the late 80s through the 90s had soundtracks in “midi” form, which really only instructs the computer on which instruments play which notes when (and at what volume). different midi devices, which were purchased separately from computers, would perform these instructions pretty differently; many of them had their own unique banks of instrument samples, so while they could all “play this note on a violin”, their violins often sounded distinct. a few of these devices were pretty widely regarded as standard (and the tunes were probably arranged with those devices in mind), but none were ever quite universal. so, surely, some people have fond memories of their particular hardware’s non-standard interpretations.

some early nintendo entertainment system emulations incorrectly performed that sound chip’s “triangle wave” sound, failing to reproduce the “aliasing” that was sort of a technological side-effect but nonetheless contributed to its distinct timbre. many people of a specific generation played nes games for the first time on these emulators and have some nostalgia for that particular inauthentic performance! in fact, some early chiptune musicians recorded their tunes using software that had the same emulation flaw — that is, the music was meant to be heard that way, which throws a wrench in our sense of authentic emulation. hmm, when we start thinking about “intent”, the gap between that idealized authentic original expression and the listener grows ever wider and murkier…

(I realize I’m going on about this for a while but really I just want to thoroughly shatter your notion (if indeed you hold it) that “preserving” means and ONLY means producing and retaining a single, authentic, definitively “correct” recording. creation is squishy, creators are squishy, music is squishy, computers are squishier than you think they are. “facts” in art are ever elusive! so we have to look for “truth” instead, even (especially!) when it comes to “preservation”…)

searching for intent

consider this: the sound hardware in the earliest model of the sega mega drive produces a distinctive distortion that some have dubbed the “ladder effect”. famed mega drive composer yuzo koshiro swears by the ladder effect as part of the authentic mega drive sound. yet in the same breath he admits that when he was writing mega drive music on his nec pc-88 (which had comparable but not identical sound hardware) he could not anticipate where and how the ladder effect would come into play when the music was sent over to the mega drive, nor did he prefer to tweak his compositions after hearing the results. are these tunes meant to be heard with or without the ladder effect? what players of “streets of rage 2” heard would depend on whether they bought a model 1 mega drive prior to the game’s release, or a revised model 2 later on. “what the composer heard when they were working” is usually a good target for authentically reproducing the “intended sound”, but this particular case confounds things.

in a neat parallel to the ladder effect story, see this photo of video game artists working with both a sharp-pixel monitor and a “preview” crt; image via, and many more interesting details at, vgdensetsu

consider also the recently-trendy “restoration” of super nintendo soundtracks. composers/sound programmers for the super nintendo borrowed instrument samples from keyboards and samplers, but those samples were reproduced by the super nintendo’s sound hardware with reduced fidelity. savvy audio-archaeologists are now hunting down the original “sound sources” for these soundtracks and recreating the music with higher-fi (but otherwise, “the same”) samples. the result is, I dunno, weird but somewhat compelling. these are “faithful” performances in an altogether different way. (I wonder if they are more faithful to how listeners remember the games sounding when they played them many years ago; or if they produce a response more faithful to how those tunes once made them feel??) many have spun these alternate performances as “what the composers wanted their tunes to sound like, if only the hardware had supported it”, or “what the tunes originally sounded like before they were compressed to fit on the cartridge” (read any news coverage of this topic for examples). in fact I am not sure how ahistoric these claims might be, but they smell a little funny to me. they could really only be verified by interviewing any number of people involved in each game’s sound production work, who may or may not still be alive and may or may not remember anything so specific.

