Re: Headphones for testing

[Steve Beet <steve.beet@xxxxxxxx>]

Dear Dominic,

I wouldn't want to scare you, but I think it may be even more complicated
than Al has suggested.

If you want to do experiments on 3-d localisation, for example, even the
position and the size of the acoustic transducer in the headphones can make
a difference. If the headphones are circumnaural, you need as large a
transducer as you can get to ensure the sound field has as flat a wavefront
as possible (as if the sound was emanating from a distant source). Even
then, the direction of incidence on the subject's pinna won't match that of
a real-world stimulus, so the pressure wave reaching the eardrum won't
(usually) be exactly the same as it would if the acoustic source was at a
distance from the subject's head. There are many other issues regarding
frequency response, phase response and distortion.

If you're concerned with musical perception and consumer headphones, the
Headroom website is worth a visit:

http://www.headphone.com/learning-center/headroom-test-lab.php

They have various measurements and reviews of headphones and the site can be
quite educational.

If you want repeatable experimental results, I'd recommend the most
expensive in-ear headphones you can afford, with the widest frequency
response and lowest distortion. A wide frequency response is usually an
indication of a good (linear) phase response. There may be hygiene issues if
the headphones are to be used by a large number of subjects, but as long as
they are fitted carefully and correctly, in-ear designs can reduce the
variability in experiments.

If your experiments will involve large numbers of subjects, you may prefer
circumnaural designs - then you'll need to consider robustness and comfort
too. I recently bought a pair of Beyer Dynamic DT-660 headphones on Amazon
for about 90 pounds. They're circumnaural with an unusually flat and
extended frequency response, low distortion, and they do show up a lot of
detail in audio stimuli. However they're not really robust enough for
large-scale experiments and they're intended as consumer headphones, so are
more concerned with making music "sound" the same as it would on speakers,
rather than reproducing and exact copy of the pressure wave at the eardrum
of the listener. Oh, and I should mention my eardrums are over 50 years old
now, so anything above 14 kHz is a mystery to me!

In summary, it's a complicated area but if you've "only" got 100 pounds to
spend, you may not have much choice and you almost certainly can't afford
calibrated headphones.

All the best,

Steve Beet

_______________________________________
Principal R&D Engineer, Aculab plc
Milton Keynes, UK



> -----Original Message-----
> From: AUDITORY - Research in Auditory Perception 
> [mailto:AUDITORY@xxxxxxxxxxxxxxx] On Behalf Of Al Bregman
> Sent: Thursday, January 21, 2010 8:56 PM
> To: AUDITORY@xxxxxxxxxxxxxxx
> Subject: Re: Headphones for testing
> 
> 
> Dear Dominic,
> 
> You didn't say what the nature of your research is.  It makes a lot of
> difference whether you are studying masking of one pure tone by
> another or the emotional effects of intonation patterns in speech.
> Not all studies require the same quality headphones.
> 
> You also have to think about loudness calibration.  Standard
> audiometric headphones with hard circumnaural cushions are calibrated
> with a coupler of known resonance characteristics.  Open-air
> headphones, such as many Sennheiser models,  cannot be calibrated this
> way.  I have used both types.  The open-air are a lot more
> comfortable, but require the use of a flat-plate coupler (which we had
> to manufacture ourselves) for intensity calibration, and the
> measurement depends on how well centered the headphone is on the
> coupler.  All this may be irrelevant for you, depending on how precise
> a delivery you require.
> 
> If you are using a separate machine for applied EQ, this can produce
> phase distortion.  Again, this may be irrelevant for you, depending on
> the nature of your research.  If you are digitally synthesizing your
> signals, and apply the equalization at this stage, there may be no
> phase distortion (except in the headphones themselves).
> 
> You have to specify your needs more exactly before you can decide on
> the headphones.
> 
> Cheers,
> 
> --  Al
> --------------------------------------------------------------
> Albert S. Bregman, Emeritus Professor
> Psychology Department, McGill University
> 1205 Docteur Penfield Avenue
> Montreal, QC, Canada H3A 1B1.
> Office:  Phone: (514) 398-6103, Fax: (514) 398-4896
> http://webpages.mcgill.ca/staff/Group2/abregm1/web/
> --------------------------------------------------------------
> 
> On Thu, Jan 21, 2010 at 3:11 PM, dominic ward
> <contactdominicward@xxxxxxxxxxxxxx> wrote:
> > Hi all,
> >
> > I am an undergraduate student looking for advice/recommendations on
> > suitable headphones (currently available) for psychoacoustic
> > experiments as part of my project. I understand on ear frequency
> > response is of great importance and is never truly flat, but would
> > applied EQ be suitable for compensation? My budget is 
> around £50-£100.
> >
> > I have e-mailed manufacturers such as Sennheiser but sadly 
> no response.
> >
> > Any support is much appreciated,
> >
> > Many thanks,
> >
> > Dominic.
> >
> 
> 
> -------
>