EMRFD Message Archive 9062
Message Date From Subject 9062 2013-09-02 19:32:54 Rick KK7B Binaural experiments--long post Hi All,
There has been a recent resurgence of interest in binaural listening techniques, with all the usual old questions being asked, and answers posted on the web by a few with some experience and many more who have may have read about it once or twice.
My own binaural listening experience goes back 35 years, and I certainly was not the first to experiment with binaural shortwave receiver outputs. The same technically adept amateurs who designed and built all their own ham gear in 1950s also designed and built their own Hi-Fi systems--The Bill Orr Radio Handbook from that period includes a HiFi amplifier. The introduction of stereo in that decade would have naturally generated questions and curiosity in a community more willing to experiment than todays audio and radio technophiles. Why experiment when you can simply go on the world wide web and find the pontifications of some expert to quote to your beginner friends and students?
So I'll begin by asserting that more experiments are needed, and the results of those experiments will likely generate more questions than answers. Observations of binaural listening include, by definition, your own personal ear-brain signal processor--and mine is different than yours. The best symphony conductor I ever played under could detect the particular pad on an English Horn in the woodwind section that needed a little work. I have designed, built and used a number of very different receivers with binaural outputs, and shared them with friends and colleagues who have put them through a wide assortment of experiments different operating environments. A few general observations have emerged. I'll use the old terms monaural and binaural to differentiate between receivers that present the same, or different information to the two ears.
Here are some observations:
Whatever means you use to generate a binaural signal, it sounds "good" when you switch from monaural to binaural.
If you are trying to copy a particular weak signal in a crowded band with interference, noise, lots of other signals, static crashes, etc. then employing the full array of modern analog and digital signal processing techniques to enhance that signal and suppress everything else is a more powerful technique than simply presenting it all and letting the ear-brain processor try to dig it out. After you have done everything you can with narrow filters, noise blankers, tone controls, etc. there seems to be little benefit to splitting the remaining signal in two and presenting something different to the right and left ears, other than possibly relieving operator fatigue.
If, however, you are interested in more than just one signal--if you want to hear the character of the band, discern how far away the static crashes are coming from, listen to a half dozen CW signals at once and instantly tell which ones are local, which ones are single and double hop, and which ones are coming over the poles, then you give ears as much information as possible.
A wide open binaural direct conversion receiver lets you instantly hear the difference between local and distant noise sources. People who are interested in noise seem to find them particularly appealing.
Binaural diversity receivers--those with identical channels and two different antennas--allow the ear-brain to synthesize a two-element array. This can have real operational advantages when operating with a number of signals arriving from different directions, with randomly varying polarization, or with a noise source that arrives from random directions--for example static crashes or ignition noise near a highway.
If you have two antennas and a single dominant noise source or interfering signal, it is generally more useful to combine the two antenna outputs in a hardware phase-amplitude network to null the noise or interference, and then send the resulting channel to a conventional monaural receiver system.
Some metaphors have been suggested for the sound of a wide-open direct conversion receiver, and "a window on the band" has been used for decades. This analogy can be extended to include both modern interference fighting systems and binaural techniques. Suppose that you want to have a quiet conversation with some interesting person. You invite him into your home, close the door, turn off the TV, and sit face to face. That is what a conventional monaural receiver tries to achieve, and we refer to it's success as "armchair copy." Background noise is the hum of the refrigerator in the next room, and interference might be a ringing telephone. These distractions, like modern DSP artifacts, are unrelated to the signal.
A basic direct conversion receiver with unrestricted audio bandwidth and equal response on both sidebands is more like opening a window and looking out at the band. There are many more signals, and they are all observed in their natural environment. Noise and interference are related to the signals, and interesting.
A wide-open binaural receiver, either IQ or with two antennas, is more like stepping outside and walking around inside a 3 dimensional band. Everything is there. On an weekday evening in the mountains with a good antenna and an open band, the experience is a delight to the senses. During the November CW sweepstakes, it's more like wandering around in traffic on an LA Freeway.
In my experience, the simplest possible implementation of a binaural receiver--a wide-open stereo amplifier with headphones, a single LO with an IQ splitter and/or separate antennas on a pair of product detectors--is the also the most capable configuration. Every technique I have added to the basic binaural receiver to enhance the ability to copy a desired signal--restricting the bandwidth, binaural post-processing after eliminating one sideband--has had a limiting effect on the 3D experience of wandering around in the band. I may still experience a 3D effect, but I'm wearing a raincoat, big hat, there's dense fog, and the only thing I can really see is the guy standing right in front of me. I use monaural techniques when I want to copy a particular signal, and binaural techniques when I want to experience what's happening on the band.
9065 2013-09-04 05:05:23 Johan Bodin Re: Binaural experiments--long post Hi Rick,
thanks for an interesting post! My experiences of binaural listening on
the HF bands are very similar to yours.
While on this subject I would like to mention ISB reception of distant
AM stations. Tune the receiver exactly to zero beat and feed LSB audio
to one ear and USB audio to the other. Frequency selective fading gives
an almost physical sensation to the brain! I have tried this a number of
times with a communications receiver which has two separate audio
outputs for ISB. BTW, it should be quite easy to modify a phasing DC
receiver for ISB reception just by adding another "summing point" after
the audio phase shifters, with one of the I/Q channels inverted, and
another identical audio bandpass filter and audio amplifier for the new
Rick KK7B wrote 2013-09-03 04:32:
> Hi All,[...]
> There has been a recent resurgence of interest in binaural listening techniques,
> A wide-open binaural receiver, either IQ or with two antennas, is more like stepping outside and walking around inside a 3 dimensional band. Everything is there.[...]
> In my experience, the simplest possible implementation of a binaural receiver--a wide-open stereo amplifier with headphones, a single LO with an IQ splitter and/or separate antennas on a pair of product detectors--is the also the most capable configuration.
9067 2013-09-04 12:35:37 Rick KK7B Re: Binaural experiments--long post Hi Johan,
Yes, I first read about the ISB AM stereo technique with locked carrier in 1974, and have since found some references going back to the 1950s. I've done some casual experiments over the years, but finally built a good receiver specifically for the technique a few years ago--phase locked channel to recover the carrier, and identical phasing channels for the upper and lower sideband. It is quite striking, particularly with wide open channels of at least 5 kHz