In-flight content experts Altitude Media recently honored Seer Systems by telephone interview.

“Another in a two-decade line of ‘firsts,’ this is to my knowledge the first coverage of a synthesizer company by in-flight radio and magazines. It brings our story to the attention of over 10 million people.” founder Stanley Jungleib said.

The segment runs on Delta and USAir for all of September and October, 2011.

Listen: Altitude Media Interview

“Patent Bust” of Seer Systems’ Electronic Music Patent Sunk By USPTO Action

San Francisco, CA – October 29, 2010 – The U.S. Patent and Trademark Office has rejected the EFF’s “Patent Busting” re-examination attack on Seer Systems’ patent #5,886,274. According to the EFF’s website, Seer and its ’274 patent were “Wanted by EFF Marshalls” for “Crimes Against the Public Domain; Willful Ignorance of Prior Art; Eggregious Display of Obviousness.” The USPTO disagreed, ruling that the core original claims of the ’274 patent are valid, as are six newly-added claims. From a technical standpoint the entire affair cost Seer only a few minor amendments of wording that rendered three of the original claims in the ‘274 patent no longer necessary.

Founder and Chairman Stanley Jungleib, responds: “Seer’s ‘274 Patent inventing scalable digital audio distribution has now withstood all reasonable tests and challenges. Seer Systems has already settled litigation with Microsoft, Yamaha, and Beatnik. Thanks to EFF this Patent has now received enough attention that no one in the MIDI or audio domain can claim ignorance of its validity and implications. Seer Systems looks forward to licensing responsible customers under fair terms. And, as necessary, we will continue defending our intellectual property consistent with our substantial record.”

Seer’s outside litigation counsel, Alex Weyand of the Weyand Law Firm, San Francisco, commented: “Why the EFF did not contact our client first, and instead chose to learn the hard way, remains a mystery.”

In recent years, the software synthesizer market has come into full swing and digital audio is something that consumers have come to expect in everything from personal computers to mobile phones and cars. In many cases, the manufacturers of these devices have been licensing technology from Seer Systems, whose intellectual property is protected by US Patent #5,886,274. (For details, see What Exactly Is the 274 Patent?)

By the time Seer was awarded the 274 patent, it had developed software synthesizer products for Intel and Creative Labs for the consumer market, and a suite of music production and playback tools for the professional musical instrument market. (As detailed in “The Best Audio Helper App You Never Heard in Your Life.”)

Seer’s retail products have not been upgraded in nearly 10 years and are still only available for Windows 95 and Windows 98. So, why did Seer shift its attention from producing synthesizer products to protecting their intellectual property? The answer is that a combination of powerful technical and business forces caused Seer to drastically shift strategy were it to survive.

First, supporting Microsoft Windows NT and 2000 was impractical. Seer’s line of professional software synthesizers used 80-bit processing and relied on access to the CPU interrupt, to which they had access until Microsoft released Windows NT. In other words, an upgrade of Reality would require more than an upgrade, it would require a rewrite with no guaranty that it would provide the same sound quality as the 80-bit resolution their customers had enjoyed. Seer considered releasing Reality as an open source product,  but legal issues have complicated that path as well. (See “Seer Considering Releasing Reality Code as Open Source.”)

By 1997, Seer’s finances were in trouble. Creative was not actively selling Seer’s software synthesizer upgrade for the Sound Blaster, so royalty payments were nowhere near their projections. Seer founder, Stanley Jungleib recalls, “I was looking for markets. I felt my job was to build something that was investable, so I didn’t work closely on Reality as a product. My product was the company.”

Jungleib had written extensively on music production systems, and even speculated about systems that could leverage General MIDI to adapt to a composition, but until Jungleib heard Seer’s new Reality engine working, even he had not envisioned the comprehensive potential of a software-based music distribution system. “I wrote it down in October, 1995: A Painful Plan for Painless MIDI. I knew it was the Grail but still too early; at the time I didn’t make a big deal about it within Engineering, because they were having enough issues dealing with Creative Labs and Windows as it was.”

Jungleib recalls how discussions with Opcode’s founders, Dave Oppenheim and Chris Halaby created the synergy to most efficiently realize the system. “They had a great sequencer with a great audio handling system and we had the best software host-based synthesizer and Windows realtime engine. We were going to use their suite of Galaxy Editors as different ‘skins’ over the Reality engine. Melding our technologies could make a tool that would solve all these distribution problems for the professional musician.

