Engineering students to present test results at Audio Engineering Society Convention

When the 139th Audio Engineering Society Convention meets in New York City Oct. 29 through Nov. 1 and exhibits the industry’s latest technology and research, Belmont students and faculty will be on the cutting edge.

Belmont audio engineering students have previously participated in AES student recording competitions, but it is less common for an undergraduate to present a session on a paper written for a class.

Senior Nick Lobel will present test results confirming a process he and his classmates developed to archive analog tape recordings twice as fast as the current conventional method.

Analog tape was the primary means for storing sound recordings from the 1950s until the 1990s. As the tapes age, the chemical structure deteriorates, which decreases sound quality.

Archivists preserve music, speeches and other significant recordings by playing the tape recording on a tape machine and re-recording the sound onto to a computer or other digital recorder.

The current process of converting the sound on the tape to a digital recording is time consuming.

“It takes a long time because for the most part it’s done in real time,” said Lobel.

Lobel and partners Jack Mocherman, Cory Wilhite and Jeremy Deardorff tested an alternative approach to the conversion.

They played tape that would normally play at 15 inches-per-second back at 30 inches-per-second.

“We played them back at double the speed, then slowed them down with a digital process we came up with,” said Lobel.

Currently, the only system for high-speed digitization of analog tapes is equipment and software manufactured to order by Otari and costs several thousand dollars. Lobel and his group used a two-track tape machine with an audio interface and ProTools – all of which would be commonplace in any studio that has analog tape recordings.

Considering there is half a century’s worth of tape recordings that have not been preserved by digitization, the impact of a successful experiment would be monumental.

The students planned to transfer tape recordings of spoken word, pop, classic rock and classical cello. They asked to complete the transfers at a studio in Nashville but were met with opposition.

“We brought it up to a mastering engineer in town because we were originally interested in doing the transfers at his studios, and he said it will never work. ‘You should just quit while you’re ahead. That’s a horrible idea,’ basically,” said Lobel.

Although the engineer did not lend his support, he pointed out a problem the group had previously overlooked.

Tape machines do not play back recorded signals evenly across the frequency spectrum. Very high-pitched sounds are not able to be played back well on tape machines. It’s not a problem when listening to a tape at normal speed because those frequencies are beyond the range of human hearing.

The problem for Lobel and his group was that doubling the speed of the tape doubles the frequencies.

“If you play things back at twice the speed, you’re getting into that range of the roll-off area where the high frequencies are starting to get attenuated,” said Lobel. “We had to find a way to get that high frequency content back and get it back accurately.”

If the group couldn’t find a way to compensate for the lost high frequencies their double speed transfers would sound noticeably different than the original recording.

That’s where Eric Tarr came in.

Tarr, an assistant professor of audio-engineering technology, helped the group figure out what was needed to compensate for missing high frequencies to avoid affecting the tone of the recording.

The solution to the problem is a signal-processing technique the group calls spectral matching.

“The purpose of it comes down to taking the signal, processing it, changing it, so that it is as similar as possible to the original recording,” said Tarr.

Spectral matching digitally boosts the frequencies that were reduced during the double-speed transfer so the new digital recording matches the original.

The group tested the system on classmates in the senior level course with four types of recordings. Double speed transfers proved to have no discernible difference when compared to the conventional real time transfers.

“The results of the listening tests were even more validation that even with four different types signals that people couldn’t hear it,” Lobel said. “That sparked the idea that maybe we can submit it to AES.”

Lobel’s instructor Wesley Bulla, and Tarr helped Lobel submit the research to AES where it was then published, peer reviewed and selected to be presented at the convention. Both Bulla and Tarr are co-authors with Lobel.

Although Lobel is the only Belmont student presenting at the AES convention, other students and faculty will also participate in the festivities.

Bulla will present his paper titled “Detection of High-Frequency Harmonics in a Complex Tone.”

Jim Kaiser will moderate two panel discussions, one about on-set recording of ABC’s series “Nashville” and the other about Kacey Musgrave’s debut album “Same Trailer, Different Park.”

Lobel and Jesse Brock will compete in the student recording competition in the modern studio recording category, while Ross Collier will compete in the traditional studio recording category.

Story by Grant Robinson. Photo courtesy of Eric Tarr.

#AES #Audioengineering #BelmontEngineering