I mean, we’re lucky to ever get detailed development info of the sort koshiro provides. it’s a somewhat rare thing. some of these composers have been interviewed once or twice, but many others haven’t. some that published behind-the-scenes details on the net did it on forums or BBSs that are now inaccessible. and unlike handwritten manuscripts, those encoded computer-musical instructions have much artifactual evidence of creator intent scrubbed away by computer optimizations, and precious few initiatives (with precious little resources) exist to locate and preserve e.g. pre-compilation computer files — which, if they were even kept by the musicians or their companies, are surely on storage media that are now approaching or beyond the end of their useful life — or other artifacts like notebooks. (some present-day chiptune musicians don’t even want that stuff available to the public. their techniques are often a significant part of their claim to fame and they may want to discourage imitators; I remember at least one pretty public conflict over reverse-engineering of such techniques.) I mean, no one then and few people now take this music seriously enough to imagine that any of those artifacts would be worth holding on to. we’re talking about an apparently secondary element of an already pretty trivialized art form!

anyway, one more interesting and similar example: some clever oldgame hackers/explorers discovered a few years ago that the iconic crunchy slap bass sample from the sunsoft game “gimmick!” actually suffers from a sound programmer’s mistake. correcting the “bit order” of the sample would have allowed the original hardware to reproduce the sound much more clearly. certainly this restoration feels more authentic than those other ones?? though some commenters still expressed a preference for the original sound, and one purports that the music is clearly designed around the distorted version. maybe none of this actually matters, because demo recordings exist for the gimmick soundtrack that would seem to be a more direct expression of the composer, given access to less restrictive equipment. it’s an interesting shift in perspective to think of the takes that appeared in the game — the ones that the most people have heard — as a derivative or subordinate expression of the music. there’s a unique sort of philosophical push and pull here between “writing for the computer, with the intent of exploiting and stylizing its unique properties”, and “conceiving of Music then wrangling the limitations of the computer to represent it as faithfully as possible”, sometimes spanning multiple iterations. perhaps most of the time both of those things are happening at once.

apparently-definitive demonstrations of the composer’s intent, such as published soundtracks, often tell interesting stories. the bulk of the soundtrack cd for the famicom game “mother” consists of unique arrangements with live performances and lyrics, followed by a final track “the world of mother” that compiles much of the music as it is actually heard in the game. other games of this era got comparable releases branded as “arrange soundtracks”. sometimes the in-game chiptune versions never made it to a CD (until the classic vgm rerelease craze of the 10s and 20s (ongoing) when, I guess, interest in those versions finally reached the threshold of a blip on the profitability radar). is one to assume that the in-game versions are, in terms of the process of creation, technological “reductions” of the music as it was imagined? no such assumptions necessary when it comes to the game “leading company”; the in-game music contains precise transcriptions of the improvised solos performed on the associated album “keys of the city”, so these jazz tunes must have been recorded first, and the (multiple, different) computer versions then transcribed from them.

some of the leading companies in the arcades in the 80s and 90s (including sega, capcom, taito, konami) had sound teams that would also perform/record the music from their arcade games as a live band. some of those same people were responsible for rearranging that music when those games were recreated for less-powerful home consoles and even-less-powerful handhelds. to me it is obvious that many of these musicians did exercise some artistry across every configuration of sound hardware they had to use, that they were often “creating computer music” at the same time as “computerizing music”. (I have long held that the arrangements of “shining force ii” tracks made for “shining force gaiden: final conflict” on the famously audio-underpowered sega game gear have an exquisite minimal beauty of their own… but again, no one cares about that stuff!! a subordinate iteration of a subordinate aspect of a subordinate art form. too many little beeps! surely no one would choose to make music this way except under duress. I swear you’re never ever going to get shining force gaiden: final conflict music on a cd. if it ever happens you can email me and I’ll buy you a copy.) I guess this is another question of “artistry” and “industry”, which really confounds things for me. the philosophy and methodology of archiving artifacts of industrial processes seem quite different from that of archiving artifacts of artistic processes. which sort of process is “producing an arrangement of the music of “gunstar heroes” that is much reduced in complexity so it can play on the game gear, because there is a market for game gear games and we’d better take advantage of it”??

this is all treating the music in isolation from the rest of the game, too, which might be a mistake. surely the visuals and interactions play a role in the listener’s experience of the music. there’s a reason those game music live orchestra concerts always have a big screen behind the players. in the early days when channels of sound were at a premium, it was common for one part of the music to cut out when any sound effects were playing, which in some games is like all the time. how can these incidental, interactive remixes be preserved?? many emulators for old sound hardware allow for individual sound channels to be disabled, which is a start. but the precise behavior (what sounds are overridden and when) depends on individual games’ sound engines (and may be yet another reflection of composer intent: which channels are the most important, and should always cut through sound effects?). and the overall, integrated effect, where rapid player actions set off bursts of visual and aural excitement, momentarily overwhelming the computer’s ability process graphics and music, an act of spontaneous collaborative creation between player and computer — demands emulation of significantly greater scope and precision. (and we’re not even getting started on music that is Supposed to be dynamic…but broader theories about preserving interactive media art are a little out of scope here so I’ll leave it at that.)