“So I wrote this 120-page specification on how that would work. How our synthesizer would interface to their sequencer to deliver a totally predictable experience for the user. I had commands such as ‘preload by bar,’ ‘preload by sound number,’ ‘unload by bar number,’ ‘unload this range of bars’…. Breaking it up so the musician would have total control over the music, over what was allocated, when, and how it was delivered. You could draw resources on-the-fly from anywhere on the net, there was a bandwidth simulator … And very importantly as a response to what was going on with mp3 theft, provision was made for the musician to protect their creation. (What you might now call DRM.)”

Unfortunately, in 1998, Opcode was purchased by Gibson, and the joint venture stopped altogether. Sales of Seer products were still suffering, and between the software piracy, and continuing issues with Creative, Seer’s ability to provide livelihoods for its staff of 25 was in jeopardy.

“Only after the company was in really dire financial trouble did I decide to ask my lawyers ‘Can I patent this?’ That was in 1997. The patent was finally awarded in 1999 after several rounds of rejections.”

Ten years on, Jungleib continues to protect Seer’s intellectual property, though some companies underestimate the extent of Seer’s investment in the technology and the validity of the patent.

“There’s a difference between imagining something (like a time machine), and actually committing resources to specifying and building it,” says Jungleib. “At the time, everyone else was committed to their own little hardware platforms. Many still are. But, with Intel as my hardware department, and a synthesizer and effects engine that could redefine itself every 10 milliseconds, I was liberated from all those arbitrary limitations.”

Seer Systems is recognized as the creator of the first professional software synthesizer, Reality.  In fact, Reality was not the first software synthesizer that Seer produced. By 1995, Seer had already produced software synthesizers for Intel and Creative Labs, though these programs were not the full-featured synthesizers that today’s electronic musicians take for granted. Stanley Jungleib established Seer Systems to develop the first software synthesizer for Intel, and his experience in leading their development helped move Seer Systems towards designing Reality and the SeerMusic system.

So what motivated Jungleib to take the leap from the world of hardware synthesizers and MIDI into the world of software synthesizers? According to Jungleib, two people played pivotal roles in that leap: Glenn Spencer and Avram Miller.

Glenn Spencer was a scientist, who in the late 60’s abandoned his career to become a piano teacher in Stanford, CA. Over the years he became a center of the mid-peninsula music community, organizing concerts and establishing the Music Special Interest Group (MuSIG), and the Stanford MIDI Users Group.

Through the MIDI User’s Group, Spencer met Stanley Jungleib, who was teaching MIDI and electronic music at Cogswell College. The two became friends and remained in contact until Spencer’s passing in 1998. Jungleib recalls, “Particularly at Glenn’s service where many spoke, I learned how enthusiastic a believer he was in every one of his students and friends.“

“Aware of his work leading the Stanford spinoff Music Special Interest Group (MUSIG), I looked up Glenn when moving from Los Altos to Palo Alto. MIDI and the new multimedia explorations I was reporting interested him so much Glenn quickly became the charter President of the local Stanley Jungleib Fan Club. He volunteered to become my secretary — which I could not accept.”

Jungleib was overwhelmed to be on the receiving end of Spencer’s support for musicians and composers. “Of Glenn’s many musical qualifications were that he spent years discussing music and aesthetics with Stanford roommate, twelve-tone composer Roger Sessions. So, when Glenn tells one that Earth Sighs is ‘the most beautiful piece he had ever heard’ and goes ‘way beyond Mahler’ it resonates nicely—and actually, they did say they found my CD in his player.”

Glenn Spencer and Stanley Jungleib at Seer Systems, 1997

Spencer’s support for Jungleib’s work in electronic music went even farther in 1991, however, when he suggested to one of his senior jazz piano students that he contact Jungleib for help with a project that had crossed his desk. The student’s name was Avram Miller. And he was vice-president of development at Intel.

Miller intended “To make the PC the music platform of the 90s.” Intel’s interest in sound and music emerged first by supplementing the 486 with a DSP board called the Mikado. The Mikado was intended to enable fully Multimedia PCs, which in 1991 corresponded to x386-based PCs with built-in audio features and FAX support.

The Mikado employed an industry-standard DSP chip, with its own OS that was completely foreign to the pro audio community. When Jungleib could not get other music software companies to bid on the project, he assembled his own group of developers and got to work. By mid 1992, however, it was determined that the Mikado board would not have sufficient horsepower to support faxing, to say nothing of the desired audio functions. Enter Ralph Smith.

