Harman is working on a new IEM target curve

[Da cynics]

imaginary room is only needed for spinorama, not for the NFS measurement system itself. That is how I understand it.

Understanding How the Klippel NFS Works

maybe this can set to lower value ? here are waterfall with 0.5 0.2 and 0.1 ms. 0.15 ms (as stereophile use)is not possible in REW. but 0.2 ms look good too. I use for all screenshots 100 segments so you can count the segment lines at 7 khz until it reach 65 db. at 0.5 ms rise time i ccount more...

www.audiosciencereview.com

 

[solderdude]

Of course the Klippel NFS does not need a virtual room. In fact it is designed to ignore the room it is in.
The Klippel is just a microphone that is spun around a speaker and records it from all around and in height (limited).

This would not be possible with headphones as it needs to be sealed.

It is the software (with gating so the room it is in does not matter) that pieces together all the measurements into the plots that can come out.
The prediction of what it might sound like in a room is what the imaginary room is for.
It combines the forward aimed response + all other directions and combines all this info, similar to what a 'proper listening room' would make of it, and produces estimates of the tonal balance.

 

[Jimbob54]

The Klippel is just a microphone that is spun around a speaker

Best not tell Amir he could have just tied a mic to a ceiling fan and saved a fair chunk of his retirement fund.

 

[Sean Olive]

With the Nura (now Denon) automated individual ear adjustment system, there is really no reason for variation based on individuals ears....

Has anyone here tested the Nura earphone that adapts to your ear canal? Or is it compensating for hearing based on otoacoustic emissions?

 

[Jimbob54]

With the Nura (now Denon) automated individual ear adjustment system, there is really no reason for variation based on individuals ears....

Has anyone here tested the Nura earphone that adapts to your ear canal? Or is it compensating for hearing based on otoacoustic emissions?

Oratory suggests the latter- dont know if he has tested them though https://www.audiosciencereview.com/...-on-a-new-iem-target-curve.48524/post-1740822

 

[Sean Olive]

The best part with using the 5128 for IEM's is that there is no official target curve for that fixture yet so when care is taken and well researched this could become a standard.

It would have been incredibly handy if the ear canal bit would be transparent so one could see insertion depth and obtained seal by eye. No such luck with 5128 nor any other fixtures.

I recently spent a weekend measuring IEMS in the 5128 vs the RA0045 coupler and with some practice was able to get consistent seal in both fixtures. I measured 20 IEMs in both fixtures including the original IEM used for our listening tests which we added a MEMs microphone to measure leakage on humans.

 

[alumnicesar]

I recently spent a weekend measuring IEMS in the 5128 vs the RA0045 coupler and with some practice was able to get consistent seal in both fixtures. I measured 20 IEMs in both fixtures including the original IEM used for our listening tests which we added a MEMs microphone to measure leakage on humans.

i heard that that the 5128 measures bass for iems much more similarly to headphones than the gras. potentially not needing an additional bass boost for the target curve. Is that true?

 

[solderdude]

I recently spent a weekend measuring IEMS in the 5128 vs the RA0045 coupler and with some practice was able to get consistent seal in both fixtures. I measured 20 IEMs in both fixtures including the original IEM used for our listening tests which we added a MEMs microphone to measure leakage on humans.

Do you use noise and a 'live' spectrum analysis to pull this off with the 5128 and check if it sits O.K. to take measurements ?
I would assume it isn't as easy as just plugging them in like with the 0045.

In the end... its the final results that count, the result being best plot fit to perceived sound.
Perhaps, some day conclusions will be revealed.

 

[Sean Olive]

i heard that that the 5128 measures bass for iems much more similarly to headphones than the gras. potentially not needing an additional bass boost for the target curve. Is that true?

Depending on the model the IEM measured on a 5128 may have 1-3 dB less bass when measured on a RA0045 based on my experience. It depends on the acoustic impedance of the IEM and the insertion depth. The same trend is seen for headphones in part due to the coupler but also leakage. 5128 is a mannequin and Gras 45 is a flat plate with a pinna which makes for a better seal and less leakage in general. A better comparison would be KEMAR

So hard to say whether the target curve for IEM and AE would be closer for a 5128. Still under review.

 

[Sean Olive]

Do you use noise and a 'live' spectrum analysis to pull this off with the 5128 and check if it sits O.K. to take measurements ?
I would assume it isn't as easy as just plugging them in like with the 0045.

In the end... its the final results that count, the result being best plot fit to perceived sound.
Perhaps, some day conclusions will be revealed.