I guess what I’m trying to argue with all of this is that preservation of this music demands open-minded discourse, continuous creation and evaluation of new and diverse performances, and interrogation of all sorts of contextual artifacts (if we can even get our hands on them!) and the possibly conflicting narratives they suggest about the music’s creation and its most authentic forms. it’s a music that asks us all kinds of unique questions about what to preserve, how, and why. to fall for the “single correct answer” trap that machines lure us into would be doing harm to future listeners’ ability to fully understand and appreciate it.

for this year’s “ray’z arcade chronology” rereleases, taito went as far as to record a soundscape of classic arcade games at their akihabara arcade “HEY” that optionally augments the in-game audio (this image of the “virtual sound map” shared by @gosokkyu from a ZUNTATA NIGHT stream). now that’s what I call preservation of original playback conditions!! more of this please


the hobbyist problem

(this section is about music primarily documented “on the computer (internet)”. it happens to also be about vgm but it probably could be about some other niche musics…)

it is interesting/important to note that most of the initiatives that exist right now for preserving vgm share most of these characteristics:

* they’re run by enthusiasts, hobbyists, volunteers (whatever word you find the most charitable) with no formal training in preservation

* their complete holdings/contents are completely free for everyone to access over the internet

* they present information transcribed directly from official sources as well as significant original research/inference/translations on the part of volunteers

* they deal with music that has been published digitally or on physical media (sometimes readily available for purchase today, sometimes long out-of-print) as well as music that was never published except as a component of the associated game

I want to make it real clear that I do not look down my nose at those enthusiasts/hobbyists/volunteers. rather, I think they have a unique and valuable perspective and deep and varied skills; and I acknowledge that the immense work they have collectively done has shaped the course of my life and rescued countless lost musics and untold stories. I have nothing but gratitude and respect for them. however! I think they have had some issues or blind spots in navigating the problems presented by this unique set of characteristics, as I’ll illustrate…

hobbyist vgm preservation has a documentation/sourcing problem

the final music paper I wrote in college was about the use of motives in kazumi totaka’s score for the game boy game “kaeru no tame ni kane wa naru”, mostly known in english as “for the frog the bell tolls”. “mostly” is an important word there — the game was never released in english, and in fact nintendo has published very little information about it in english officially (just the names of a few key characters and places, in crossover/cameo appearances), such that consensus on the correct translation of the title has not quite been reached. in one miiverse post about the game’s main character appearing in smash bros for wii u, smash director masahiro sakurai referred to the game as “the frog for whom the bell tolls”, which seems a somewhat less graceful if more direct allusion; some take this as the definitive answer. many fans often shorten it to “frogbell”. an unofficial fan translation exists, which draws some names from official publications and provides original translations for others. a little confusing!

the situation is significantly more confusing when it comes to the game’s music, which was never published officially at all (except a couple tracks on a couple compilation CDs), let alone in english. this leaves those enthusiast-archivists — who record or rip music from the game, edit and tag it, and upload to some online repository for distribution — totally in the dark. in researching for my paper I investigated nine different rips/uploads of the soundtrack across multiple sites in both english and japanese, each of which had some features distinguishing it from the others. the different uploads disagreed on the titles of tracks, the ordering, which music to include or exclude, whether to divide certain pieces into multiple tracks or merge them into one. (and I’m talking about some Real titling discrepancies. in one japanese upload the track that is usually just “end credits” or “staff roll” is called “mille-feuille magic”, where’d that come from? all the english uploads agree the final boss music is called “slay the snake!” but clearly someone just made this up at some point because the two japanese uploads have two other, quite distinct titles.) a startling number of uploads included tracks that appeared in no other upload. and when I watched a video of a complete playthrough of the game, I identified a few further jingles that had been left out of every upload.

