When I was a budding sound engineer with the U.S. Air Force Band during the 1990s, one of my mentors was a consultant in the Washington, D.C. area named George Weber.
George was a somewhat eccentric guy and definitely a dyed-in-the-wool audiophile. He had done some modifications to the band’s K&H studio loudspeakers, and those things did sound amazingly good.
George once invited me to his home and played some record albums on his super-high-end stereo system. The turntable alone was amazing – it had an air bearing for the spindle and a linear-tracking arm also with an air bearing.
To keep the air noise to a minimum, the air pump was in a separate room. I don’t believe I’ve ever heard LPs sound that amazing. And if I remember correctly, he had a personal pair of K&H loudspeakers in ceramic housings.
For the listening session, George suggested I bring some of my favorite (and familiar) albums, which leads me to my point: the material we choose for critical listening and sound system tuning should satisfy several criteria.
First, we should very familiar with it. Second, it should be well recorded. And third, it helps if we like it.
George stressed something else about critical listening: limit the sections of music to specific passages, and repeat those passages over and over again during evaluations.
What’s so important about being intimately familiar with the playback material? The most obvious answer is that when we are familiar with something, we notice differences very readily.
In a general sense, we best know the sound of the human voice, particularly the voices of friends and family. Thus at least one of the recordings to use for evaluation of a system is something containing a familiar voice, because right away, it should clearly show problems, particularly in the midrange. (As a side note, this is one of the reasons loudspeaker designers work so hard to get things right in the midrange.)
Voice alone usually isn’t enough, since it covers a limited frequency and dynamic range, while we’re usually dealing with musical sources. But the same rule of familiarity applies when choosing material that covers the whole range, because problems with the bass, drums, guitars, etc. should be obvious if we really know the tracks.
One of my favorite tracks was “Famous Blue Raincoat” performed by Jennifer Warnes and engineered by George Massenburg. The voice is so clearly represented, and so are the strings and upright bass – no drums though.
Robert Scovill share a different but related approach with me back in those days. For a Tom Petty tour during the mid-1990s, he recorded each raw microphone input to ADAT (remember those?). Then for the next show’s sound check, he would play the tracks from the previous show to help suss out the system and acoustics of the room.
I thought that was a unique and cool idea, and it certainly satisfies the familiarity issue, with the added bonus of being recorded from live instruments.