Tag Archives: multiverse

Multiverse, save me: Love Story (Taylor’s Version) – Taylor Swift

In multiverse posts, I examine alternate recordings of songs. Yes, I know I’m a nerd.

For any artist, power – and money-making potential – lies in owning your masters. If you own your master recordings, you can fully control what is done with your work. If you don’t want it licensed for use in this film or in that advert, you can say no. Moreover, the revenue earned by your records goes to you, not to a company that then pays you a small percentage of what they’re making.

Under the terms of her record deal with Big Machine, Taylor Swift did not own her masters. This was not in any way exceptional; few artists actually do. Young musicians, signing their first contract, are negotiating from a position in which they have no power; granting ownership of master recordings to a label is almost always a necessary price of signing a record deal.

In 2019, Scooter Braun, manager of Justin Bieber, Demi Lovato and others, bought Big Machine. The deal to buy the label included Swift’s masters, which the label owned. Braun subsequently sold them to a private equity firm called Shamrock Holdings, owned by the estate of Roy E. Disney, for a sum said to be in the region of $300 million dollars. Under the terms of the deal, Braun continues to derive a percentage of the earnings from the masters. He is, we can presume, a very rich man.

Swift had, by her own account, been trying to buy her masters for a while, and she was furious at the deal. Coverage of the story made the mainstream press, with lots of ink spilled over what it all signified, who was the goodie, who was the baddie, etc.*

Early on in the dispute, Swift signalled her intent to rerecord her old catalogue, or at least part of it, so that she could own the masters to her old songs at last, derive fair value from them, and control how and when they’re used. And so, two years later, we have this: a re-recorded version of Love Story (parenthetically titled Taylor’s Version – that is, not Shamrock’s version). A new version of Fearless, her second album and pop breakthrough, will follow next month.

Love Story in its original form is only 13 years old, and pop music mix topologies haven’t changed as much as you’d think in that time.** Even so, the two versions of Love Story are appreciably different, despite sharing an identical arrangement. This is because the new version of Love Story is, to my ears at least, rather less pop than the original, and significantly more indie.

Over the last year, Swift has worked extensively with the National’s guitarist Aaron Dessner, and from him she appears to have developed a taste for reverb, which the National slather all over their recordings to give an air of profundity to their lugubriousness.

A reverb haze on everything is ubiquitous in indie, but is seldom to be heard in pop music. The density of the mix topologies works against it. OG Love Story is, as these things go, relatively organic, but the wall of sound in the choruses (nine acoustic guitars, according to producer Nathan Chapman, plus nearly as many electrics, as well as fiddle, pedal steel, splashy hi-hats and backing vocals) renders long reverb tails an unwelcome distraction, if a creative mix engineer were actually able to make them audible above the wall of sound in the first place.

Taylor’s Version of Love Story, then, is most noticeably different from the original in its treatment of the vocal. There is audible reverb throughout, which is especially prominent in the first verse. The harmony vocals are downplayed throughout in favour of the lead, and that lead vocal is heavily Auto-Tuned, in a way the original lead vocal surprisingly wasn’t. We’ll get back to that.

The results are, if you’ll pardon the pun, mixed. There are things I prefer about the new version. I like that it sounds more like it was recorded in a shared physical space; the disjunct between the dry vocal and wet drums in the original recording is unpleasant and jarring to my ears. I think Dessner has overdone it with the reverb, as usual, but there’s been an attempt to make the mix coherent in terms of space, and it does work better than the 2008 recording.

However, the praise Swift has received for the greater level of control*** of the new vocal in many reviews strikes me as pretty clueless. Love Story was written when Swift was 16, and her youth is always apparent in the original recording – both in the song’s sentiment and the sound of Swift’s voice. The breathless, barely-in-control quality of the original vocal works in the song’s favour by getting it over some of the awkward turns of phrase: the lines that don’t have enough syllables to fit the metre, the lines that have too many, and the lines where an unnatural word or syllabel is stressed to fit the rhythm. The new Auto Tuned vocal is bland – the robotic sound of it doesn’t match the lyric or emotion of the song. It could have been delivered by any early-rounds American Idol contestant, and would have been subjected to the same brutal processing.