the uploaders can’t be blamed for this; without a soundtrack publication or in-game sound test, their only option is to derive an original tracklist based on the context of the music’s appearance in the game. this practice has a long history and has largely been uncontroversial. but in more extreme cases like this one, the uploader’s decisions have a non-negligible impact on how the listener/researcher understands the music. and many of the most prolific uploaders mostly make good decisions, to my mind! but what I’d like to see is careful accounting of these decisions and broader awareness of the subjectivity.

(yes, “decisions” is the right word — there’s often nothing objective about this process! even when there is a seemingly technically sound method of retrieving all the music from the program data, e.g. some kind of internal table that points to the memory locations of each tune, often important bits can be left out. in the case of kaeru, I speculate that certain short jingles missing from certain uploads are internally classified as sound effects, not music, and would be neglected by a method like this (which might seem unimportant, but in my case these overlooked jingles actually contributed significantly to my argument!). the credits sequence also dynamically strings together a few discrete looping segments. I think some uploads, derived from such a method, failed to recognize this bit of context from the game and list the segments as separate tracks. when my pal lssq made a video about the search for the secret “totaka’s song” in this game and others, many many comments said something like “why not just look through the files to find it?” well, not every piece of music is A File, man!! in kaeru no tame, totaka’s song plays after you listen to one area’s music for like 4 minutes without leaving… the uploads also disagreed about whether/how to include this.)

some sites are good at this; for example, vgmrips rips come with .txt documentation that often identifies the source of the track titling/ordering, at least. but when these rips are reformatted and make their way elsewhere on the internet, possibly reaching a much larger audience, this documentation is often lost. and as these rips are duplicated and reduplicated, an instance of the music that is in some ways the product of just one person’s interpretation (see “exhibit b” for more on this) takes on false authority and blinds listeners to alternatives — again, some of the most popular uploads of the tunes from kaeru no tame are missing quite a few little bits of music that happened to be very valuable to my paper. (in fact, some recordings have emulation differences that actually affect the audio itself… more on this in “exhibit b” too. my go-to recording while I was researching my paper has all kinds of little volume attenuation pops and clicks that I later learned are maybe not supposed to be present, or at least not so audible.)

(the one area where kaeru no tame presents fewer problems than some other games is attributing the music to a composer. the game’s credits confirm that all the music is by kazumi totaka. lots of older games list obscure aliases in their credits, or even credit a vague larger entity like “capcom sound team”. other games might credit multiple composers but not publish any information about which tunes are the work of which composers. sometimes soundtrack cd releases clear this up; other times an enterprising researcher reaches out to someone involved in the game’s production that can provide that info. on vgmdb there are talk pages for each release that sometimes, but not always, detail the source of track attributions, which is nice. but again this attribution info can get reproduced across sites without any context or a link back to the source… in any case, vgm heads seem to LOVE to track down/speculate about this stuff. I get it! I’m definitely interested in composer histories/styles…)

what does irk me is that the somewhat (and I don’t throw this word around lightly) pretentiously named “Video Game Music Preservation Foundation” website is a major offender of this sort. tons of assertions about not just tracklists, recordings, and attributions, but also composer biographies and development details that are often totally unsourced. totally barren discussion pages, edits are rarely annotated… is this really preservation?? on some prolific editors’ user-talk pages they admit to just making educated guesses at some of these things. but that often isn’t made clear on the pages that present those guesses! and certainly on the pages dedicated to some topics I’ve done a lot of research on (e.g. ryu umemoto) I can spot some dubious and some outright false info. I can’t deny the scope of the site and the amount of work that has clearly gone into it. it would be a deeply useful resource if I felt like I could rely on any of the information to be verifiably true! kinda disappointing, and emblematic of a broader problem. I’m sorry, I really don’t mean to like Start Beef here. but someone had to say something.

