Headphones, Headphone Amplifiers, Tools and Methods

[DualTriode]

What do you think qualifies a headphone to be a good chameleon?
I would translate that into a consistent inter-subjective seal and PRTF.

By the way...

View attachment 102447

I hope you don't mind the meme.

In my opinion leave the ear transfer function out of it. there are too many ears including artificial ears to confuse objective measurements of the headphones by themselves.

All the high frequency reflections in the ear and off the ear removes any objective measurement of the headphone frequency response.

Transfer functions can be measured / calculated later.

From my reading of the Harman multiple regression statistics applied in the chameleon research different shaped ears did not have much weight in determining preference.

Thanks DT

 

[thewas]

In terms of data some other brand has some catching up to do to prove itself.

Rather on pseudo science marketing.
But at least they finally managed to make one headphone is quite closer to their own target (the 371) than the old 25 year old Sennheisers were.

 

[DualTriode]

First, multiple Regression statistics published in an AES paper is not pseudo science marketing.

Hello All,

UPS does make deliveries in the rainy North West. Late yesterday the big brown truck delivered a Lynx Studio Technology E22 PCI Express Card - Audio Interface (2 x Analog / 2 x Digital) sound card. This card is recommended by Audio Precision to operate with their Flex version of APx500 V 6.0 analyzer software. The self, closed loop, noise and distortion floor of the E22 is nowhere near the APx555 however still better than the self noise and distortion of the 45CA-9. The object here is to get rid of the environmental noise on the test bench generated by the APx555 cooling fan. A topping L30 headphone amplifier will replace the APx1701 and its noisy fan. Room noise is much improved without the instrument fans. A little side note, the Topping L30 headphone amplifier has a much improved SINAD over the 1701 power amplifiers.

Next is to install the E22 sound card in the Coffee Lake computer and do some closed loop tests of the sound card with the APx555.

Thanks DT

 

[thewas]

First, multiple Regression statistics published in an AES paper is not pseudo science marketing.

It isn't, but this is https://www.audiosciencereview.com/...man-curve-for-loudspeakers.18782/#post-614592

 

[DualTriode]

Hello All,

Today I played around with ASIO sound cards and APx500 Flex software. The goal is to run headphone measurement without the LOUD APx 555 and APx1701 cooling fans. Electrically as you know the APx555 is the most quiet. The APx1701 compared to the APx555 is just so so.

To get rid of the 1701 cooling fan noise a Topping L30 headphone amplifier will take its place on the bench, today’s test runs were with the RTX 6001 inputs and outputs connected together closed loop at 1 volt output. We will setup the test gain and plot scales later.

We are not looking to best the APx555, we are trying to turn it and its’ cooling fan off to make the test ambient noise a lot of dB’s quieter. The goal is to have less ambient noise and still have a margin between the ATX 6001 THD+N and the test microphones in the 45CA-9. We have met that goal with both the Lynx E22 and RTX 6001. So far the RTX 6001 tests better than the Lynx E22 card.

The RTX 6001 has a SINAD of 94.8dB and a THD+N of 0.0018%, much better than the 45CA-9 and preamplifiers will perform. The Lynx E22 will also do well.

See the attached RTX 6001 FFT plot.

Thanks DT

 

[DualTriode]

Hello All,

Whoops I forgot to pack the APx 500 Flex software USB dongle, the AP Flex software will not function without it. So meanwhile the APx555 including fan is the stand-in AP dongle required to operate the ASIO controlled third party analyzer hardware. The AP Flex USB dongle will be back in hand in a few days.

Thanks DT

 

[DualTriode]

Hello All,

An update about tools. the more recent versions of AP500 software have a added function called Transfer Function.

I will be using the white noise Transfer Function to test and report headphone measurements on a flat plane test fixture. 45CA plots will also be posted.

Yesterday I purchased a nice pair of used professional GRAS 46AO pressure microphones to use with the AP white noise Transfer Function with the soon to be introduced flat plane test fixture. No fake ears need apply.

Thanks DT

See attached GRAS 46AO Pressure calibration curve PDF

 

[pozz]

Seems like good progress so far.

