Hacker News new | past | comments | ask | show | jobs | submit login

The original article in Scientific American is much better https://blogs.scientificamerican.com/observations/is-pop-mus...

The author of the article highlights a potential bias with this study which is worth pointing out:

But I did wonder if there was a selection bias in play here. The Million Song Dataset, huge as it is, may not provide a representative slice of pop music, especially for old songs. Its contents are heavily weighted to modern music: the database contains only 2,650 songs released between 1955 and 1959, but nearly two orders of magnitude more—177,808 songs—released between 2005 and 2009. That’s because it draws on what’s popular now, as well as what has been digitized and made available for download. And the songs of yesteryear that people enjoy today (as oldies) may not be the same ones that people enjoyed when those songs first came out.




Yeah - this could potentially be a huge bias. Songs that are top 40 hits for a month or so and songs that have staying power to be memorable 40 years from now are probably not quite the same category.

It also echoes my (baseless) assumption that the main reason that we have this idea that music a few decades ago or more was somehow "better" is that anything that wasn't at least above average is completely forgotten by now i.e. survivorship bias.


That's true. Terrible songs (like Sylk-E Fyne's "Romeo and Juliet") are completely forgotten very soon while the really good songs are never gonna give you up, never gonna let you down, never gonna run around and desert you.


90s style and bad lyrics aside I think old hip hop etc like Romeo and Juliet you mentioned or a lot of Akon, dr. Dre stuff etc had a lot more melody and was much nicer to the ears than most of what we have today, even from good (subjective) artists.


People are really still doing this? I can't stand the fact that back in 1998, The Undertaker threw Mankind off Hell In A Cell, and plummeted sixteen feet through an announcer's table.


It was done to make a point in a clever way. I think you should appreciate the level of skill it takes to RickRoll Hacker News and not get voted off immediately.


If you have to point out how clever something is, it was not, in fact, clever.


No. You're thinking of jokes.


Clever.


It's an interesting experience listening to '90s hits now that they are oldies. I certainly enjoy them a lot more now than then, and looking inside my soul I can guess at three reasons:

1. Pure nostalgia.

2. The selection effect.

3. I've grown more tolerant. I now happily (though not ecstatically) bop along to songs which I hated back in their heyday. Ice, ice, baby...

Probably (3) is actually partly nostalgia as well. But it is also that I am not as caught up in the petty jealousies and genre-fan-feuds that I had been as a lad.


4. You haven't heard the song 3 times a day for the past month.


This is a big one for me. When someone listens to recent pop on the car radio I want to rail against how cheesy and annoying it is. Then I'll be the one driving and flipping stations and I'll land on some 70s or 80s pop tune and happily sing along.

I realize that it's no less cheesy or annoying but it's familiar enough yet I've not heard it in a couple of years. I can get down with a lot of music but hearing the same damn mediocre song every week, it goes from "eh, whatever" to "dear god why is this still around??"


The BBC regularly repeats old episodes of Top of the Pops, a music chart programme that was broadcast between 1964 and 2006. It's quite close to a random sample of weekly Top 40 listings - about half the episodes in the archive are unsuitable for repeat broadcast, because they were presented by notorious sex offenders.

There was a lot of crap in the charts, but I think the overall level of quality was higher. I think the reason for that higher quality was essentially negative - the routes to an audience were much narrower, so artists were practically required to get into the charts to sustain a professional career. A far broader range of artists were playing the major label game, because they didn't have many alternatives.

"Pop" is a much narrower term in 2018 than it was in 1978, because musicians don't need chart hits or radio airplay. Digital recording and distribution have created a much more fragmented music market, because artists can connect directly with audiences rather than relying on a very narrow funnel of publicity and distribution. Good music is somewhat more difficult to find today, but I think there's more of it and more definitions of "good".

https://en.wikipedia.org/wiki/Top_of_the_Pops

(proxy required outside the UK):

https://www.bbc.co.uk/iplayer/episodes/b00704hg


maybe also because in the past you had to be a real musician to actually play something (no hardware to help you). So many were trained in making music (instead of self made) ? Not that being trained equals quality, but somehow, maybe it helps... (IANAMusician :) )


Pop music of the past had plenty of people lip-syncing their concerts, house bands that played all the instruments, and so forth.