as funny as I think it is to see certain internet groups calling themselves communities, “foundation” is truly on another level, wow. you will not catch me naming anything I am responsible for a “foundation”. please become a member of the jaxcheese musical research foundation by pressing the follow button at the top of this article. anyway this image is the header of the one of the oldest pxtone sharing sites, which was pre-ptweb and mainly for the japanese scene, inserted here because “collective” is Such a better word. oh also I wanted to mention here that some of the record labels leading the aforementioned “vgm rerelease craze” are also doing this sort of work, which is interesting. they’re not hobbyists; they are subject to additional restrictions and priorities like “how many tunes can we fit on a vinyl record”; and they have varying degrees of cooperation/information from the originating company and/or composer. somewhat of a different story!

hobbyist vgm preservation has an access/copyright problem

a couple of years ago I was introduced to the work of hiroyuki mizuno when his soundtrack for the 1999 visual novel “ano, subarashii o mou ichido” was ripped and uploaded to vgmrips. he was before then mostly unknown on the english-speaking net; I think this upload and a few others that happened around the same time across a few sites are primarily responsible for his becoming more known outside of japan. I found it to be really compelling music! he wrote a couple of really thoughtful and charming fm synth soundtracks for games on the old sharp x68000 computer and continued to write fm synth tunes even as his studio moved on to making games for more capable computers (a man after my own heart), eventually producing a lot of really sophisticated and awing jazz/fusion-style stuff. I tracked down and bought some CDs of his music, some soundtracks that were already circulating online like ano subarashii and some others that weren’t. when I find music I like, I do this when possible as a means of supporting them (or even, if the CDs are long out of print and only available secondhand, just a private means of showing respect for the value of their work), and also because I am an incorrigible collector.

the soundtrack to ano subarashii in fact comes on a dvd containing music files as well as a “manual” with commentary from the composer and concept art. the manual opens with a series of greetings to original purchasers of the dvd, secondhand purchasers, and illegal downloaders (the latter reading simply “buy it”); and a pretty extensive letter describing the studio’s feelings about piracy. purchasers of the dvd are specifically instructed to make copies only for personal use and to destroy any copies upon reselling it. this kind of thing feels different when it’s coming directly from the composer, right? I truly always try to support musicians however I can, but I can’t deny that I discover and download a lot of stuff through dubious channels — in this case, channels that the composer specifically disapproves of. if it weren’t for the vgmrips upload, though, I don’t think this music would ever have made its way onto my radar, and I never would have bought the DVD. how do we reconcile this? (spoiler I don’t really have an answer to this, I just want you to think about it.)

I think a lot of listeners, researchers, and are happy to totally ignore copyright. maybe it smells a lot like an oppressive arm of the publishing industry, and it seems like it would be progressive to reject that (I dunno, maybe it really is!). but reframing it as a matter of respect for the artist and their work makes me think about it differently. a lot of the sites doing good vgm preservation work right now have escaped legal enforcement of copyright by mostly flying under the radar and by following a “just upload it first, then respond to complaints if any come in” approach which seems to be pretty tried and tested. and this is definitely a good thing! as much as anyone else, I don’t want these resources erased by big corps flexing their legal muscles, and I do appreciate the productivity, greater coverage and broad access that is enabled by their cavalier policies… I just, I dunno, think it might be a little more complicated than that?

like, what about cases where rips are being shared of a soundtrack that is also available digitally through a channel that directly supports the artist (e.g. many old sega soundtracks that are lately available officially on bandcamp via data discs). vgmpf cautions its editors against uploading audio that is directly pulled from official soundtrack releases, but there’s no reason to assume that audio otherwise ripped or recorded from the game is not ultimately precisely the same item (especially when it borrows the same tracklist from the official release) — purely a philosophical difference. sure, these rips have unique research value (again see “exhibit b”), but many many downloaders just see this as a repository of free music to listen to. khinsider has plenty of unique, original recordings from games that never got soundtrack releases; but it also has tons of uploads directly from commercial CDs and digital downloads, many still in print, mixed in and often indistinguishable from the rest! does the so-free-and-so-easy availability of so many of the masterpieces of this art form reinforce its marginalization as Not Real Music?? (tough but important question!! which I am not prepared to answer either way) if we respect this stuff enough to spend all this time preserving it, it’s worth thinking critically about how to respect its creators! and maybe in a post-capitalist society paying respect to artists would not be so tangled up with paying money to artists. but we’re not there yet!! (I’m not really an expert on how to get there though…)

