Notes & Lines discusses the
intersection of math, science, and technology with performing and visual arts.
Topics include bizarre instruments, technically-minded musicians, and cross-pollination of science and art.
Ikutaro Kakehashi, an electronic music legend who founded the Roland Corporation in the early 1970s and pioneered digital music standards in the ‘80s, died last week at the age of 87.
Kakehashi started his musical career in Japan repairing electronic organs while running his own electrical appliance shop. In 1958 he decided to devote his career to designing an ideal electronic musical instrument and founded Ace Tone in 1960. Ace Tone primarily manufactured electronic organs and guitar amplifiers, but Kakehashi’s passion was developing an electronic drum. After a few prototypes, he patented an “Automatic Rhythm Performance Device” in 1967. Kakehashi used a diode matrix circuit involving a number of inverted circuits connected to a counting circuit. The circuits’ synthesized output became the desired rhythmic sounds. He commercialized the design as the FR-1 Rhythm Ace, essentially the first drum machine, featuring 16 preset rhythms and four manual percussion sounds.
Kakehashi left Ace Tone to start the Roland Corporation in 1972. He continued to improve his drum machines and introduced several new models before releasing the iconic TR-808 in 1980. By this point drum machines were becoming popular in New Wave and electronic music, and several rival manufacturers were already making digitally sampled models. The 8-bit Linn LM-1, which debuted that same year, was the first device to use digital sampling of acoustic drums and also one of the first programmable drum machines, but it retailed at a whopping $4,995. (Here’s a video demo of the LM-1, showing its superior sounds.) Despite the fact that microprocessors were becoming more common, the TR-808 used analog subtractive synthesis common to synthesizers of the previous two decades to reduce manufacturing and retail costs.
The TR-808 had an odd design history. Kakehashi hired rogue American musician and engineer Don Lewis, who was known to extensively modify electronic instruments for his own use, to design Roland’s drum machines. In the spirit of Lewis's unconventional approach, Kakehashi deliberately ordered faulty transistors to give the drum machine a characteristic sound. The unique cymbal tone was supposedly discovered when a Roland engineer spilled tea on a breadboard prototype, and the design team had to work for months to recreate the pssh sound.
Despite retailing at less than half the price of digital drum machines, the TR-808’s unrealistic sounds made it a commercial flop and production ceased in 1983. It picked up a cult following, however, and is heard on many hit songs starting with Marvin Gaye’s “Sexual Healing” in 1982. Most hip-hop groups of the ‘80s used the TR-808 to create their beats, and it’s still heard regularly in mainstream pop and hip-hop. Rapper Kanye West's 2008 album 808s & Heartbreak is an ode to the device, and West used it on every track.
During the TR-808’s run, Kakehashi began promoting the idea of digital music standardization. Talks with other electronic instrument manufacturers led to the introduction of the Musical Instrument Digital Interface (MIDI) standard in 1983. MIDI allowed interoperability between any type of digital instrument, provided a standard voice library for digital keyboards, and allowed a single controller to play several digital instruments at once. Over 30 years later, MIDI is still the dominant technical standard for digital music.
Kakehashi retired from Roland in 2013, and the company still manufactures keyboards, synthesizers, stage pianos, recording equipment, and digital music software. While his name may never be a household one, his influence and legacy are felt and heard throughout the digital music world to this day.
About a year and a half ago I started working as a professional organist on the weekends. In the course of that work I’ve met people who work a day job and play on the side, and many people who don’t play any instrument at all but are fascinated by pipe organs. I’ve found that a disproportionately large number of these people are professional engineers.
Pipe organs sort of scream “engineering” more so than any other musical instrument. They’re fantastically complex, often with thousands of moving parts, but also extremely simple in that they’re basically just a mammoth set of scaled flutes controlled by valves operated by a piano keyboard. A typical instrument, with its own unique tone and personality, is individually designed and built to diverse yet precise specs. Compared to a piano they’re musically clunky and incapable of on-the-fly volume changes, but a large organ can produce tones ranging from a whisper to a deafening roar.
It’s probably no surprise, then, that many engineers attempt to build their own organs. One of the first home-built organs to surface online was Matthias Wandel’s. Wandel had next to no musical knowhow but decided to attempt his own organ in 1992 while still in college. Using the resources available in his father’s extensive woodworking shop, he managed to build a pretty impressive working organ. Wandel fashioned his pipes from wood, which is a common organ pipe material anyway, and first provided wind using a vacuum cleaner motor and hose, later upgrading to a small motor and precision blower. Wandel shared his plans on his website, showing that aside from the blower motor his instrument is completely mechanical, much like all pipe organs were until about 1880.