Yes. I used a 30 Hz square wave and a scope to check the seating. It was the only way I could get a repeatable measurement with a 5128. I also found a pair of tips that seemed to get consistent results. With a RA0045 it’s much easier

 

[Sean Olive]

Yes. I used a 30 Hz square wave and a scope to check the seating. It was the only way I could get a repeatable measurement with a 5128. I also found a pair of tips that seemed to get consistent results. With a RA0045 it’s much easier

I have been thinking of 3D printing a better interface for an IEM than otherwise matches the acoustic impedance of a 5128 coupler.

 

[solderdude]

30Hz SQW would be a lot 'faster' than using noise and spectrum analysis.

I have used 40Hz in the past and noticed how quickly you can see seal getting worse (over-ear headphone DT150 in this case.
Can imagine this works very well with IEM's.

I have been thinking of 3D printing a better interface for an IEM than otherwise matches the acoustic impedance of a 5128 coupler.

So only the outer-ear part and where the earphone is inserted nice and round and the rest of the ear canal and mic being slanted similar to 5128 or some mold of an actual ear canal that comes close to 5128 ?

An idea I once had was to make the spot where an IEM is inserted from transparent material, perhaps even with gratings on it so one can see insertion depth and seal.

 

[Sean Olive]

30Hz SQW would be a lot 'faster' than using noise and spectrum analysis.

I have used 40Hz in the past and noticed how quickly you can see seal getting worse (over-ear headphone DT150 in this case.
Can imagine this works very well with IEM's.
View attachment 322511


So only the outer-ear part and where the earphone is inserted nice and round and the rest of the ear canal and mic being slanted similar to 5128 or some mold of an actual ear canal that comes close to 5128 ?

I tried noise and moved quickly to square waves. I also observed and later confirmed with calibrator there is a 2.7 db difference between left and right 5128 ear. Apparently this within manufacturers tolerance. I guess headphone manufacturers shouldn’t sweat 2-3 db mismatch between channels when a $40k piece of test equipment allows this ;(

It would probably have the same shaped ear canal with ear drum impedance but a simple cone shape canal opening. Maybe clear so you can see ear tip. Thinking out loud here.

 

[amirm]

I tried noise and moved quickly to square waves. I also observed and later confirmed with calibrator there is a 2.7 db difference between left and right 5128 ear. Apparently this within manufacturers tolerance. I guess headphone manufacturers shouldn’t sweat 2-3 db mismatch between channels when a $40k piece of test equipment allows this ;(

I have also heard that there is a lot of drift, requiring constant recalibration. Have you seen this?

 

[_thelaughingman]

I have also heard that there is a lot of drift, requiring constant recalibration. Have you seen this?

If that is true, that's absurd and poor engineering. I work on multi million dollar instruments in the medical field and they can hold their calibration for months on end.

 

[Sean Olive]

I have also heard that there is a lot of drift, requiring constant recalibration. Have you seen this?

I don't have enough experience yet to say .. but will keep a look out.

 

[deadkrillin]

I also observed and later confirmed with calibrator there is a 2.7 db difference between left and right 5128 ear.

You found this difference on every IEM you tested, regardless of seating & other factors? That's absolutely atrocious quality control, if so.

I have been thinking of 3D printing a better interface for an IEM than otherwise matches the acoustic impedance of a 5128 coupler.

The 711 coupler absolutely seems like the more convenient and repeatable measurement device as of right now, and I'd be ecstatic to see something that takes 5128s more accurate acoustic impedance and makes it easier and faster to use. Clear material allowing the engineer to be able to see the ear-tip would be very cool too.

 

[solderdude]

You found this difference on every IEM you tested, regardless of seating & other factors? That's absolutely atrocious quality control, if so.

The difference is in the 5128 fixture (tolerance) Dr. Olive has.
It is possible to match L and R drivers when measured on the same coupler but here too there are production tolerances and usually only more expensive ear/headphones are matched. Cheaper ones usually are not.

Then again my left and right ear canal are different as well in shape so are bound to have a different acoustic impedance and am probably not the only human with differing ear canals.
The brain knows and deals with that through daily experience. A fixture just reports the SPL and has production tolerances.
They should be within a small band for the prices one pays for those fixtures.

 

[Phosphenetre]

Use real life sounds. Record accurately, reproduce so it sounds like what was recorded.

Do you have any reference-level recordings that you use to fulfill this purpose?

 

[solderdude]

The one doing the testing should use real life sounds as a reference. So this requires a studio or knowledge of how instruments recorded should sound.
The recording of it is the difficult part because microphones and their positioning also has an impact on the recording.