It is not obvious that you had to be a trained musician to succeed. Many who succeeded were, but many who succeeded are not.


I think another reason is that pop music has been reduced to a formula to some extent, and that's exactly what TV shows like the X-factor are looking to capitalise on.


Almost everyone thinks the greatest music ever made was made in their late teens and early 20s, and it has always been that way. I think I'm somewhat unusual in that I spent a lot of time DJing in my 20s, so I developed habits around constantly searching out new music, but even so -- I'm in my early 40s now and I still reach back to dance music and rock music from the late 90s quite often to listen to. I've just resigned myself to the fact that late 90s trance music will make me happy in ways that no new music ever will.

As to the topic of the post --

In music production, you can think of the 'stage' available to the listener as a box. Pitch goes up and down, stereo pan goes left to right, and volume goes front to back. When you're producing, you can't have two sounds in the same space without them merging together in the listener's ear. In the early days of stereo music production, producers generally tried to place the sounds as if you were standing in the room, watching the band perform, and were fairly conservative with giving everything space, so that sounds wouldn't clash. As people got more sophisticated with production (think Dark Side of the Moon), producers started experimenting more with the listeners since of space -- surrounding them and enveloping them with sound -- but still generally trying to capture the sound of physical musical instruments. Even when they used synths, they tended to try and make them sound like "real instruments".

Once digital music production started becoming more widespread and sophisticated, producers started abandoning the pretense that they were faithfully recording the live performance of instruments and focused purely on sound as sound, and the new tools allowed them to perfectly place all the sounds next to each other on 'the sound stage' and take up all the available space. Which, if you're filling up all the space 'up front', as it were, necessarily means that the song is going to sound louder, if that's what people want to do.

Now, there are plenty of genres of even dance music who play with dynamics more -- think Deep House, for example -- even there, they're going to make the bass as loud as they can, because they still want to make speakers boom for the dancers -- and they can push them as loud as they can be, very precisely because ableton or whatever makes it simple to do.


Producers often use reverb for front to back space as well.

I with agree that its got more precise, typically when producing a kick for techno I make sure that the power is exactly around the point where the sub is going to be able to play it back with the most impact.

Also with tools like Neutron we can make sure there is no overlap between channels.

I am probably a similarly age to you, but I hardly ever listen to anything more than 5 years old apart from as a learning exercise or if there are people in the same room as me.


>It also echoes my (baseless) assumption that the main reason that we have this idea that music a few decades ago or more was somehow "better" is that anything that wasn't at least above average is completely forgotten by now i.e. survivorship bias.

Where does the idea that music output is equally good in all ages comes from?

The fine art world, and the literary world know very well from historical experience of centuries that there are periods of huge masterpieces and lots of great creators, and periods of drought. Heck, the classical music world knows that too.

Why would this not hold in the pop music world?

Music is another activity that is historically influenced by period, trends, culture, etc, and all of these have their ups and downs.


I think it's fair to consider that our current view on classical music could be due to survivorship bias as well. Consider all of the composers who were lost to time because their peers or musical historians did not consider their music noteworthy.

Whenever I see discussion on classical music, it's a fun mental game for me to imagine it's a forum for metal instead, and these (assumingly) well-to-do academics are all metalheads who have a strangely archaic way of describing the specific merits of songs.


>I think it's fair to consider that our current view on classical music could be due to survivorship bias as well. Consider all of the composers who were lost to time because their peers or musical historians did not consider their music noteworthy.

I don't think that's much of a concern in practice. We celebrate lots of works that people didn't find particularly noteworthy at the time but were appreciated at a later age (Bach, for example, was out of fashion and nearly forgotten for a century or more after his death, Satie was not especially appreciated at his time, there are lots of other examples). In other fields too, e.g. consider Van Gogh.

Most of the works of obscure composers are known to fans of classical music (and musicologists), they are still occasionally played, but are still considered ho-hum.