In 2002 a relatively non-musical mechanical engineer named Raphi Giangiulio found Wandel’s plans and was inspired to go a bit further. He quit his job in 2003 and ended up taking almost five years to build an actual room-sized organ with two separate windchests, complete with five sets of pipes and a foot pedal keyboard, completely out of wood. Giangiulio’s website has an extensive set of plans, audio and video samples, and the mathematical formulas he used to design his pipes and tunings. Like Wandel, Giangiulio benefitted from his father’s large woodworking shop. Perusing the site makes it pretty obvious that this was a “quit your job and get it done” kind of project. Giangiulio has become an amateur organbuilder in his own right and has finished three small organs, two of which he's sold, since he completed the big one.
As a former longtime brass player I’m probably more inclined to construct a much-simpler hosaphone. Easier to build, but not even close to the sound of a homebrew organ.
John Philip Sousa—composer of the well-known “Stars and Stripes Forever” march and namesake of the sousaphone—famously hated recorded sound. He almost always refused to conduct his band when recording equipment was present, and in 1906 he said this during a congressional hearing: “These talking machines are going to ruin the artistic development of music in this country. When I was a boy...in front of every house in the summer evenings, you would find young people together singing the songs of the day or old songs. Today you hear these infernal machines going night and day. We will not have a vocal cord left...”
Today, as Sousa predicted, recorded sound has rendered most forms of live music-making irrelevant, or at least not strictly necessary. And while digital music is worlds away from the wax cylinders he griped about, its sound is still very different from live performances.
An open-source article published this month in the Journal of the Audio Engineering Society attempted to examine the relationship between audio compression and emotions. A group of four researchers from Hong Kong University of Science and Technology played a group of subjects audio samples of eight different orchestral instruments using three varying degrees of audio compression. Subjects voted samples into ten emotional categories, from happy and heroic to scary and angry. The study found that reducing a sample from 112 Kbps to 32 Kbps resulted in a reduction in “positive” emotions and an increase in negative ones. It also found that the trumpet samples were affected most, while the horn was least affected by far, probably because of the difference in overtones between the two instruments.
Several audiophile blogs and websites have already picked up on the study, claiming that it corroborates their belief that vinyl’s still the ideal listening format. The research is problematic in a few ways, however. The group postulated that 32 Kbps audio compression adds a background “growl” that caused respondents to negatively reclassify their samples. As demonstrated in the video below, the difference between 112 Kbps and 32 Kbps bit rates is monumental, and I’d call the “growl” they're referring to straight-up distortion. The bit rates chosen by the team are surprisingly low in 2016 terms, and even the 10-year-old inexpensive digital recorder I use at home records at 320 Kbps. Also, studies examining emotional responses are often highly subjective to each participant. To be fair, the researchers consistently describe the study as an introductory one that could lead to further, more substantial investigation.
Lossy audio compression—the technology that makes it possible for hundreds or thousands of songs to fit onto a matchbook-sized MP3 player—has changed the sound of recorded music dramatically. Tech-savvy listeners who are more concerned with sound quality can use lossless codecs to convert their music, but these tracks are likely to be up to five times larger than ones that used lossy compression. I find it astounding that, only 110 years after Sousa’s comments, one can access millions of hours of free, high-quality music on YouTube at any time of day in any part of the world (with Wi-Fi). But I’d be interested to see more studies on the relationship between compressed music and user emotion; it would likely prove that there’s still no convenient substitute for hearing it live.
On October 31 construction work on the Hamburg Elbphilarmonie concert hall officially ended. The magnificent glass façade sits atop an abandoned warehouse on the river Elbe and is now the tallest inhabited building in Hamburg. The 2,039-seat hall will officially open in early January.
Despite its grandiose appearance, the Elbphi (as it’s popularly called) was something of a budgetary disaster. It took over nine years to build, six more than planned. While construction was initially estimated to cost €77 million in 2007, the final cost of €789 million was 10 times the original figure. The project was subject to some blistering media criticism and public scorn as a result of these setbacks.
Anyone well-versed in modern music hall construction would expect nothing less from such a project. Building a concert hall is characteristically expensive and time-consuming, often beginning with reasonable estimates that quickly spiral out of control. For example, the famed Sydney Opera House broke ground in 1959 and could have been finished as early as 1964 at a cost of $7 million Australian dollars. Due to a variety of setbacks, including weather, site drainage and miscommunication, the hall was opened in 1973 after construction costs ran to $102 million AUD, nearly a billion AUD in 2016. In Los Angeles, the Frank Gehry-designed Walt Disney Hall took 12 years to build because of the need for a $110 million underground parking garage and stalled funding in the mid-1990s.
Why are concert halls so difficult and expensive? Many critics point to the demands levied by famous conductors and classical music impresarios; indeed, all three halls discussed so far are iconic, breathtaking spaces worthy of hosting great music (and huge egos). Others point to the often faulty economics of selling classical music: a hall must have enough seats to justify its cost, and perhaps be architecturally grandiose enough to attract people to fill them.
