PNW Chair Dan Mortensen started the meeting with general business and had all attendees briefly introduced themselves. Facility host Glenn Lorbecki spoke about the Recording Academy (the Grammy award folks) and its Producers and Engineers Wing. Glenn is currently serving as co-chair (with Bob Ludwig) of the P & E Wing. PNW Committeeman Bob Gudgel introduced Dan Lavry.

Lavry's work goes back to the early 70's designing circuits for telephone traffic measurement gear. Some of his work included high speed IC test systems for IBM and Toshiba, a time interval generator for Hughes, an AD converter (8 bits at 100MHz) for EH Research, a programmable filter and isolated AD for Nicollet Biomedical, and an array of AD and DA converters for instrumentation, medical and telecom industries (HP, ITT, ATT, GE, Siemens and more).

He acquired an appreciation for timing accuracy, including jitter issues when designing some specialized tail pulse generators for physics research labs. He then spent a few years designing super accurate time interval meters for the aviation industry. That project was all about fine timing, clocks and low jitter issues.

With his background in conversion, and a love for music, he got into audio in the early days of digital audio. Dan's first design was an early 16 bit stereo AD and 18 bit DA. Early products that were designed for Analog Solutions were sold to companies such as Ampex, Otari, New England Digital and more. Ultra Analog, who manufactured Dan's DAC module under the name DAC D20400, sold modules to OEMs Wadia, Pacific Microsonics, Mark Levinson and others.

Dan partnered with Apogee Electronics and designed the electronics for the first Apogee AD-500 and DA1000 and contributed nearly 2 years into the dither project with his mathematician friend, Jerry Goodwin.

For the last 14 years Dan has been designing conversion equipment for his own company, Lavry Engineering, formerly known as dB Technologies. The company name was changed in 2002.

Dan began the meeting by discussing how people might claim to hear differences in S/PDIF (Sony/Philips digital interface) cables. He used his laptop to project graphics using MathCAD. He showed how factors such as long cable lengths, the AC coupling, cable reflections, and the preamble in the data stream could distort the signal and contribute to high jitter or clock accuracy error. His recommendations boiled down to: use the shortest cable possible and use word clock on digital devices if possible, rather than the signal clock.

A short break was held for refreshments, and door prizes were awarded.

• Lanyards: Bob Cavanaugh, Graham Jordan
• Tote bag: Diann Skye
• Polo shirt: Rick Rodrigues
• Hat: Jon Ulrigg
• Coffee mug: Jared Boone
• CDs: George Thomas, Bob Smith, Mark Edman, Ryan Wells, David Christensen, Michael Major, Lindsay Smith
• Digital multimeters (donated by Rick Rodrigues and Fluke Corp.): Marc Lavry, Elise Kates

He also described synchronous (integer multiple, such as 44.1kHz to 88.2kHz) & asynchronous (any other multiple, such as 96kHz to 44.1kHz) SRC. While synchronous is relatively easy to do in hardware, asynchronous conversion must streamline the calculations and do limited coefficients to work in real time. A good software SRC can be a good alternative, but you can't monitor in real-time.

Dan continued with some comments about "inferior digital sound" and reviewing fundamental pulse code modulation (PCM) theory. He feels a perfectly implemented digital system should not have any "digital sound." Since nothing is perfect, we are hearing only the difference in how close to perfect implementation we have achieved.

He discussed distortion, and fallacies about harmonic distortion and n-ordered functions. He said digital is good for linear processing, but not so good for non-linear operations. He cited software plug-ins for effects such as tube emulation and hard limiting that he feels are not good. There is a possibility of the signal being aliased because of the harmonics being generated in the software unless the programmer is very careful to have this not happen.

Finally, some questions led to discussions of adequate bit depth for processing, and the reality of 20 vs 24 bits.

Special thanks to Glenn Lorbecki of Glenn Sound studios for use of the studio for our meeting.