the gamer/internet-young-people crowd (which includes many of the people who work on and use the resources I’m talking about) seems often to reject this kind of notion about copyright and access. last year leading video game preservationist frank cifaldi announced that he works with wata to first dump the contents of unique prototypes etc. before they are processed, but under their agreement cannot release these dumps online in their entirety. (wata is a kinda-reviled collectible video game “grading” company that encases pristine games kinda-irretrievably in plastic slabs to “preserve” their value for collectors.) an indignant and voluminous response forced a couple of threads of justification out of him. this is really instructive stuff and I was glad to see that it seemed to get through to some people. access is a priority for preservation! but I think it’s easy for people accustomed to the open-access paradise of today’s internet to see unrestricted and instantaneous access as The Only goal, the only priority, the only qualifying threshold of preservation — if I can’t see it now it’s not preserved. whereas my recent archives training and testimonies like this have shown me that it is but one of multiple competing priorities (privacy! conservation! resource allocation! and so on). plenty of archivists seem to hold a healthy resentment for copyright and restrictive copyright agreements that hold back open access, but they recognize the reality that they simply must push back and forth with these things, make compromises, if they are to continue their work and be productive.

I mean, I don’t want to look like a killjoy/narc/oppressor here… I think my being artist and archivist gives me a useful sort of double perspective on this. to me it doesn’t seem a total fantasy that copyright can encourage creativity and protect the livelihoods of working artists — and thus protect their legacy in a way that will have a positive outcomes for future access! yes, it is often mis-wielded by big corps to the opposite end. and in the short term it is an obvious frustration to archivists. but I just can’t find my way to seeing it as righteous and progressive to toss the whole thing out, at this moment. I can’t imagine that with less regulation, the same people with the power to exploit copyright maliciously would not simply exercise that power even more absolutely…

in a kinda similar case earlier this year, over the course of several threads and hundreds of little branched-off discourses, one favorite writer of mine tirelessly hashed out a related access/copyright debate (originally “is it ok for ai to train on copyrighted images” but eventually something more like “is copyright actually good for the average artist”) with a whole mob of twitter commentators. I found this really fascinating and a little dispiriting to witness…a lot of these folks seem to have a real cynical, incurious, antagonistic attitude about the relationship between artists, their artworks, and the public. I think it boils down to a similar sort of internet-era entitlement-disguised-as-freedom: “the only right thing is if I can do whatever what I want with other people’s art now”. it’s an attitude that seems a little pervasive in this internet vgm preservation scene, and one that I think could hold it back from achieving greater legitimacy and productivity if it is not carefully interrogated!

photographic evidence of my purchase of the ano, subarashii o mou ichido soundtrack. which again is really really good. you should find your own copy… I can only share mine with you if you come to my house and listen to it.

am I doing this better or worse than everyone else

I had to wrangle with these questions when I started working on my ryu umemoto fanpage, which aims to document all the works of the pioneering visual novel composer. I’m happy with some of my decisions and still uneasy about others. my solution to the sourcing problem was just to link and/or cite sources as often as possible! the page has some original and speculative research, like track attributions I guessed at by ear; I have tried to label these clearly and explain my reasoning. of course, often the sources I’m citing are secondary themselves and not clearly sourced, so sometimes I’m just passing the buck in terms of reliability. but I try to compare as many sources as are available for confirmation.