Taking a hall’s acoustics into account creates an interesting interplay between architecture and seating capacity. Most of the concert halls judged to sound the best—including Vienna’s Musikverein, Amsterdam’s Concertgebouw, and the Birmingham (UK) Symphony Hall—incorporate a “shoebox” design in which their auditoriums are simple rectangles with the performing group stationed at one end. In a shoebox hall, the first sound perceived by a listener has reflected off either wall, so the sound heard in each ear is subtly different. Because the sound from each wall takes slightly longer to reach the far ear, the signal attenuates to bend around the head and gives the listener the impression of being enveloped with sound. One perceived drawback of shoeboxes is that listeners in the very back, farthest from the performers, experience a less satisfying aural and visual performance.
The alternate design pioneered in Berlin in 1963 is the “vineyard” hall in which the orchestra is effectively enveloped by their audience. This design is sort of the musical version of a theater in the round, with terraced groups of listeners spread throughout the space, including behind the orchestra. Vineyard halls have a few cool advantages, including the ability to see the conductor’s face from your seat behind the orchestra, but acoustics are not among them. To begin with, many instruments (including trombones, trumpets, clarinets and oboes) project their sound straight out from a bell or pipe. Those seated behind the ensemble would perceive much less of these instruments but much more of the horns, whose bells face backward. Most vineyard halls have angled walls, so they don’t experience the same enveloping reflections of shoebox types. To make matters worse, vineyard halls were originally conceived to increase the amount of seats (and tickets sold). Bodies and clothing act as natural dampers, so the larger the audience, the quieter and duller the sound.
A glance at London’s famous halls shows that designing an economically successful hall that sounds “good” has always been a difficult proposition. The famed Royal Albert Hall, which hosts the popular BBC Proms, is typically judged an acoustical failure due to its enormous, 5,000-seat size. (Granted, it was built in 1871, before acoustic design was a true science.) London’s Royal Festival Hall was built within a reasonable time and on-budget (18 months and for £2 million) using state-of-the-art acoustical figures in 1951, but the designers failed to take audience absorption into account, so it’s now judged as one of the driest—and worst-sounding—halls in Europe.
Sir Simon Rattle will become conductor of the London Symphony Orchestra next year, and it’s no surprise that he’s already lobbying for a brand new £100-200 million concert hall to replace the vineyard-style Barbican Centre. If he gets his wish it’s likely that the new building will be an architectural behemoth, hopefully with just enough seats to satisfy the orchestra and the audience.
Image credits: a as archictecture | Santa Fe University of Art and Design
Music is a no-brainer when it comes to AI research. It has a finite set of rules, a relatively limited scale, and pretty strict limits as to what “sounds good,” at least based on a researcher’s subjectivity. The concept of musical AI is also at least 50 years old: well-known futurist Ray Kurzweil appeared with a music-writing computer of his own invention on I’ve Got a Secret in 1965. However outlandish a music-writing computer might’ve been in the 1960s, Ray unfortunately stumped only one panelist.
This summer has seen a bumper crop of headlines about musical AI. The most notable example is probably IBM’s Watson, the Jeopardy-winning supercomputer. IBM is leveraging Watson to create Watson Beat, a new app designed to boost a musician’s creativity. A user feeds the app a musical snippet and a list of desired instrument sounds, and Watson more or less remixes and alters the sample, choosing its own tempo and doing its own orchestration. Richard Daskas, a composer working on the Beat project, says the app could be helpful for a DJ or composer experiencing “musician’s block” in that it “generate[s] ideas to create something new and different.” An IBM researcher working on the project says Watson Beat should be available commercially by the end of the year.
If there’s a developing tech-related area, how can we not expect Google to be in the ring? A few months ago the tech giant released a 90-second melody created by its Magenta program. Google is using Magenta, a project they first announced in May, to apply its machine learning systems to creating art and music. Similar to Watson Beat, a user feeds Magenta a series of notes, and the program expands them into a longer, more complex sample. Magenta relies on a trained neural network, an AI technique inspired by the brain, to remix its inputs. Google’s efforts at neural networks have already tackled visual artistic development: its DeepDream algorithm was the basis for a visual gallery show early in 2016.
Recent research from Baidu, the Chinese tech giant, takes a different tack and combines AI, visual art, and music. The company’s experimental machine learning algorithms analyze patterns in visual artworks and map them to musical patterns, creating a kind of “soundtrack.” (Check out this video, in which the AI Composer tackles Van Gogh’s Starry Night and other images.) Baidu says the program first attempts to identify known objects such as animals or human faces within the image, and analyzes colors for perceived moods (red=passion, yellow=warmth, etc.). AI Composer contains a large musical library categorized by “feel,” and it draws upon these musical samples to piece together an original composition in the mood of the image.
A large grain of salt is necessary when evaluating AI developments, at least in my opinion. It’s exciting to see artificial neural networks doing their thing, but even considering the subjective nature of art and music, it’s hard to see how Watson, Magenta, or AI Composer have produced anything “good” or worth listening to. Granted, they’re all in the early stages, so who knows? Maybe we’ll see a day when composers come up with basic ideas and let computers do the rest. I for one hope that day’s far off over the horizon.