Feb 7, 2013

WSA 160m (1)

WSA160m, choosing a headphone

Weak Signal Audio 160m, a new project

When searching for progression, sometimes there is a need to look at matters from a totally different perspective.
In our efforts to dig deeper in the band-noise, opening up another layer, making new CW QSO’s possible, we decided to do a different investigation and perform some tests. This project does not touch on antennas or any RF/IF hardware. Instead, those steps from the detector to the 'decoding by human brain'. From your AF stage, via any transducer, passing ear-wax to that final conversion into gray matter signals.

Why discuss the use of headphones?

There's several potential topics to be discussed along this extended audio path. At some point we need to look at headphones, and perhaps pick one or more relevant types who might serve during future tests. Questions arise; are there relevant differences in headphones when it comes to making that extra QSO? What other parameters exist next to the obvious ‘wearing comfort’? Is there a single winning model or do you individually need to test-find a match for the operator?

List of headphones under test

No effort was made to add any specific commercial popular model or brand. We just grabbed as many HAM-Radio, Hi-Fi, DJ, Advanced Gaming and/or your typical MP3 headphones as we could get. All full-range stereo models. Some equipped with in-line volume control(s).

Medium sized “on-ear” models:     Philips, Sony
Large sized, fully ear-surrounding models:    Heil, MFJ, Beyerdynamic, David Clark, Philips
Small sized “in-ear” models:    Koss, Sony, Philips

We started with more than 20 models, slowly narrowing it down to about a handfull which offered small advantages in sudden areas.

Just a small bunch of the tested headphones

Comparison test conditions

The sound source was 160m recordings from CQWW-CW- 2012, using the Microtelecom Perseus. Filter taps were set to 32, as this enhances weak signal detection and avoids typical digital harshness which is common on some DSP's . During the first weeks the internal sound card of the laptop was tried, evolving to external DAC units, different external (Hi-Fi) amplifiers, finally settling for the premium quality Objective 2 Headphone amplifier. The O2 is an open source amplifier designed by NwAvGuy, with emphasis on benchmark performance and low cost.
http://www.jdslabs.com/item.php?fetchitem=o2full Later we also added the O2 USB DAC combo version. 
You can write a separate document on the required amplifier performance and correct matching to any type of headphone. Read more at: http://nwavguy.blogspot.nl/2011/07/o2-headphone-amp.html 

The Objective 2 headphone Amplifier

The listening test

From Early December till mid-January we spent two evenings per week listening to many different, but carefully selected time segments of the contest recordings, offering:
  1. Weak signal alone in the band noise
Extreme weak signals in/out of the band noise being extremely difficult to read, let alone decode for a valid QSO. Sometimes under lots of QRN and/or slow QSB. Listening for a longer duration to any CQ’ing station being called by far-far-away-dx. Sometimes we would eaves-drop on a CQ-√≠ng big-gun, switching his 4SQ 180 deg. back and forth.
  1. Weak signal close to a loud signal
This is also a dynamic range test; how does it (and your hearing) deal with weak signals in an environment of the complexity of loud sounds and quick volume changes.

Our main goal was to distinguish between:
  • How easy/difficult is it to pick and decode a weak CW signal?
  • Which is offering you a better reading resolution while tuning the band?
  • What other audible effects, good or bad are noticed?

It was not about the offered sound quality as such. This was about making an extra QSO Y/N, next to being able to listen over prolonged periods of time, either comfortably scanning the band or monitoring QSO after QSO of some running Big-Gun.
Surely we were subject to the typical learning/recognition process when repeating sound clips, listening level differences and the inevitable ‘sound memory effect’ when doing endless A/B/X comparisons. But along the way several headphone models started to show their merits in one or more areas. This was not about doing a headphone shootout, but achieving our goal by learning from all aspects.

Noise cancellation headphones

A few years ago we did several tests, using different in-/on-/around-ear active&passive NC models, both at home and our club station PI4TUE. During 48hrs contests with many operators available, we received a lot of feedback. At best the NC models would offer some cancellation of the high environmental noise at low frequencies. Rumble and fan-noise from > 20 network servers in the contest room was attenuated to some extent and did offer some relief. But the active models showed other nasty side effects like additional internal amp noise, sound distortion and dynamic range issues. The passive versions, with ear surrounding gel-pads on the other hand, performed outstanding. Today, 3 operators still use these at home and in contests. Only the passive noise cancellation gel-pad version headphones have been used in this test.

Test results

High sensitivity HP reveals internal AF amplifier noise &hum
Every hum, noise, pitched distortions (audio birdies) of the AF amplifier are being heard. And this effect becomes much worse going to premium and high-end range models. The more expensive the headphone the higher sensitivity they usually offer. You need a fairly loud source input volume to overcome this amplifier QRN. 
AF amp internal noise/hum level must be very low
That is why we quickly went to search for the best available headphone amplification ham operator budget can afford. You can build your own low-noise headphone (distribution) amp, but don’t take it too far; read the chapter at ‘Designer Components’   From my past years at service and design of consumer audio Hi-Fi equipment, I recognize these statements all too well.
At future test we will show AF spectrum outputs of various HF transceivers.

