I've been teasing this for a while. On the record, I am blaming NaNoWriMo for my being slow at actually publishing it, but off the record, the right way to frame everything so it made sense and wasn't overwhelming was simply proving elusive. I would love to receive feedback on this if anything here is weird or doesn't make sense because there's... several disparate concept coming together in this article.
This article discusses the way the meta of managing a music library shifted from 1992 to 2002, how Sony had a hand in creating that change, how Minidisc as a format adapted, and how users (in the USA at least) used the format itself differently in 1992 and 2002. (strap in: ~3300 words)
The music and audio world (and the media world at large, including written text and video) when minidisc was introduced was almost exclusively physically oriented. People bought pieces of media and listened to them. People would buy or borrow vinyl records and dub them to tape. Or bought and listened to CDs and commercially pre-recorded tapes. One piece of physical media tended to correspond to one piece of.... logical media. Whether a specific album or a mixtape, a tape full of voice memos, a movie on a VHS or beta tape, or a tape full of recorded episodes of The Simpsons. Multiple pieces of logical media sharing a piece of physical media was possible (especially with VHS: you could fit like 6 hours of video on one of those tapes in the long play mode) I couldn't tell you how common it was outside of home taping of TV, due to the run-time of the tapes and the potential for automation of the recording.
Minidisc wasn't a big leap. In the 1990s, when you recorded a minidisc, you probably spent some time preparing and/or you'd bought or borrowed a CD, and then you spent some time dubbing, and potentially splitting and labeling the contents of the disc. (even if you use a computer as the source) By the end of the decade, minidiscs had fallen enough in cost as to be nearly negligible. If you bought five CDs a month, a five-pack of new minidiscs wasn't a significant ask. With all that work going into it, I tend to use the term "author" even for discs that weren't an original composition or mixtape.
At some point, it became possible (typically through accessories that were likely costly) to transfer CD-TEXT to a minidisc, but CD-TEXT itself wasn't released until September 1996, and was vanishingly uncommon, so if you bought like fifty CDs during 1999, it was likely only a couple of them had it.
This is all doubly true if you don't spend the money on a high-speed dubber of some type, because otherwise, minidiscs could only be authored in real-time, and prior to around 2000, titling and editing could only be done after recording was finished. Even with a high-speed dubber, the process will take proportionally longer if you're pulling tracks from multiple CDs.
All that is in service of some of Minidisc's best features: As with CDs your discs can have multiple tracks. You can re-arrange tracks once they're recorded, you can have non-contiguous tracks if a particular mixtape is an effort over the course of some time, and you can add text labels to the discs and tracks, to quickly identify media. So, even though it was tedious and annoying to do, there were benefits to actually doing editing on a disc.
In addition: Minidisc's success in Japan in particular is often described as being a direct result of the fact renting a CD is significantly cheaper than buying one, so Japanese minidisc users took all of this a step further by not having the original source CD anymore, so a Japanese minidisc user can't (easily) re-dub from the original CD, even if they wanted to.
Once you made a disc, you might label it or write the track listing down on a J-card, and then you probably kept it and it'd enter your rotation of different discs you'd cycle through. You'd go buy more when you wanted to make more mixes or when you'd bought more CDs you wanted to use on the go. I know all of this because I have MDs from the the 1990s and very early 2000s from both Japan and the US and the same usage pattern is largely true for both, but there's always exceptions.
Even once Sony introduced the MDLP format extension in 2000, to make a minidisc, you had to dub it from a source. High-speed dubbing hardware (such as CD-MD decks, bookshelf stereos, and boomboxes) was common everywhere outside the United States, but not here. Any US user of the format was significantly dedicated to the cause as they were recording up to 80, 160, or 320 minutes, sometimes adding track-marks and titles by hand, usually on a portable device. (even set-top decks with PS/2 ports for keyboards and control and titling were uncommon compared to portable recorders.) (This will come in handy later: ATRAC3, the new codec used as part of the long play modes is pretty good, Sony claims the 132-kilobit "LP2" mode sounds as good as the old 292-kilobit "SP" minidisc mode and for my part: this is accurate. The 66-kilobit "LP4" mode is passable but most minidisc machines shipped with cheap tiny earbuds similar to the ones from iPods, and different people notice different things.)
Now that, with MDLP, each disc is at least twice as much work to make: once you go to all of the effort to record and edit a disc, you're significantly more likely to leave it, label it specially for what it is, write-protect it, and go buy more discs. For minidisc, the cost of new discs is is significantly less than what it would have cost in time to re-do the dubbing and titling on, say, 3 CDs on an LP2 disc. Although considered unaffordable in the early '90s, the cost of discs did eventually come down: even Sony's own discs were $2 a pop or less by 1999 or so. You could likely get third party discs for less.
All of that is... until June 2001, when Sony announced NetMD, a new option for writing audio from a computer to minidisc at faster than real time. NetMD makes very heavy use of the MDLP modes and the ATRAC3 codec, for 160-320 minutes of audio per disc.
