TheRadioBoard

I've been doing a lot of listening tests to try and assess the different headphones I've collected over the last few years. One problem (if you have not got any sound level measurement equipment) is remembering the loudness of different 'phones during the intervals of time it takes to change headsets. I'm very bad at this and a bit of research says that people in general make big errors on this kind of comparison.

I remembered some lab. work I did in my twenties as part of a psychoacoustics course. The task was to find the 'just noticeable difference (jnd)' in sound level at a particular frequency. One way of doing this was to vary the sound level on one ear so as to find the minimum change that was just detectable. The sound level on the other ear is held constant so you are making an uninterrupted comparison - no unreliable memory task involved. (You have to reverse the measurement across the ears to balance out any difference in sensitivity between the two ears.) It turns out that the auditory system is quite sensitive to very slight sound level changes when measured in this way. Less so if you have significant hearing loss in one ear!

All this means one can use different 'phone elements on the left and right ears to make (admittedly rough and ready/subjective) loudness and tonal comparisons. Of course the phones need to be fed from the same sound source (Xtal set). Differences in sound level manifest themselves in the form of an apparent shift in the location of the sound source between the two ears. If the sound seems to come from in the middle of the head the phones are producing almost (or actually) equal sound levels. There may still be a physically measurable difference in sound levels but the great thing is you can't hear it!"

It wouldn't satisfy those who want hard figures in energy levels and decibels but then most of us only want a subjective assessment of which phones work best.

Comments and criticisms welcome.

_________________
John

I use a headphone horn speaker and a relatively good microphone fed into my stereo to do a comparison on the recording DB meter.
High fi headphones sound louder to me probably due to my poor low end hearing but are not always the highest on the DB meter when compared with vintage headphones.
Any comparison method is worthwhile. Variances in results can a lot of times be logged to the age and condition of the phones. I've also found variances in the same brand/model headphones that are made a little different.

My pair of vintage Brandes 'Matched Tone' headphones are my favorites to use with crystal radios. I have a pair of Stromberg-Carlson Sound Powered headphones, but I can't seem to make up my mind if they are really better than the Brandes or not. It's hard to beat those Brandes!

Norman
KA1GUK

You are comparing a set of phones designed for fidelity to a set designed for voice only, blocking outside industrial and military noises (and most of the music range). RCA big cans give me a headache when using them for easy listening.

I have also collected a few but I didn

Hello John,

As you said accurately remembering a certain sound level is just about impossible for most people. The use of a reference phone element is a great idea but is has its problems too.

When comparing phone elements a good impedance match is required. If the elements are of the same type the matching issue is less important. But if a 600ohm element needs to be compared to a 2000ohm element, two matching transformers are required and both transformers must have identical efficiencies.
If the transformer efficiencies are not identical, or not known, you can still do the comparison but you’ll have too realize that you are comparing two matching transformer+phone element combinations and not just two phone elements.

I could happen that the most sensitive system (matching transformer+ phone element) doesn't have the most sensitive phone element but has highly efficient matching transformer.

Corné

Yes Corne, of course you are right. I had overlooked that. I was mainly think of comparing different examples of the same sort/make/model of phone.

Even in such cases I am now aware there can be significant variations in impedance values.

John

_________________
John

Like 29RAdHead75 I too have a set of the AGS Reflex Model V's and find them to be quite sensitive and very comfortable for lengthly listening sessions. I use the 8 ohm winding (the pink pair) of a Bogen T725 to drive them on the tip and ring connections only.
This puts the two elements in series which works the best, but unfortunately I don't know what their impedance is. Any ideas on their Z or how to measure it would be appreciated. For thirty plus years old they're a great set of 'phones.

_________________
To be where I am, I had to leave where I was!

My background includes alot of experience in broadcast and production studios. Also I like headphones.

As others have said:
- sensitivity is a function of the whole audio chain, so you probably need to treat as combinations - what headset with what transformer, and have the tested sets roughly equivalent in terms of presented load.
- "Better" is subjective. Quantitative assessments are possible (instruments and experiments) but some 'phones may simply sound clearer or "nicer" without an easily-measured reason.

You might want to rig up an A-B switch using a DPDT switch and binding posts or jacks, which would allow you to quickly switch between two headset/transformer combination, as you quickly change headsets on your head.

I have several different types of headphones around the house, from pink crystal earpieces to AKG K240's. For crystal radio or small regens I have a pair of Vietnam-era HS-16A phones and an old UTC output transformer that seems to work well. But they still hurt after a while. Of course, for the price of 4 transistors, or one op-amp, or a LM386, you can drive just about any headset. My most-used general radio-listening headphones are a battered set of mid-range SONY MDR-V3, on a custom headband, with donuts cut from a mouse-pad as earpads.

You need an audio oscillator, an audio amp with speaker out, a proper audio voltmeter, some resistors and a handful of clip-leads.

I'd go into more detail here, except I found this link which explains it all much better:

http://www.bentongue.com/xtalset/2meaphi/2meaphi.html

You can actually do the tests with one resistor instead of a custom impedance box, and use voltage divider equations to solve for the effective impedance.

This is how I measured the impedance of my headphones.

Corné

Kenn & Corne,

Thanks for the links.....I'll see what I can do with my audio signal generator, which can output close to 30 volts into a 600 ohm load, and my DVM.
It's more a case of curiosity than anything else, since I'm happy with them. I use some telco elements with ear defenders as my main 'phones.

73, J.B., VE3EAR

_________________
To be where I am, I had to leave where I was!