I'd say good works tend to rise to the top, even if they're not popular at their time.


I think your comment highlights the biggest take-away from the article. Finding statistics to fit your narrative is a trope in journalism, XKCD has even pointed it out: https://xkcd.com/1845/

I think one way these articles could be more honest is if they did not compare the survivors to all modern top 40 hits, but rather to the most curated selection of culturally-relevant and generation-defining songs. I'm sure any avid listener to any genre/time period could find you 10 songs that will make you reconsider your perhaps preconceived negative opinion on that genre/time period, and 10 songs that will make you continue to generalize the genre/time period as low-quality music.

But this is non-trivial and prone to more bias, so we'll probably never see it happen. In the mean time we can continue to use this example of pop-statistics in journalism to remind ourselves to be wary of what we read.


I’m surprised it’s only two orders of magnitude. Compared to today, it was insanely expensive to record music professionally more than 20 years ago, especially when the recoding is more than a single microphone (or stereo pair) capturing a complete live performance in a single take.

There would be orders of magnitude more music published today because the tools are so cheap and available.

What you can do today with $2,000 of equipment is better in every single metric than what you could do in 1990 with $20,000 of equipment. (I’m excluding the costs of studio space here for simplicity, as that can be highly genre dependent: on one hand the cost of concert halls for a symphony orchestra probably hasn’t changed much; on the other hand new electronic music doesn’t need a physical studio at all.)

I don’t know about the economics of recording back in 1960 but I can say with confidence that even the absolute best equipment found in the top studios would be technically inferior to today’s $2,000 electronics. (Though they did use their genius to innovate sounds which we still love to emulate today, so those old studios may well be creatively superior. But we have all that creativity in the form of plugins and effects today.)


>I don’t know about the economics of recording back in 1960 but I can say with confidence that even the absolute best equipment found in the top studios would be technically inferior to today’s $2,000 electronics.

While this might be true for recording media (e.g. a DAW with high end SnR vs 4-track tape overdubbed to death), we still use e.g. the same microphones as back then, and they're still expensive. Ditto for all kinds of analogue musical instruments (guitars, pianos, etc).


>Ditto for all kinds of analogue musical instruments (guitars, pianos, etc).

Quality guitars, amps and microphones are vastly cheaper, because of the astonishing improvements in automation and far-eastern manufacturing. A $400 Squier Vintage Vibe strat is 95% as good as an American Fender strat from 1990, at about 20% of the cost in real terms. Budget condenser microphones simply didn't exist before the mid-80s, so you'd need to spend about $2,000 for a decent LDC mic; today you can get a very good LDC from Rode, SE or Aston for about $300. The same goes for headphones, monitors, outboard and all sorts of other gear - it's just ludicrously cheap in historical terms. A grand piano is still an expensive item, but you can get a beautifully detailed and utterly convincing multisampled piano plugin for under $100.


>"we still use e.g. the same microphones as back then, and they're still expensive"

Good microphones were almost prohibitively expensive in the old days. Today you can buy something like an SM57 for €100, and that's normal retail price, no special offers or anything. Second-hand, they're usually less than half that.

Decent acoustic guitars can be had for less than €200, brand new. Sure that won't be a fancy name-brand guitar, but it'll sound good and play well.

All kinds of great outboard gear like the FMR Really Nice Compressor is €200, and it's one hell of a piece of gear, even at twice that price. Combine with an inexpensive mixer with decent preamps, and you're already well on your way. Yes, even Behringer is decent these days.

And you can run a pro-grade DAW like Reaper on an ordinary run-of-the-mill PC. An audio interface with a good amount of input channels is going to cost a bit, but if you're OK recording one or two tracks at a time, a Focusrite Scarlett 2i2 is €120. And even inexpensive studio monitors have astoundingly good sound quality, compared to what people used to swear by.

You can get really far on a budget nowadays.


True, but today's cheaper gear and instruments are much much better than back then, if even they existed. I keep reading nothing but positive comments about this brand (no affiliation whatsoever) http://kaminstruments.com/index.htm

Also cheap electronics makes diy solutions more affordable. One could mike a drum set using cheapo low end Behringer dynamic mics paired with a self built preamp to get very good results for the money.


