Last year my son Noah, then 11 years old and in sixth grade, was looking for an idea for a science project. As I had recently (a few months) completed the Troels CNO-T25 (http://www.troelsgravesen.dk/CNO-T25.htm) and was experimenting with 3 channel stereo, I suggested an experiment. The experiment was prompted in part by my observation that it was not always obvious to me whether I was listening to mono or stereo (from two speakers), and the difference was even more obscured with sound coming from three loudspeakers.
I went to some lengths in earlier posts to explain that mono does not mean “sound coming from one speaker” – mono can mean sound from multiple speakers, it is just the same source signal. I will go into some detail about the experiment, but as a brief summary subjects were not told the purpose of the experiment or other details, and they showed no awareness of when they were listening to mono versus 3 speaker stereo, and in some cases they preferred the sound of mono! I think this experiment has important implications for how all-on-one speakers (e.g., iPod docks, Zeppelin Air, etc.) are, or should be, designed. More about that later.
Before describing the actual experiment, I would like to explain what experiment I ideally would have liked to conduct, and why we did a compromise experiment.
The DESIRED EXPERIMENT
Have people listen and compare the following setups: 3 identical front speakers being fed a mono signal (same signal to all 3 speakers) versus a level matched stereo signal divided by Dolby Pro Logic II to the center/left/right speakers (level matched as the resulting volume of 3 speakers in mono might be louder than 3 speakers in Dolby Pro Logic mode).
WHY WE DIDN’T DO IT
We didn’t have 3 identical front speakers, and subjects might detect a difference between 3 speaker stereo and 3 speaker mono based on tonal or other differences between the center and left/right speakers. We also had no ability to precisely level match between mono versus stereo trials. We could have compared two speaker mono with two speaker stereo, that would have been less interesting.
The ACTUAL EXPERIMENT
Test Question: Can people tell the difference between mono and stereo playback (using 3 loudspeakers for playback)? Is mono or stereo preferred?
Procedure: Two sample songs were imported into a MacIntosh computer from compact disc recordings using Apple Lossless format. During playback, the Mac was configured
to either play “stereo” (different left/right channel sound) or “mono” (the left and right channels are combined so both are the same, however there are still two channels of output). The songs were played with iTunes, and the output of the Mac was sent by digital optical cable to a Rotel processor in “Dolby 5” mode, which sends the left channel input to the left speaker, the right channel input to the right speaker, and the combination of the two channels to the center speaker. When switching the Mac from mono to stereo, there is no change in volume or balance between the speakers, the difference is that, coming out of the Mac, the left and right channels have different signals in “stereo” mode, whereas the left and right channels have identical signals in “mono” mode. In terms of what the listener hears, the center channel speaker produces the same sound in either stereo or mono mode. For the left/right speakers, however, mono/stereo is different: when the Mac is configured for mono, the left and right speakers produce the same sounds, whereas when the Mac is configured for stereo, the left and right speakers produce different sounds.
Test subjects listened to a one minute portion of song A three times, and were told that two of the times would be alike and one time would be different. Then they listened to a one minute portion of song B three times with the same protocol as for song A. The subjects were not told the purpose of the experiment, and were not told what was different between versions (no mention of mono or stereo or playback settings). The subjects were asked to identify which of the three playings of song A was different, and which playing they liked best.
Two people conducted the experiment, my son and I. To have a double blind procedure, one person operated the computer and one person interviewed the test subjects and
recorded results; only the person operating the computer knew whether the song was being played in mono or stereo. The person operating the computer played the same
song three times so that (in random order) one time was mono, one time stereo, and one time was either stereo or mono.
There were two sample songs: song A was the pop song “Beautiful Monster” by Neo. While the recording for song A is “two channel” with different sounds in the left and
right channel, it is not an acoustic stereo recording, and likely most of the vocals and instruments were recorded in mono and then artificially mixed to create the
final two channel recording. Song B was a jazz tune “China Boy” that seems like a “real” stereo recording with Ken Peplowski (clarinet) and Frank Vignola (guitar). I
found a youtube clip of Ken and Frank playing Tiger Rag, sort of gives the idea of China Boy: http://www.youtube.com/watch?v=9MUwb3eNZzE.
In the next blog post I’ll discuss the results.