I’m a little less happy with my solution to the access problem and still rethinking it as I continue to work on the site. there’s a pretty even distribution of games with and without official soundtrack releases; and out of the latter category, a mix of digital releases, CDs that are still easy to find, and others that are long out of print and/or were produced in very small numbers. but as of right now, regardless of official release status, I’m linking out to rips/uploads available for free on the internet. the composer having passed away tints the situation in an interesting way…because he is no longer around to benefit directly from sales, it could be argued that to expose more listeners to his music is the better way to respect his legacy. and I like to imagine that a source like this on a relatively niche topic does not steal any sales, since a lot of its visitors are unlikely to even know about the music before their visit, and it may in fact convert some of them into paying fans (and I do include purchase links and/or info about CD releases when those are available!). but this all kinda flies in the face of what I wrote before about respect. I can’t deny that I’m being a hypocrite here. but, on the other hand, I’m not the uploader, I’m just linking. and if there are sources out there it seems silly not to link to them in the name of completeness! passing the buck again. as long as everyone is happy to pass the buck, though, these entrenched problems just get perpetuated without closer examination…


the ordeal of being documented

(I guess this is section is about music published “on the computer (internet)”, rather than “computer music” per se. maybe it’s not even really about music but about creative things in general! but there’s plenty of overlap between those categories and this felt like an good/important final segment. bear with me once more)

where I live, at least, cultural awareness and rejection of the collection of our personal data by big corps and governments has steadily grown in recent years. and last year there was a pretty high-profile internet conflict focused on a forum dedicated to compiling information on certain targets to facilitate stalking and harassment. to me the ultrapopular voluminous youtube documentaries on the Internet Crazy Person of the hour are precisely the same thing (often with the same subjects, if that tells you anything), sold as entertainment to a broader audience. on the yet-smaller scale, it is a pretty common practice now for people in conflict to compile and publish private exchanges and embarrassing old postings in order to discredit their opponents. (sometimes they deserve it! (?) I dunno. I don’t wanna complain about “cancel culture”. I’m definitely not talking about celebrities or politicians, for one thing…) the cons of publishing information about myself on the internet have seemed to me to be approaching the threshold of outweighing the pros. so again I will question the internet preservation ideal of preserving as much documentation as possible on as many subjects as possible and allowing for access as publicly as possible…

of course, “publishing personal information” is not really the same as “publishing music”, though music is often deeply personal… all I can say is that over the last ~five years, a handful of computer musickers that I admire have tried to erase all evidence of their work as a result of some personal conflict or other. because in all of these cases the music was only accessible in the main through a handful of internet outlets, they were all able to do it pretty thoroughly. some would later restore some of the deleted tunes; others remain totally erased. some were engaged in feuds that aired out a few too many personal details a little too publicly, and deleted to escape humiliation or danger. one, I think, simply faced occasional crises of self-esteem. and one was the subject of one of those “broadcasting as many private details as possible of someone else’s disaster for my personal gain is fine because we all agree they deserve it (don’t harass them though I guess)” multi-part youtube documentaries, and deleted to escape the harassment campaign that somehow resulted — a campaign that benefited significantly from the thorough documentation and wide sharing of their personal info. functionally speaking it’s not hard to argue that preservation of a sort is happening here. these posts and videos are, after all, among the only remaining documents of this artist’s existence. it’s preservation that majorly Sucks though!!

the question is: should this self-deleting behavior be discouraged or even disallowed? if repositories exist that back up the work of these self-deleters, is that good or bad? how should we feel about efforts to restore access to an artist’s work that go against that artist’s wishes to be left alone and forgotten? does it matter how historically important and influential the artist/work is; i.e. do influential artists waive their “right to be forgotten”?? I think these are tricky but important questions for preservers… for my part I retain the work of these favorite vanished artists of mine in my personal library and will share it privately with people I deem trustworthy. I see this as a safe, local (community??) sort of preservation activity that does not go against the artist’s wishes. it’s the compromise I’ve settled on, but I’m still not really sure about it…