 

[DualTriode]

Mulling stuff over for the flat plane headphone measurement fixture.

The GRAS 46AO Pressure Microphones I have coming are 12.7 mm in diameter, or ½” at the preamplifier section and 13.2 mm in diameter at the microphone capsule. 13.2mm / 25.4mm := 0.5196” diameter

I am thinking of two layers of ¾” plywood glued together for about a real world thickness of 1.41 inches thick. On the bench is a 17/32” drill bit or 0.5135 inches in diameter. The 46AO microphone with a fat end of 13.2 mm will just slide through the 17/32” dia. drilled hole.

I am thinking that the microphone and a layer or two blue 3M painter’s tape will be just snug in the drilled hole in the layered wood block.

Just thinking, I will let you know.

I am also thinking about a 1mm thick sheet of closed cell neoprene, a fake skin over the wood block. The 45CA has a hard smooth surface. So perhaps the fake skin should not be included. Still mulling.

Thanks DT

 

[DualTriode]

Hello All,

This time I brought along The AP Flex V6.0 software disk and dongle.

No problem doing the software installation.

Running the Flex Software with the Lynx E22 soundcard in XLR loopback gave promising test results.

The Lynx E22 sound card gives up a lot to the AP APX555 analyzer. This is an effort to switch off the APx555 with its cooling fan.

The self distortion and self noise of the GRAS CA45 with couplers and amplifiers will still allow a distortion and noise margin to the LYNX E22 soundcard.

The large advantage for the LYNX E22 Soundcard over the APx555 is no APx555 cooling fan.

The Lynx E22 SINAD is 96.7dB And THD+N is 0.0014%. Also see the attached FFT.

Thanks DT

Later, include the Topping L30 headphone amplifier, a headphone and the 45CA to the loop.

 

[DualTriode]

Hello All,

Just musing about stuff.

I should be measuring, quantifying and comparing the noise performance of the 3 analyzers: APx555, Lynx E22 with APx500 Flex V6.0 software and the RTX6001 with APx500 Flex V6.0 software. Also I will look at the noise and distortion of the headphone amplifiers. Since the test loop includes microphones the Audio Precision cooling fans should also be included in the noise measurements. I am not getting fancy with this setup. I will just set it all up on the 72” by 36” bench top and run it there.

The microphones will be the standard microphones in the GRAS 45CA-9 test fixture and the GRAS 46AG microphones coming to be installed in the now being assembled Flat Plane Test Fixture (FPTF).

Both pairs of microphones will use the GRAS 12AA power supply and amplifier. Unfortunately I am waiting for cables to go from the single end BNC output of the Grass 12AA amplifier to the female XLR inputs of the Lynx E22 and RTX6001 analyzers.

Meanwhile see the attached Loopback plot of the APx555 analyzer. The RTX6001 and Lynx E22 do not perform nearly as well as the APx555. I will measure and document the self noise and distortion of the GRAS microphones and 12AA electronics.

Thanks DT

 

[DualTriode]

Hello,

The GRAS 46AO CCP powered microphones. The only CCP/amplifier power nearby is the APx1701 on the bench which requires the APx555 or sim to operate. The microphones are out in in the hall covered with a big fluffy pillow. Only the CCP/Amplifiers are turned on, no power amplifiers.

Take a look at the noise floor of this setup. The loopback noise of the APx555 alone is ~ -160dB's

Thanks DT

 

[restorer-john]

The Lynx E22 SINAD is 96.7dB And THD+N is 0.0014%. Also see the attached FFT.

Those numbers are a bit disappointing. I was expecting considerably better for that Lynx E22 card. You have the fundamental around 16dB down from 0dBFS- was that a sweet spot input/output wise?

What sample rate and bit depth did you/the software set the card for the loopback test?

 

[DualTriode]

Those numbers are a bit disappointing. I was expecting considerably better for that Lynx E22 card. You have the fundamental around 16dB down from 0dBFS- was that a sweet spot input/output wise?

Yes that was sort of a sweet spot for the XLR to XLR loopback. Above that spot the distortion climbed steeply.

I will take another go at it when the Geshelli Archel2.5Pro amplifier with XLR inputs arrives.

Thanks DT