Whether Taylor’s Version of Love Story or any of the other songs from Fearless will supplant the originals – on radio or in the hearts of the fans – only time will tell. Early signs from fans are encouraging, but radio may be a tougher nut to crack in the long term. The Nathan Chapman-produced original was machine-tooled to work on radio, and it does work. The National’s sound – bigger low end, less mid-range info, lots of reverb – is more of a headphones/home-stereo kind of sound, so is less attractive to radio stations whose primary concern is being audible on a small stereo in the corner of a shop or office, or in a car. And those original versions are not like the original cuts of Star Wars that George Lucas pulled when he released his Special Editions. They’ll still be out there, winking at radio programmers who’ve loved them all this time.

Then there’s simple attachment and inertia. Despite Jeff Lynne’s best efforts, DJs are still playing the ELO recordings from the 1970s and ’80s, not the versions he’s recreated at home by himself, despite them being more or less identical. If anyone has the fan power to buck this trend, it’s Taylor Swift, but I remain sceptical that it can be done.

*My own take (simple version): the baddie is clearly Scooter Braun, who comes over as a complete prick.

But here’s the more complex version. Record deals with most labels are profoundly weighted in favour of the label in a way that would be judged illegal in almost any other industry. She’s been shafted, no doubt, and maybe a really sharp manager might have been able to do a better job for her when she was a kid, but probably not. As I said, all young artists get shafted.

However, old school label deals did pay off handsomely for a small percentage of artists, and Swift was one of those. She has an aboslutely gargantuan reported income and a considerable fortune that increases every year. If she wants to play the long game, she’ll have more than enough stashed away in five years to buy the masters from Shamrock, pretty much whatever their price. The only question will be, is she willing to pay it. But there’s a price for everything. The masters may increase in value, but they can’t go on tour, they can’t sell merchandise, and they have no claim on the money Swift makes from her post-Big Machine records, which it’s safe to assume will earn her tonnes more cash.

Re-recording your old songs is one way for Swift to try to regain control of her work, but it’s a very time-intensive way and, long term, there’s no guarantee it would be successful in terms of supplanting the old recordings, especially as she gets to songs released relatively recently, where the differences between an old and new versions would be minimal and the new recordings would benefit less from the curiosity factor that is undoubtedly helping Taylor’s Version of Love Story.

There’s also the opportunity-cost element: what writing and recording work isn’t Swift doing while she’s doing this? Surely cracking on with that will be more fulfiling, and with every radio play and Spotify stream, she’ll be closer to buying back her old masters in future, too.

**There are differences, of course, but a 2008 recording and 2021 recording have a lot more in common than a recording from, say, 1963 and 1976, 1982 and 1995, or 2000 and 2013.

***Her much the “control” comes from the singer’s performance and how much from the mixer is – literally – impossible for the listener to judge. The vocal is Auto-Tuned to within an inch of its life – that much will be apparent to basically every listener. But when mixing in the box, you can do almost anything to almost any musical element in the mix. That includes shortening the length of a given syllable by hundredths of seconds to make it tighter if you choose to. If you’re good at it, you can do it without notable artefacts, too. It’s simply not possible to tell, as the listener, exactly what editing and processing has been employed during the mix, and any critic who makes confident claims on behalf of Swift here is overstepping the limits of what they can know.

Once More into the Multiverse – R.E.M.’s Monster remixed

Warner Brothers’ ongoing programme of 25th-anniversary editions of R.E.M. albums has reached 1994’s Monster. Part of the package is a remixed version of the album. Let’s see what a reconsidered 2019 mix from original producer Scott Litt can do for the band’s divisive, guitar-heavy used-bin staple.

Monster always was quite an odd-sounding record.

Coming out in 1994, it seemed like a slightly delayed reaction to the dominance of alternative rock, most of which up to that point had been based on scorchingly distorted guitars. In truth, it was more of a reaction to inter-band politics. At some point in 1993 or so, Peter Buck had put his mandolin and dulcimer in the cupboard, turned up the tremolo and distortion on his AC30, grabbed a Les Paul and rediscovered the joy of simple, swaggering rock riffs. Drummer Bill Berry had already threatened to leave the group if the next album wasn’t louder than Automatic for the People and Out of Time, and if the band didn’t go out on tour to promote it. R.E.M.’s follow-up to Automatic was going to have be a loud rock record or there would be no follow-up at all.