Presence of AF background noise affects the ability to distinguish weak signals
We are subject to back ground noise in daily life and have become used to it. The fixed level of AF amplifier noise in the background creates an unconscious distraction. This became very apparent when using the in-line volume controls; making them less sensitive effectively pushed the AF noise level below the hearing threshold. You only recognize this benefit when you free yourself from the distracting noise. However, now offering much higher impedance to your amps output might affect its performance in other areas. Again a reason to switch to a quiet amplifier, also capable of handling a wide impedance range.

Lower range of audio spectrum HP allows for disturbance, like rumble or typical key click thumbs
Again, premium headphones reveal them pretty loud. Today, many on- and in-ear models are ‘tuned for extra bass’. Next to your CW passband filter settings, additional high pass filtering is a possible solution. But not used during this test.

Upper range of audio spectrum HP does not contribute to readability
Obvious and confirmed while using some experimental settings of filtering, sound equalization etc. A dedicated lowpass filter is under design, but never used in this test.

Open system on HP allow for distraction of environmental noise, decreasing focus over time
No problem if your shack is a remote, dead quiet man’s cave. Whatever your preference, you need at least some closed ear-protection, so you can concentrate on the job. We find open systems only usable for what they are intended for; playing music.

Premium High Fidelity HP offer an aggressive time envelope on noise/impulses/key-clicks or any radio communication sound. This is much fatiguing.
Whatever the great detailed sound and high dynamics offered by the premium to high-end quality headphones, they have a serious downside. They closely follow every quick rising signal and reproduce a ‘truth’ which becomes quickly annoying. Interesting during the first minute, but soon becoming unpleasant for your hearing. It’s like listening with AGC OFF, times 100. Even clear band noise slowly becomes fatiguing for your hearing. Let alone any nearby key-clicks, pops etc. For those using any antenna pre-amplification, be warned; it’s like sticking a needle in your ear every time the neighbor flicks a light switch. The overall result is a rough, harsh sound experience. Despite the much more detailed sound, there was never an advantage while trying to detect/decode CW weak signals.

Communication drivers/cheaper HP models seem to have a limited rise-time, thus masking the aggressive time envelope noise/impulse effect
For recognizing and decoding weak CW signals, the low-end or cheaper models performed equally well, less the harsh and quickly fatiguing sound experience. A cheap hamradio boom-mic headset was preferred over its premium competitor. I recall that exact same experience when I bought the high-end model after having used the former for several years. But the larger, much more comfortable ear pads and sturdy mic-boom did make a valid difference. I only use it when making a phone QSO on 50MHz.
Mind you; the time envelope of CW signals is relatively slow. And their sound environment (band noise, off-pitched CW signals)is such different, that we do not need any additional sound resolution to distinguish them better.
Premium headphone models just offer too much; less seems better.

Communication drivers/cheaper HP models seem to have a mechanical threshold. This can offer less detail at very low listening volume, perhaps also masking weak signals by amplifier noise
Although we were splitting hairs, a few low end models showed this effect. It became more apparent while using noisy amplifiers and/or connected to some of our HF rigs. Turning up the volume cleared the effect.

Louder is NOT better
Turning up the volume for completing that difficult QSO is not enhancing your hearing ability.
On the contrary: your hearing ‘Auditory Filter’, which allows for max distinction of isolated signals, is negatively affected: http://en.wikipedia.org/wiki/Critical_band Read more about your hearing at: http://projekter.aau.dk/projekter/files/9897666/Auditory_filters_at_low_frequencies_07gr1062.pdf 

Interesting remark: “another important effect of loudness perception is that not every sound that is above your threshold of hearing can be heard. If a sound is given, while a near louder one is also present (within the auditory filter) then it is possible you do not perceive the lower one. This process seems intuitive, but on the psychoacoustic and cognitive levels it becomes very complex. The term for this process is masking.” In short: your hearing ability has a minimal bandwidth and a varying curve.
MP3 encoding uses this masking effect of human hearing to lower the data rate, without affecting the overall ‘sound quality perception’.
Be cautious about treating your hair cells. You want your personal auditory filter curve sharp and symmetrical. Preserve your hearing!
The above shows a typical hearing of a young, healthy person. Now see what has happened to those suffering from hearing impairment:
Guess how you are going to compensate for that lack of hearing ability
Right; you will turn up the volume even more....
And exposure to loud listening volume slowly might activate your ‘Acoustic Reflex’, the attenuation reflex in your middle ear. And again you compensate by reaching to that volume knob.
Although turning up your listening volume seems effective sometimes, it certainly has more to do with your current focus and perhaps the continuous distraction from environmental noise.

Closed HP systems with passive noise cancellation allow for high focus at an average lower listening volume. Separation of weak signals is easier at lower audio levels here.
The ear surrounding gel pads model from the professional aviation headsets offer clear, isolated reception of weak signals in noisy environments. They avoid listening fatigue in lengthy contests. The gel pad aviation model was friendly for the ears, even after many hours of contest use. But one of our clubstation operators perceived the ‘ear sealing effect’ as uncomfortable. Another instead, uses a generic hearing protector around his in-ear headphone, like drummers do in their studio.

And some final remarks

'Clean' TX signals, with less sidebands/key-clicks, resulting at AF level in more approaching a true sinus character, are easier to decode as a weak signal
Sound timbre and tonal coloring are a matter of personal choice, as is wearing comfort
Passive/active sound moderation like limiting and/or low-pass filtering may have a pronounced effect, but are to be proven as we have not used these in this test.

Rens, PA3FGA
Mark, PA5MW