w. boyd rayward’s commentary on h.g. wells’ utopic “world brain” concept — an “encyclopedia” that would centrally collect (and preserve!) all human knowledge — deftly exposes its inevitable spiral into fascist fantasies of global subordination, of a system that would track every individual and (per rayward) “provide the information necessary for the suppression of dissent and diversity.” art, to my mind, is, always has been, and must always be a form of resistance against fascism. so where global, permanent, forcible documentation is a tool of fascism (it’s hard to argue against this, right?? even the most vapid and superficial depictions of fascism like 1984 include this), it seems an essential component of art that the artist be allowed to erase their work, to un-document themselves. or perhaps to exhibit and preserve their work only through the methods they deem appropriate, and only within trusted communities, as I explored in “exhibit a”. is this too radical?? perhaps time will tell. but it seems clear to me that at this moment, people are much slower to interrogate preservation in this way when it comes to computer- and internet-borne artifacts. I think this could come back to bite us (if it isn’t already sinking its teeth in…)

wow this is the second time I have criticized 1984 on this blog. I don’t even hate it that much. I guess it’s one of those things like electroswing that is just fun and rhetorically useful to make fun of. by the way, none of this article applies to electroswing, I don’t care if gets preserved. it would probably be for the best if humanity forgot about electroswing actually. by the way I ran out of ideas for images so here’s another screenshot from the JagOfTroy longplay of teleroboxer on world of longplays


whew, made it!! this one sure was a lot of work but I’m glad to have it all out of my head and on the page, even if not so many people will see it. and looking forward to writing something a little shorter and maybe more fun next. I hope you will look forward to it too.

as I was wrapping up writing this blog, a minor controversy erupted on the net over (specifically sega game gear) sound emulation in the new “sonic origins” compilation of classic sonic the hedgehog games. people uploaded side-by-side comparisions of audio from sonic origins with the ““original audio””. I wonder where that original audio comes from?? my guess is it’s just some other emulation that has somewhat arbitrarily been established as more authentic. I almost tried to work this into “exhibit b” but, man, “many such cases”!! if I tried to write about em all this thing would have been even more unwieldy than it is. but I thought it deserved an honorable “current events” mention down here. these arguments have leaked into popular games discourse for years and will continue to do so…

one of the things I like about writing this blog rather than trying to be a real academic writer is that I can Engage With The Literature as little or as much as I want and make lots of claims that are only anecdotal and possibly dubious. I did cite a lot of things here (more than usual!) but there also plenty of bold, insufficiently supported assertions… I realize this one is pretty provocative and I would really like to talk with you about it if you have opinions. so one last time, Please Email Me at jaxcheese@gmail.com, or comment or otherwise contact me. I’m @jaxcheese on everything except pokemon go where someone else got it first.

many thanks are owed for this one. foremost thanks to my draft-readers sydney, who encourages my writing tirelessly, and sidedishes, who listened to me ramble about many of these ideas before they were organized at all. thanks to the pxtone community for its deep grace and creativity, and again to our community archivist sidedishes, and to daisuke amaya who started it all; and thanks/apologies to the chiptune archivist. who cared about our music and gave me many things to think about. thanks to my professors at the simmons department of library and information science, especially profs pratt, bettivia, and botticelli, whose lectures and assigned literature informed a lot of my thoughts for this one; and to prof mason at wheaton who showed me many new ways to think about music in general.

thanks also to all the authors whose writings/tweetings I cited: kimberly christen, sarah norris, heslop/davis/wilson at the national archives of australia, alexander overington, crtpixels, yuzo koshiro, bradsmith, frank cifaldi, doc burford, and w. boyd rayward. major thanks to the composers of all the music I referenced, whose music never stops charming and inspiring me. thanks to all contributors to vgmrips, vgmdb, snesmusic, the hoot archive, khinsider, and even vgmpf (and many other precursor sites lost to time), who I believe, despite my little grievances, are doing vital, pioneering work which is certainly the foundation for everything I do (and anything I will ever do) in this area.

finally, respect to those composers who left the net and never came back. your work will live on in my heart forever.

(p.s. I’m sorry I forgot to make a “hardly know her” joke in this one, the right opportunity never presented itself I guess. uhh, “computer music?? I’ve never even heard her music!”)




he/him * vgm composer * drummer/violist * pxtone proponent * librarian in training * fan of turtles, sci-fi, jazz, mega drive, string qtet, etc. * jaxcheese.net