I have never given up recorded music. There are two states for me. One is enjoying music on just about any sound system, ranging from my clockradio to Bose 901's that hang in a bar I used to frequent. The other state is where I listen to music on a sound system for which I am responsible. That can be a nagging experience and it is only since about 10 years that I build speakers which are good enough to not be distracted by stuff that does come out of my stereo yet shouldn't be there.
You have to place the board directly in front of your head with your nose barely touching it. Not between the speakers. 100x80cm is probably too small.
Place the speakers as you normally would (equilateral triangle) about 2-3m in front of you.
Then listen to Paul Lansky "Chatter of Pins". First without the barrier and then with the barrier. Close your eyes in each case and try to localize the sounds.
No difference?
I measured at least four of different sizes, the smaller ones like the Apple Earpods are pretty flat. A search through some database show that most have Fo between 100~200Hz probably 6 months ago I just picked one curve to use as a basis to flatten the impedance curve and tried it on different head/ear phones, improvements were of varying degrees depending on the quality of the head/ear phones used. There was one test circuit that seemed to work best for the majority of head/ear phones that I used, but I have no explanation why. Plan to try some different things next year. My guess is reducing back emf into the driving circuit. Currently I would speculate the smaller devices introduce back emf over a wider spectrum, but I have not yet tested this.
Have you listened to the Apple EarPods?
Last edited:
How? If you are prepared to use a windows PC then Convolver Convolver — a convolution plug-in will do it with 4 suitable IRs.
For linux there is Jconvolver and BruteFIR.
Raspberry PI would do it to file, but I have not managed realtime, as yet.
The wide variations of frequency responses would seem to be a bigger problem:
Learning Center - Build a Headphone Graph | HeadRoom Audio
See also:
http://www.davidgriesinger.com/Binaural_hearing_and_headphones.ppt
I have held a pillow in front of my face with good results, but it is not comfortable to listen long times in that position.
This is a very good presentation. A very important point is that the ear canal transmission has no directional dependency but it significantly differs from person to person causing serious spectral problems.
We would need to measure at the eardrum to circumvent this issue. Pretty tricky (and dangerous).
Last edited:
I would be real curious if anyone has actually measured the impedance rise with differing headphones and not just assumed that they would all be the same. I would think that an in ear would be very different than an on ear system and a major difference would show up between a dynamic driver and an electrostatic model.
On listening to headphones in general there is just no way with a conventional headphone to create the signal path anything like a stereo pair of speakers. Human hearing does not follow the model of two ears in isolation, you can never get there from here kind of thing, unless of course we are going to include open air headphones and lots of bleed to allow for cross talk for both ears at least.
On listening to headphones in general there is just no way with a conventional headphone to create the signal path anything like a stereo pair of speakers. Human hearing does not follow the model of two ears in isolation, you can never get there from here kind of thing, unless of course we are going to include open air headphones and lots of bleed to allow for cross talk for both ears at least.
David,
No I have just popped in from time to time to see what is being said. I did see the comment a couple of posts back not from you about using a simple zobel of the same value on multiple headphone models. That seemed incorrect on its face, just assuming they all follow a similar impedance path. I will go back and look at your links and see what you have posted. But I have read enough other studies to understand that a headphone can not actually recreate the entire sound field through any type of electronic manipulation. It is the problem with how the ear works and bone vibrations and other aspects that you just can't duplicate with headphones. I am not saying that all headphones sound bad, not in the least, but it is not the same experience of hearing without the isolation that any headphone is going to produce. I still remember some very expensive Sennheiser electrostatic headphones years ago that were about as good as they come, very good but not dynamic speaker sound or live event sound.
No I have just popped in from time to time to see what is being said. I did see the comment a couple of posts back not from you about using a simple zobel of the same value on multiple headphone models. That seemed incorrect on its face, just assuming they all follow a similar impedance path. I will go back and look at your links and see what you have posted. But I have read enough other studies to understand that a headphone can not actually recreate the entire sound field through any type of electronic manipulation. It is the problem with how the ear works and bone vibrations and other aspects that you just can't duplicate with headphones. I am not saying that all headphones sound bad, not in the least, but it is not the same experience of hearing without the isolation that any headphone is going to produce. I still remember some very expensive Sennheiser electrostatic headphones years ago that were about as good as they come, very good but not dynamic speaker sound or live event sound.
With a blocked meatus recording I don't think you need to know what it does, as long as the headphones do not disrupt its natural resonance., as long as the EQ is right for speakers. An ear canal earphone certainly would, changing an acoustic open circuit into a short circuit..
I guess it would depend on the enclosed volume, mainly. You will get refections between the pinna and headphone driver/case, but that is another issue. I have added some damping, which helps, but you lose some top end. Ear speakers should be good, but I am not about to buy Stax. It has to work reasonably well for the poor, like me, or it is not going to catch on. Some headphone response standard would be a huge help, I think..
Shhhh......
That's giving it away. I'll be using the convolver in my PC player. Already have some nice spaces lined up for the impulse responses.
No problem. I thought people might have missed the fact that you can display impedance. (although not the complex components ). You got the nudge..
I disagree that what you seek is not possible over headphones. I think it is within spitting distance. It depends to some extent on whether the 'bone conduction' sound ends up at the oval window the same as it does with the air path, or whether we hear the 'sound' part of the vector with our ears, and the direction part (which has no sound itself) with say the hairs on our feet. Feet seems unlikely, since the direction detection is not degraded by wearing wellies. Similarly you can remove (or greatly attenuate) all your HRTF aparrt from one ear by immersing all exept one ear in the bath, and still hear direction. On the other hand ypu can screw your pinnae up, fit ear plugs, and wear industrial ear muffs, all at the same time. If you can still hear any sound, you can tell its direction. Confusing or what!
I might just address a point that picowallspeaker referred to:
Frank
When "special" sound is achieved this type of appraisal approach becomes completely irrelevant. It would like sticking an air speed instrument out of the window of a Ferrari doing 150mph, to ascertain whether the car was going impressively fast or not -- if you can't trust your physical senses to send you the message that things are hummin', then it ain't happenin' ...
Frank