If you would like to hear what was possible in the 1960s with creative genius, I would suggest this:

https://en.m.wikipedia.org/wiki/Tomorrow_Never_Knows


Revolver was their best album. I can’t wait for the modern stereo remix.


Further down in the Scientific American piece, one of the researchers does respond to this:

>Serrà acknowledged in an email that a bias due to the “test of time” effect is possible but argued that its influence should be small. For instance, he noted, the long-term patterns and trends that he and his colleagues identified also hold over relatively short—and relatively recent—time periods (say, 1997 to 2007), where the “test of time” effect should be minimal. “The same happens with close and not-so-recent time periods (e.g., 1960 and 1968), where both years could partly incorporate such an effect,” he wrote. “Since the trend is consistent in short time spans where you assume the ‘test of time’ bias is minimal and, furthermore, the trend is also consistent for longer time spans, we can assume it is a general trend and, thus, that the ‘test of time’ effect is really small.”


Well, if the effect is there, it should be visible after incorporating the rest of the pop corpus. Straightforward to test; easy to reproduce.


True. I wonder if now, six years later, more older music, not just the biggest hits, has been digitized as well, so that the sample does not skew so heavily.


Sadly, this makes “pop” an inherently political thing to judge.


I found a much more interesting measure is simply to take the Billboard Top 40, wind it to an arbitrary date and see how many songs you know.

By that measure, older music was much better. Sure, the chart has flashes in the pan and bubblegum garbage in every age, but the difference in memorability between the Top 40 now and the Top 40 even 15 years ago is quite staggering.


So the BBC have been performing a variation of this by playing 30 year old "Top of the Pops" programmes on BBC4, and it's an interesting experience to see what has and hasn't stood the test of time. Most episodes have at least one spectacularly terrible "hit" on. Bonus points if the presenters intro them as the next new thing. Back then most of the performances were mimed to the track, with occasionally hilarious results.

However, at no point did someone as annoying as Ed Sheeran dominate the charts...


TOTP wasn't just mimed.

For political, financial, practical, and historic reasons, the "live" tracks were supposed to be rerecorded specially for the show. So often you'll hear a unique version on TOTP.

After about the early 70s it was damn near impossible to record and mix a single in less than a day. So what actually happened is that the TOTP producer responsible for the rerecording would be taken out for a very generous drink and/or lunch while the "session" was happening, and then they'd return to find that a master tape had magically been created in their absence, and not at all brought in by the band's manager, having already been recorded and mixed earlier.

It was a weird sleight of management that kept everyone happy.

Then the bands would go on and throw themselves around with no instruments plugged in. The vocals were often live, but everything else was literally just for show.


> However, at no point did someone as annoying as Ed Sheeran dominate the charts...

Debbie Boone? The Beatles before Dylan told Lennon that his lyrics suck?

Anyway, the issue with Ed Sheeran is that he appeals to the Tweeny Girls who are the only group still spending money on music. And, as such, the powers that be are going to flog him senseless until he stops printing money.

The issue is less Ed Sheeran being annoying than the saturation marketing reaching comical levels.


Is it possible that today’s top 40 is at a disadvantage because the songs haven’t yet been as extensively recycled through advertising, movie soundtracks, re-releases and other mechanisms by which the top 40 of 15 years ago has become more embedded in our culture and hence more memorable?


Pop music doesn’t have any music in it anymore just sound fx. I speculate it has to do with lawsuits over the very few melodies available. First case I remember was the one with George Harrison.


The top 40 means so much less today than 15 years ago. I don't know that many people that even bother with it.

Older music was just as bad/good as today.


>And the songs of yesteryear that people enjoy today (as oldies) may not be the same ones that people enjoyed when those songs first came out.

That's trivial to fix, just take the actual at-the-time top-20 of each year. Yawn.


Thanks. Url changed from https://www.smithsonianmag.com/smart-news/science-proves-pop..., which points to this.


Wow, that was really not a good article.


Smithsonian is usually better than that.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: