Although there might have been other music produced or recorded in the 1980s, we may never know of its existence due to the cacophony of all of the various keytars, drum machines, and other synthesized music playing nonstop throughout the decade. There was perhaps no more responsible synthesizer than the Yamaha DX7 either; it nearly single-handedly ushered in the synth pop era. There had been other ways of producing similar sounds before but none were as unique as this keyboard, and for ways beyond just its sound as [Kevin] describes in this write-up.
Part of the reason the DX7 was so revolutionary was that it was among the first accessible synthesizers that was fully digital, meaning could play more than one note at a time since expensive analog circuitry didn’t need to be replicated for multiple keys. But it also generated its tones by using frequency modulation of sine waves in a way that allowed many signals to be combined to form different sounds. While most popular musicians of the 80s used one of the preset sounds of the synthesizer, it could produce an incredible range of diverse sounds if the musician was willing to dig a bit into the programming of this unique instrument.
There were of course other reasons this synthesizer took off. It was incredibly robust, allowing a musician to reliably carry it from show to show without much worry, and it also stood on the shoulders of giants since musicians had been experimenting with various other types of synthesizers for the previous few decades. And perhaps it was at the right place and time for the culture as well. For a look at the goings on inside the chip that powered the device, [Ken Shirriff] did a deep dive into one a few years ago.
While FM synthesis is digital, my original DX7 sounds very analog synth. I don’t know why. It has some smoothness and suddle noise that probably makes that way.
The sounds produced by Dexed sound different, they are very sharp but on DX7 the same sounds are not that “harsh” at all.
Leaky cap in the output stage ;)
It’s the 12-bit D/A. I’ve had DX7s of the mark I and mark II varieties, and aside from some operating-system improvements, the big change was to 16-bit D/As in the latter. Strangely, a lot of people don’t like that sound as much. It is brighter and clearer and more “synthetic”.
The old DX7 had grit, and it’s the grit of “inferior” D/As.
Dexed does a couple of algorithms, though, including one that supposedly mimics the 12-bit DAC. Give that a shot?
too bad modern music doesn’t go much beyond one octave and autotune for the poorly written concept of what makes a SONG.
Are you also shaking a stick in the air while saying ‘get off mah lawn’ to those dang kids? ;)
There’s plenty of good modern music, if you know where to look. The mass-produced pop crap is designed to be done cheaply and quickly to make as much money as the labels can get their hands on, so it’s little wonder that it falls short of expectations for many (clearly, not enough, as it’s still making the labels money!).
The same can be said of any era and any art form. We always get inundated with the mostly formulaic pap of the timeframe, but in the long run only the gems are remembered.
Indeed, we in 2024 are hardly unique.
I’ve also noticed that a lot of trends you see in music often don’t apply outside of pop – for instance, in the genres I prefer to listen to, I’ve heard any number of 20+ year old songs that could easily have been done today. While no genre is completely free of trends, outside of pop, at least, it usually becomes much harder to identify a specific year range for any given piece of music unless you already know when it was released.
The 2010s-onward are not historically unique, they follow a known pattern of art-death that we sadly never learn significant lessons from. Not well enough to stop it from happening the next time.
Culture has become stuck, stultifying, suffocating, and geriatric. And exceedingly conformist, afraid of social opprobrium. In short, we are in the part of the cycle where the squares have massive influence (and the squares are not the same people from cycle to cycle. Now the squares are exactly who is most acceptable to society, which is very different from who that was in, say, the 1950s).
i’ve been making my wife go into the cd store and pick a disc at random. one of the most exhilirating aspects is listening to something new and unknown, totally convinced that this is some 20 year old brookyln hipster aping some 1970s fad in a vain effort to prove sophistication, only to discover it is in fact a 50 year old album, or the newest album from some septuagenarian. people are constantly blurring these lines and the ambiguity is fun
No, this “well actually” doesn’t deserve to exist. Art has declined in a real way and it’s kind of tiresome to reddit-deny this fact. “You can find good stuff if you know where to look” is always true, because there’s always an exception, but the RATE of that exception has become exceptionally small in past decades.
Did you keep statistics? If not, chances are this is just survival bias. Things weren’t much better 50 years ago, but we only hear and remember the good stuff. And even much of the “good” stuff of those days sounds horrible now. I have bought a juke-box, filled with records of what I suppose were popular songs in the 1970’s. More than two-thirds of that is crap to listen to now.
Compared to 80s music and technology
2024 has had less than .00000000001 good stuff to find
The instruments alone sound cold and unmusical
The popular voices sound like overdramatic let me seduce you with my whiny nasal lack of musicality
It is worse than a haunted house
And who pays to appreciate good music now that nothing of true value is created?
It’s a mess
Sorry
It just is
Just because you don’t listen to new music you enjoy, does not mean there isn’t good music coming out. The radio as we knew it is dead, and with the internet we have collective access to more new music than ever before in human history. Please don’t yell at clouds.
I’m serious. Whatever kind of music you enjoy, someone has just released an album full of it on Bandcamp or SoundCloud, and they’re bummed that no one is listening to their masterpiece. Go find it.
Confirmation bias – you don’t remember all the crap songs that came out 20+ years ago, just the greats that stood the test of time. People have been churning out low-effort junk forever.
man, this attitude and the responses here demonstrate exactly the same perceptual problem we have with people lamenting the enshittification of the internet. the world is constantly producing awesome things. the kids are alright. but if you lose track of ‘the long tail’ then before you know it you’re stuck in a pop monoculture wondering where all the cool went.
there’s always an abundance of garbage for people to point at. that has nothing to do with the question of whether or not there’s anything else. and there is always something else.
Rock and Roll is dead!
It’s the sole reason for why I ride Honda and never Yamaha. |-[
A Honda or Yamaha will never get you into Hells Angels.
Because you like dressing up in leather?
Because I’m obsessed with railroads like Sheldon Cooper and they told me I could help pull a train?
I dunno Jerry, seems like a lot of tuff talk goin’ on.
Stopped being fun when they stopped being crank-fueled young murderers. Now it’s just a club for old diabetics and sufferers of gout
I owned one of these as a kid. It was a great synth. It ended up failing somewhere in the power management module. This was prior to my introduction to electronics; I wish I could have repaired the beast.
Which reminds me that I’ve been meaning to take this for a spin on one of the Raspberry Pis I have laying around:
https://github.com/probonopd/MiniDexed
Google MicroDexed will get you FM on a microcontroller.
Had a DX9 once upon a time. Pretty cranky beast, but made some good sounds. Made a MiniDX7 and I think it holds it’s own and I can add custom sounds as I require.
FM synthesis gets a bad rep. Many folks claim it is too cryptic to learn. I took Trig and read an FM guide, and its eazy peasy. Freq ratio of the modulator/carrier is key. Integer ratios will produce repeating wave patterns. Irrational ratios produce “noise” waveforms. The noise isnt really random however since it is composed of a few sine waves. The DX7 was entirely digital, it sounds good and warm bc of its DAC.
“Warmth” is mostly the absence of higher odd harmonics, which can be a product of the analog filtering or limitations of the DAC, or how you compute the waveform in the first place.
there’s a lot of aspects of fm synthesis that are easy. the hard part is being intentional when refining a tone. if you have a tone that you like, but you want to make it (e.g.) “brighter”, it’s not always clear what attribute you have to change. some desires are easy to translate into parameters and other desires are more subtle or confusing. if you’re willing to ‘just play with it’ then you will discover a lot of awesome tones, but it’ll involve some trial-and-error.
anyways, i made an algebraic soft synth and played with the fm tones i could make with it, and that’s my impression. some things were very easy and i got exactly what i wanted, as if i had designed an instrument. and other things were just poking around in the dark and i was never satisfied, but golly i made some fun weird sounds
Years ago I became obsessed with the DX7 and the math behind it, eventually generalizing it to 3 dimensions so I could waste hours and hours generating 3D models using it (https://unsplash.com/@rick_rothenberg)
What really killed the DX7, IMO, was the outright user-hostile programming interface. It only displayed one value at a time, but the envelopes alone had four levels and four rates, each repeated for six voices: 48 parameters just in the envelope section, and you have to keep it all in your mind somehow during sound design. Now step through all of these one at a time.
It was a bit better on the DX7 Mk IIs, and even better still on the SY77/SY99 series. But it was never convenient to just experiement around with, which is what makes working on knob-per-function synths (analog or digital) so much fun.
Check out Dexed for the ultimate DX7 interface (IMO) even though it’s in software. I’ve had a long-term project in my brain to make a physical potentiometer box front-end for it, which would actually give you the synthesizer that the DX7 was meant to be, after all this time.
Or pen and paper.
BTW: Ken Shirriff
https://www.righto.com/2021/11/reverse-engineering-yamaha-dx7.html
https://www.righto.com/2021/11/reverse-engineering-yamaha-dx7_28.html
The DX7 did multiplications using log-lookup tables and addition, exponentiating at the very end. And a whole bunch of other silicon-saving design stuff.
“none were as unique as this keyboard”
So one 1 is more 1 than all the other 1’s??? If you mean “it was the best” it’s okay to say that.
It’s always amazed me the skill ceiling on this tech. Most people developing game music for FM chips ended up using simple pre-sets that sounded horrible with a few notable exceptions. Now give someone an Adlib Gold or SB16 and a tracker and they can do amazing things.
The math of music is a hell of a rabbit hole to go down.
So an octave is a doubling or halving of the frequency. So what’s a fifth? Multiplying by 1.5. So far, so good. What if you go up two fifths? Three? Do it twelve times on a piano and you land back on the same note.
But wait! 1.5^12 != 2^7. It’s close, but not equal. And it can’t be because the former is an irrational number.
The difference is called the Pythagorean Comma and it’s the flaw that makes musical perfection mathematically impossible.
“But wait! 1.5^12 != 2^7. It’s close, but not equal. And it can’t be because the former is an irrational number.”
In what universe is 1.5^12 irrational? (3/2)^12 = 531441/4096 I sincerely don’t understand how you get to that being irrational.
129.746337890625
Sorry… I was thinking of the 12th root of 2, which is the ratio you use for equal temperament, which is irrational.
Still, the Pythagorean Comma is still a thing and it does make musical perfection impossible to achieve in practice (alternate temperaments can be used but also tend to result in key specific tuning).
Ahh, thanks, I knew about 2^(1/12) but didn’t connect the dots; was worrying I might be losing it.
Fun fact: the product of two rational numbers can never be irrational but the product of two irrationals can be rational.
https://www.mathsisfun.com/calculator-precision.html
There’s also a Mac and I would guess also UNIX cli tool called bc which is quite fun. https://www.geeksforgeeks.org/bc-command-linux-examples/
The “bc” utility was a Unix tool a long time before Apple Macs joined the club. I was using it in 1988 on SunOS and Apollo systems (albeit Apollo spelled it differently).
Filed under “A” for “Aren’t I pretty mommy?” but thanks for the history lesson. Please note that I allowed for this.
“History. bc first appeared in Version 6 Unix in 1975. It was written by Lorinda Cherry of Bell Labs as a front end to dc, an arbitrary-precision calculator written by Robert Morris and Cherry. dc performed arbitrary-precision computations specified in reverse Polish notation.”
bc could probably calculate exactly how much I don’t care but again, thank you.
Filed under “A” for “Aren’t I pretty mommy?” but thanks for the history lesson. Please note that I allowed for this.
“History. bc first appeared in Version 6 Unix in 1975. It was written by Lorinda Cherry of Bell Labs as a front end to dc, an arbitrary-precision calculator written by Robert Morris and Cherry. dc performed arbitrary-precision computations specified in reverse Polish notation.”
bc could probably calculate exactly how much I don’t care but again, thank you.
Are you talking adjusted ‘piano’ tuning mode or a proper mode such as Mixolydian?
I’m talking about temperament. The comma is a thing in Pythagorean tuning, but equal temperament is close to universal nowadays, and that uses a half-step ratio of the 12th root of 2.
heh i am just as astounded by that formula but i call it 1.5^12 =~ 2^7 wow!! the error factor doesn’t move me, but the almost-equivalence is fantastic
i learned an awful lot of music theory and sampling theory and so on before i saw that so clearly expressed, and it is just astonishing to see the harmonic series, the circle of 5ths, and the 12-tone scale simply leap out of the mathematics as an inevitability. the vague ‘why??’ feeling of arbitrariness that dominates in music 101 students suddenly evaporates under the clear light of a simple piece of math
but yeah the error factor does explain why temperment is so tricky :)
All Yamaha DX series synths use FM synthesis. It was developed in the late 60’s at Stanford University, California, by John Chowning. His algorithm[ was licensed to Yamaha in 1973. The formula for calculating a sample using FM synthesis is basically A * sin(Fc *t + B * sin(Fm *t)). It requires several multiplication and sin() calculations. Given the state of the art for microprocessors in the 1970, these calculations were done by table lookup. The values for Fc, B, Fm determine the spectrum (sound) of the waveform and A determines its amplitude.
Best synth I had was a moog that was stored in a damp room where the roof was leaking and must have developed some constructive corrosion of the internals … the the sounds it produced were much more nuanced than the original.
Out of curiosity I just checked on TAS (That Auction Site) and Yamaha DX7s are listed around $400
“it nearly single-handedly ushered in the synth pop era” – nah, was already well-established by then – Moogs, Prophet 5, JP8, etc. – it certainly did make a big impact though, with a lot of artists switching to using one., many of them using it for it’s superior ability to mimic real instruments, like a piano or a saxophone.
Being “digital” does not mean playing more than one note at a time. The word you are looking for is polyphony and the DX 7 was neither the first nor least expensive synth at the time with polyphonic capabilities. Two examples that come to mind are the Juno 6 and 60, both of sold for less than the DX 7. Arguably, synth pop was firmly established prior to the release of the DX 7 ( Depeche Mode, etc.,) and the only contribution made to ’80s synth pop by the DX was to make every song after its introduction sound the same.
The 1980s was when I was a teenager so that music should be my favorite. It is not. I remember at the time hearing a song on the radio by a band I liked that had just gone full synth and lamenting “is music always going to sound like this now?” Fortunately the 1990s eventually arrived and I loved that sound: synths were abandoned and allowed to mature. But all through the 1980s I felt like an outsider to the entire culture. And Ronald Reagan drove it all home! These days I actually like synth music, as it no longer sounds like bloops and blurps.
In the late ’80s/early ’90s I worked with a guy who repaired all manner of electronic musical gear. He specialised in, amongst other things, Hammond organs. We used to split them in half for touring bands. Abig, tedious job – mainly because he had a huge roll of black wire that he insisted was used for every single connection from the bottom part to the top part, via a huge wire to wire connector. I think an Amphenol lump. Anyway, he was pretty famous in the Hammond world. Was called Bill Dunne, aka The Organ Doctor. They were happy times. The workshop was freezing, but he had a huge paraffin space heater, which he would run indoors until one of us was almost in a coma, when he’d shut it down and open the door for a bit to get some air back in. He also like to get to the workshop by 7am at the very latest. That was painful.
Bill absolutely hated DX7s. Because they never broke down. He earned very little money from them 😂