The band cut the basic tracks for Monster live on a soundstage, and Scott Litt’s finished mix always suggested to me a degree of overthinking. Having the guitars forward in the mix was a good thing, given how crucial Buck’s tone (and on a few songs temolo) was to the sound of the record, and I’d argue that dropping the level of Michael Stipe’s vocal was a sensible thing to do too, but on some of the songs the weight of the guitars pushed the drums so far back that they became tiny. I’ve always felt the masters contained a more energetic and more satisfying mix, with the drums a bit more prominent.

Sadly, Scott Litt’s remix isn’t quite that, and goes a long way to convincing me that what might seem “wrong” with Monster when listened to critically is actually right in a greater, more fundamental way.

We can surmise from Litt’s new mixes that he felt his original mixes left the vocals too quiet and the drums too processed and too quiet. The new mixes correspondingly give us a whole lot more Stipe, and a less polished drum sound.

For evidence of the latter, A-B the intro of I Don’t Sleep, I Dream – the EQ-ing on the toms in the 1994 mix is absent (or reduced), giving them a perceived higher fundamental, and less detail in the range of stick impact; they boom less, and they cut less. Of course, these decisions are personal, but I prefer the 1994 mix as far as the tom sounds go, and it’s not even close. On the plus side, the snare is EQ’d differently, with a less present, less hyped-sounding top end. It’s an improvement.

Unfortunately, on many songs you don’t really get the benefit of it. One of the issues with distorted guitars is the amount of sonic real estate they take up. Monster‘s guitar sound is crazy huge. This necessarily leaves less space for the drums. Perhaps the top-end hype on the snare on the 1994 mix was to try to bring it out against the guitars. In the 2019 remix, Litt goes a different way: he adds more compression, to flatten the transients, turn up the sustain of the drum and position the reshaped snare as a solid block in fixed audibility against the guitars. But he goes rather too far for me. On What’s the Frequency Kenneth, the drums actually feel like they lag behind the beat due to the heavy compression as they fight against the wall o’ Buck and the newly prominent Stipe. They have no transient left at all. I’ve never previously heard an R.E.M. record and felt like Berry was dragging. If anything, he tended towards being a little early. The new mix is, on the loudest songs at least, extremely unflattering to him. The decision to take off the little bursts of tremoloed guitar in the choruses, meanwhile, merely removes one of the song’s best supporting hooks. A strange choice.

Other weird choices abound. The main guitar and drums crushed into the middle on Crush with Eyeliner, while the sides are crowded with clean overdubs and Thurston Moore’s backing vocal is drowned out by multi-tracked Stipes? Definitely odd. Anchoring Tongue with a tom-heavy drum track right from the intro? Yep, strange again. Other choices, such as remixing of the guitars on Let Me In, are just misguided. The whole point of Let Me In is that incandescent distorted guitar sound, presented so ambiently that actual strums are hard to make out. With only a minimum of pick attack and volume change to tell you where the beats were, the guitar sound became disortientating and weightless, but also uncanny and beautiful. The new version sounds all too earthbound, with Stipe mixed so dry it sounds like he’s singing into your earhole from six inches away. Being brutal, it almost suggests Litt didn’t get what worked about the song first time round.*

Of course, this is just a bonus-disc remix, a parallel-universe version (a Bizarro World remix, if you like). It doesn’t replace the actual album mix of Monster. But it does spotlight the choices made by the band and Litt 25 years ago, and reinforce to the non-audio-engineer fan that so much of what we hear when we listen to recorded music is mediated by mix engineers and producers. When different choices are made, the result is a different album.

monster

*Just to prove how subjective all this stuff is, Scott Aukerman and Adam Scott talked about the remix on their podcast, R U Talking REM Re Me? Both preferred the remixes to the album mixes for the majority of songs, and both felt Let Me In is the biggest improvement. To which all I can say is, whaaaaaaaaaaaa?

 

2016 Clip Show Post

New Year’s Eve again? They come round quickly, don’t they?

This year I’ve not been able to devote as much time to the blog as I would have liked, which I’m looking forward to remedying in 2017. Thank you for hanging in there with me this year. I really appreciate that people spend their time reading my incoherent ramblings.

I’d like to leave 2016 behind, if I may, by pointing some of my newer readers back at some of the pieces I enjoyed writing this year.

I’ll be back on Monday. Have a great weekend, whatever you have planned.

Bert Jansch

Farewell to the Glad

The Dolphins – Fred Neil

The musical multiverse – alternate versions, demos, outtakes, mixes

Joni Mitchell from Blue to The Hissing of Summer Lawns

She’s Gone – Hall & Oates

Their Back Pages