With NetMD, you use Sony's (or a compatible third party's) jukebox software to rip CDs or convert MP3s or WAVs and organize your music collection, and then send the audio to minidisc at significantly faster than real-time. The first few NetMD machines could write LP2 audio at 16x real-time and LP4 audio at 32x real-time. In trade, Sony limited the number of times you could write a track to a disc. Their software referred to these as check-outs - you could regain one by using the software to delete the track off the disc.
Between MDLP encouraging much longer discs, how much faster NetMD (especially via MDLP) authoring is, and the check-out restrictions, it no longer made sense to have a very large library of discs, or put a lot of effort into presenting them physically. I see a lot of what look like "someone's complete MD ecosystem" on eBay (and have bought one or two) and most NetMD-era setups include a recorder that someone used for a number of years and an single five or ten-pack or discs. Sometimes the labels aren't even applied, or the discs might be numbered, like, one throuth five.
This all solidifies the comparison of late-era minidisc to early file-based digital audio players. Similar overall capacities, removable media, and a presence as an extension to a computer audio ecosystem.
This all part of what I'm now calling File Orientation (he said it, the name of the post!) - an ecosystem where the music is organized and exists primarily as files on computer storage media. The core idea here is the same whether your files are MP3s you drag onto an SD card, AAC files you sync onto an iPod, or ATRAC3/3plus files you write to minidiscs or a memory stick. The computer, not individual pieces of media, is the center of your ecosystem.
ATRAC3 itself was introduced a couple years before MDLP and NetMD, in September 1999, as part of the announcement for the NW-MS7 - the first Memory Stick-based Network Walkman. The NW7 is relatively unassuming by modern standards. It looks a lot like Sony's modern voice recorders. It took a full-sized memory stick, had a built in rechargeable battery, and a screen to display the details of the current file. This should sound familiar: you load audio onto the NW7 by ripping CDs or converting MP3s using Sony's jukebox software and then copying the files to a memory stick, which consumes one of three "checkouts" in the jukebox software.
At launch, an NW7 was the equivalent of around $430 (in 1999, which is almost $765 in October 2022 dollars). That's thoroughly "early adopter" territory. Though, the original Minidisc recorder cost at least as much when it was new in late 1992.
At first glance, "this music player can hold like 2 hours of 66-kilobit music, or one hour of good sounding music" isn't very exciting. Especially given supplementary memory sticks would have cost hundreds of dollars at the time. In theory, you can buy and use multiple memory sticks. (In fact, the original Memory Stick iteration even had a reasonable amount of space for labels)
But, the NW7 was, again, thoroughly an early adopter product and as with many early adopter products, the limitations are the point, or at least not a deterrent. In 1999, you can use the NW-MS7 as a way to get a taste of what portable digital audio would've been like a few years in the future, even if there was some compromise at first. Thinking about that, I don't believe Sony ever legitimately imagined people buying Memory Sticks in bulk, and 1:1 replacing MDs with MSes. Instead, (opposite of what I asserted in the MZ-R910 post) Sony knew the future of portable digital audio was centered on computers.
While a 64-megabyte memory card may have cost "hundreds" - hard disks were rapidly coming down in price. In mid-late 1999, you could get a 20-gig IDE hard disk for $500. A bare IDE disk would've cost a bit less. Bigger disks existed, but cost proportionally more because they were considered specialists.
And, you can fit a lot of ATRAC3 audio on a 20-gig hard drive, so I think the idea is you're supposed to just constanly be refilling one or two memory sticks, as opposed to having a bunch of different sticks. This is... different from the old way. On one hand: loading a 64-meg flash stick, even over USB 1, should be faster than writing an MD, tape, or even a CD (at least in 1999). On the other hand, if you're away from the computer for more than a day, 64 megs of music could get old fast. You only get 2 hours of audio at the lowest quality setting.
As ever, technology evolved, things got better rapidly, and a strategy emerged. Flash got bigger, cheaper, and more types of ATRAC (and ultimately, MP3) players appeared. Here's a sampling of what you could get just three years later in late 2002, here in the US, from Sony:
- $130 D-DJ01CK ATRAC/MP3 CD player
- $230 MZ-N707 NetMD Minidisc Recorder incl. dock/remote
- $299 NW-MS11 Memory Stick walkman incl. 128-meg stick
- $350 MZ-N1 NetMD Minidisc Recorder incl. dock/remote
By this time, I imagine Sony's overall strategy was you could pick what tool you want to use based on the context or your budget. If you're doing really intense exercise, it may be worth spending the extra on a Memory Stick walkman, and the limited capacity might not bother you. If you're doing moderate exercise, want to cheaply carry a lot of music, or want to be able to record things in the field, you can get a Minidisc recorder. If you're on a limited budget or want even more audio and you don't need all of minidisc or memory stick's shock protection, you can get an ATRAC CD player and burn CD-R or CD-RW discs of music. (mobile CD players had gotten Really Good by the early 2000s and typically had decent shock protection, this should've been even better with ATRAC or MP3 files, plus CD-R and CD-RW media was incredibly cheap at the time.)
That is a lot of flexibility. The next step (and one someone already did) is hard disk based mobile jukeboxes.
Apple's October 2001 launch of the original 5-gigabyte iPod (1,000 songs in your pocket!) was, in retrospect, a vindication of the computer as the Digital Hub of the Computer Era and (for better or worse) a better execution of both the jukebox software and the idea of a high capacity portable audio player. Sony eventually followed the iPod with hard-disk based ATRAC players of their own, but continued the multi-prong approach. (Disclaimer: I have no idea what the UX on Sony's HDD-based ATRAC players is like.)
Apple ultimately added branches to the iPod family, following some of the same concepts as the early Network Walkmans. The iPod Shuffle was a low capacity device built around the idea of filling it with random files out of your library. The iPods mini and nano were up to a couple gigs apiece and were focused on being budget friendly and/or smaller versions. The nano was the first iPod based on solid state storage, for example, so Apple also framed it as a workout buddy where you might use your big iPod on the home stereo or in the car, as well as a cheaper version you might give a kid.
iTunes itself didn't have most (any?) of the DRM restrictions Sony's jukebox software did. You could copy the files at will, convert them to any other format iTunes supported, copy them onto more than one iPod. Just about the only thing iPods couldn't officially do is let you copy music back to a computer. (Although: iPods themselves can be used as generic removable storage so if you're at your friends house it may make more sense to copy the files there, which in retrospect makes the one singular restriction much funnier, to me.)
So that's File Orientation. The ultimate impact for minidisc is people didn't bother to buy lots of extra discs. Sony did its best to control the redistribution of files, likely at the behest of the record label attached to the electronics company. I rarely see NetMD machines accompanied by a large number of discs.
Eventually Sony relented on some of the DRM SonicStage enforced but ultimately the ways NetMD, file orientation, and OpenMG/SonicStage had changed how people managed their music libraries and treated minidiscs had already happened. Even if you could now write unlimited copies of Boulevard of Broken Dreams to a minidisc, nobody did. I know this because I've bought up people's entire NetMD ecosystems. People who bought into NetMD and weren't already heavy MD users had ten or so disks at absolute most.
I often talk and write about Sony making its own decisions, which impact how minidisc sells. Sony chose not to market minidisc. Sony chose not to bother selling much, if not most of the minidisc ecosystem here in the US, and it found its niche regardless. Sony chose to enforce its vision of a computer-oriented music ecosystem via NetMD. You could still dub CDs. The first round of portable NetMD machines the US got all had analog/optical line input and editing conrols.
But nearly nobody did - NetMD was too big a convenience factor to pass up. The format must have done relatively well in this time -- well enough Sony ultimately built several budget-oriented models just for the North American market with no audio line inputs at all, making minidisc as a format significantly more comparable to "a weird MP3 player". Sony was right about the way things were going, but their extremely late acknowledgement of MP3 managed to make them a niche player in modern portable audio. Sony ultimately moved ATRAC to the back seat and years after killing off minidisc, removed it entirely from new Walkman products.
HiMD in 2004 intensified all of this. The 1GB discs were still cheap but at $7 a pop, not disposable or, say, cheap enough to merit buying extras to give to your friends. Portable CD players had gotten really cheap and really good, so they were imminently reasonable as a primary on-the-go music format. People with computers could quickly, easily, and inexpensively (pennies a disc) copy CDs at high quality, if they were worried about ruining an original copy of a CD or if they still wanted to, say, make a custom mix. (ATRAC/MP3 CD players also existed in a bunch of different form factors and 650 megs of MP3s is functionally nearly indistinguishable from 1 gig of MP3s.)
Flash memory had gotten bigger and cheaper (although HiMD was still cheaper, a 1GB CF/SD/MS card was around $70 in 2004-05) and the capacity on HiMD was so high most people only needed one. I'm not convinced anybody (read this as "most people") using HiMD as an on-the-go music player ever bought more than a single 1GB disc. I have seen very few anecdotes talking about actually owning multiple (let alone "several") of the 1GB discs. It wouldn't surprise me to find out most people who slotted HiMD in after NetMD just reformatted a few NetMD discs or ultimately used it in classic NetMD mode anyway.
As part of The Start Of The End for minidisc, in 2005-06, some of Sony's last new minidisc machines in the CMT-AH10 (2005, a bookshelf stereo with a HiMD writer) and the MZ-RH1 (2006, a portable recorder) using the newest version of SonicStage gained the ability to rip classic format minidiscs. The community took this as something Sony should have put in to start, but I'm pretty sure these things were meant to be used by the Japanese market as a transition tool, a way to rip large minidisc libraries (and remember: no CDs to go with them) to complete the move to newer formats. (Prior to this, HiMD mode recordings could already be transferred from a HIMD mode disc to your computer, Sony even sold a couple models as pro field recorders.)
The last few MiniDisc machines sold a surprisingly long time. The MZ-RH1 (the ripper) was sold globally until 2011, possibly alongside one of the last Japanese-market NetMD machines, and decks and stereos sold for a few more years, with a TEAC CD/MD dubber deck remaining "technically available" until December 2021. (Although all indications are that TEAC stopped manufacturing that unit in 2017 or so.)
But I'll talk in a future post more about HiMD, some of the late-stage minidisc hardware, and more generally about the end of the minidisc platform in general.