Smith held Intel badge #14 and was Jungleib’s main engineering contact at Intel Architecture Development Labs in Oregon. When the Mikado’s shortcomings became evident, Smith asked if Jungleib could get the code to run on a host processor instead of the DSP board. What incorporated as Seer Systems in December of 1992 got to work, and in March, 1993, delivered Satie, the first x86 host-resident real-time software synthesizer.

Jungleib credits Glenn Spencer and Avram Miller for being the two people without whose efforts he would never have established Seer Systems. He maintains photos and press clippings in the Seer Systems archives to recognize their contributions to their friends and communities. “Our relationship with Intel was long and complex, but everyone I worked with there modeled the highest standards of professionalism. Their cultured demanded being open to reason,” he concludes.

On March 23, 1999, the US Patent Office issued Patent #5,886,274 to Seer System. The title of the patent is “System and Method for Generating, Distributing, Storing, and Performing Musical Work Files.” (Check out this scan of the patent from the Seer Systems archives.)

Due to the EFF’s “Patent Busting” campaign, there seems to be considerable confusion around what exactly the “274 patent” describes. The EFF has gone so far as to call it “bogus” and “illegitimate,” but when pressed for their justification, the EFF respond that “free speech” is sufficient justification. They also point to Seer Systems’ founder, Stanley Jungleib’s book General MIDI as “prior art” that makes the patent invalid by describing the same music publishing system (in the section titled “GM2000″).

Does General MIDI describe the system described in the 274 patent? Does it even describe the SeerMusic system that Seer Systems released three years after General MIDI’s first publication?

Not at all.

In 1994 when Jungleib completed the first draft, and even in 1995 when the book was finally published, the devices with which musicians worked were still hardware devices – mostly keyboards and drum machines. Computers were used for sequencing and for editing synth programs, and Internet access was becoming increasingly commonplace, but software synths were simply not the sort of thing you could buy at a retail music store.

Jungleib wrote GM2000 to help synth manufacturers understand the difficulties facing composers and to encourage manufacturers to support General MIDI. GM2000 proposed a future version of GM in which a playback device could help hardware devices determine the most appropriate sounds for performing a composer’s work. If all else fails, the synths would have GM-mapped sounds to fall back on. In other words, the system would make the “worst case” a bit less bad.

Jungleib did not describe an end-to-end software solution that ensures that a composer’s work will sound identical on another system. Jungleib envisioned that system in October of 1995, months after the release of General MIDI. He then reassembled the Seer Systems design team and began work on Reality, the first step towards the SeerMusic system, in which he implemented the system that would eventually be protected by the “274 patent.”

If you’ve ever considered publishing a MIDI file of your own, but gave up on the idea because of all of the variables between you and your potential listeners, the difference between GM2000 and the 274 patent are obvious.

BTW – If you don’t have a copy of General MIDI, you can read it for free online and see for yourself.

When Stanley Jungleib wrote General MIDI in the mid-1990′s the electronic music landscape was filled by hardware synthesizers and software sequencers. A decade later, real-time software synths -which Seer Systems had invented in 1993- seem to dominate recording studios, but the MIDI foundation remains intact. USB and Firewire cables have replaced a lot of MIDI cables, but computers and keyboard controllers continue to speak with one another over the MIDI protocol.

You can now read General MIDI in online form for free. Read it here.

Published in 1995, General MIDI is still recognized as an essential reference for anyone interested in the evolution of the MIDI industry. Check out the reviews from Computer Music Journal and Audio Engineering Society in the Seer Systems archives.

In 1999, Seer Systems released ReMixer, a free helper app that rendered musical performances published in the SeerMusic format. Composers could create SeerMusic files with Reality, importing and creating musical elements from multiple sources. The SeerMusic files then included all of the information needed to let ReMixer recreate the performance exactly as the composer intended – track levels, synthesizer patches and parameter changes, and even effects processing were performed in real-time by ReMixer – a capability still well beyond the capabilities of any free helper app available on the Web.

Here’s a demo video from 1999 that demonstrates ReMixer stand-alone application user interface. It shows how you can change the mix: level changes, mutes, solo, reverb on each track and even a global tempo and transpose setting for the entire song. Experienced electronic musicians will note that most of the tracks are being played by the software synthesizer built into ReMixer, including the lead guitar.

The Moog and Chimes demos show how Web application developers could embed synth controls in their user interfaces. Check out the way you can adjust the filter settings. This is not a simulation – these controls actually change the filter cutoff, resonance, etc. 10 years later, the SeerMusic platform is still the only one that enables this type of user experience.