In the digital world, today’s hottest innovation can be obsolete tomorrow. But some inventions endure and are built upon and improved for so many years they’re hard to live without. Two of those inventions, created at Stanford, are FM synthesis and fluorescence-activated cell sorting.
Can you hear me now?
John Chowning, professor emeritus of fine arts and music, came up with the idea for the FM synthesizer in 1967.
“I kind of stumbled on it,” he recalled.
At the time, Chowning was experimenting with localization — how to make computerized sounds seem to come from different places in a room. To create the illusion of a sound coming from somewhere other than directly from a speaker, he had to understand how we perceive sounds at a distance, which depends on the structure of the sound and how it reverberates off walls. And with the relatively low processing power of computers at the time, it was no easy task.
“I began searching for this sort of complexity,” Chowning said, by examining a sound’s pitch. Increasing vibrato — the oscillation between two pitches, like the wobbly sound of a violin note — can change the actual note you hear.
But Chowning’s computer was no violin. In the 1960s, PCs weren’t capable of more than simple “beeps.”
“Imagine how the sound of that is simply a tone sound,” he said. “So I simply modulate that sound with another wave that is sinusoidal [similar to a sine wave] in shape.”
By combining multiple tones with different frequencies, Chowning could create unique sounds.
“So that, basically, is frequency modulation, and I was experimenting with the computer to generate signals that had some internal liveliness that helped me localize” the sounds, Chowning said.
Chowning noticed that as he increased the vibrato’s rate and depth, he no longer heard just a change in instantaneous pitch, but a complex timbre, much like the sound of a musical instrument.
“I knew enough about the world at that time to realize that was unexpected — to get such complex timbres by other means demanded lots and lots of oscillators or waveforms and filters — lots of gear,” Chowning said.
He followed with some “semi-rigorous experiments.” Not only did it become a useful experimental tool — “it generated lively signals, lively sounds,” Chowning said. “But then I realized a few years later that it was probably useful to the electronic music instrument world.”
Chowning signed the patent for FM synthesis over to the University, and Stanford’s Office of Technology Licensing (OTL) marketed the idea to a number of organ companies that already made basic electronic instruments. After being turned down several times, the OTL “finally found one who understood what I was doing,” Chowning said.
And in 1983, the 61-note Yamaha DX-7 synthesizer was born.
“One of the great advantages of FM over competing technologies . . . was that it demanded relatively little processing — it was a simple formula that demanded little memory,” Chowning said.
Less memory meant smaller hardware. The DX-7 was no bigger than its descendents that are sold today. And regardless of whether or not one is a musician, FM synthesis matters to everyone who owns a cell phone.
“Probably its biggest application ever was ringtones, believe it or not,” Chowning said.
Getting to the FACS
Genetics professor Len Herzenberg “used to complain that he couldn’t look down a microscope; he didn’t like counting things in the microscope, it was frustrating to him,” said his wife and lifelong lab partner, Lee Herzenberg.
His solution, which built on older, less efficient machines, was the fluorescence-activated cell sorter (FACS), invented in the 1960s. It started out as an instrument used to measure atomic fallout by judging the size of atmospheric particles collected after bomb tests. Herzenberg changed it into a machine now present in almost every modern hospital and microbiology lab.
“As most good inventions are like, he didn’t have to invent the wheel to make an automobile,” Lee Herzenberg said. “But he took a wheel that was known to work on a cart, and [he] improved it to work in an automobile.”
Building on the Los Alamos machine and other technologies, Herzenberg created the FACS.
Rather than picking through Petri dishes for the products of chemical reactions, “we were trying to get subsets, subpopulations of these cells and understand them,” Lee said.
After all, “you can’t study it if you can’t isolate it and recombine it with things,” she added.
Fluorescence is one way to identify and separate the cells into different batches. In the original FACS, a black light hides particular wavelengths of visible light so that only a handful of wavelengths chosen by the researcher shine through. The light excites a particular molecule, marking the outside of a cell and making it glow another color.
More advanced versions of the machine, used today, shine lasers.
“What FACS does is it has cells run in single-file in a stream and it shines laser light,” — which is high energy, but emits a narrow set of wavelengths — “and if they have a fluorescent molecule that can absorb where the laser is lighting it, they absorb and emit then at whatever their characteristic color is,” she said.
After they’re hit with the light, a detector, which is often covered by a filter to allow in only the desired wavelengths, checks each cell’s color. Then, they are separated based on how they glow. And all that happens over the space of about three inches, from where the cell passes the detector to where it comes out of the nozzle.
State of the art sorters can handle many different types of cells at once.
“These days, [there are] up to seven lasers and 20 filtered detectors, and then we have electronics that orchestrate everything,” Lee said. “We may look at a million cells in a minute or so.”
The most popular proteins to “tag” cells with are antibodies.
“The reason we use antibodies to tag things is they have microscopic eyes. They can detect shapes of things sort of like a blind man recognizing somebody’s face by detecting the configuration or the shape of it,” she said.
When the antibody recognizes a shape that it’s looking for, “it’s like a lock in a key, it finds a shape that matches it, they stick together very tightly,” she added.
Antibodies made by the body, combined with human cells, allow FACS to recognize, among other things, HIV infection and B-cell leukemia, because different combinations of antibodies attach to cells in different conditions — and all of them look different to the machine’s eye.
“Technology can in turn grow the field by making it possible to see these things and therefore opening up new questions to be answered,” Lee said. “So it’s kind of a spiral, if you like.”