Introducing: Antdroid's Preference Target Curves [UPDATED]


Update: This was updated to reflect a new calibration file that more accurate represents the lower treble region of the curve for headphones -- thanks to Crinacle!

A couple years ago, I hand-sketched a target curve in Microsoft Paint and called it my Antdroid Target. It was partially a joke, but it actually was semi-based on what I thought what I considered neutral. This ended up being very similar to the Moondrop Blessing 2 when it came out, though that IEM is probably a touch brighter.



Since then, I wanted to come up with what I thought would be an ideal preference curve for in-ear monitors and later headphones. Fast forward a bit, and this year, I upgrade my IEC-60711 coupler to a new version of the knock-off GRAS unit, and added an ear and cheek simulator to the coupler microphone, and am now able to make headphones and IEM measurements that are pretty comparable to industry research.

**Please note that this isn't taken with an official GRAS RA0045 + ear/cheek sim but I've compared my headphone measurements to others and it's pretty close, and maybe 1-2dB off in treble. (usually a little less, but could be effects of pad wear and silent revisions -- Hifiman!!)





I've measured hundreds of IEMs at this point and a few dozen headphones and had a pretty good idea of what my preference targets were, but I wanted something I could use to compare to other headphones and to just have a target equalizer setting to work to. So I came up with an approach to come up with the Antdroid Target Curves.

It was quite simple actually. I took my favorite IEMs and Headphones I owned, measured them carefully and then took an average of each population and then made tweaks as necessary.

Easy! Let's see how it turned out!

Headphones

I came in thinking that my preference target was pretty much a flat bass and midrange to about 1KHz and then a rise similar to the Harman/Diffuse Field target with a little reduced gain and then a smoothed out treble response. The results were pretty close!

I used a combination of headphones that measured pretty differently. In this case I used my flagship and warm ZMF Verite with BE2 Leather pads, the brighter Hifiman Arya, the punchy Denon D5200, the portable Audeze Sine, and the legendary reference headphone: the Sennheiser HD600.




Once I aligned all these varying measurements and took an overall average, it came up with a nice looking curve that was basically what I was expecting. The average graph below is smoothed out a bit to make it more presentable, but it falls in line with exactly what I thought it should be. It is plotted below with the current 2019 over-ears target from Harman International's headphones research.



In-Ear Monitors

For IEMs, my own set of IEMs are actually very similar to each other for the most part. That is until I got the Unique Melody MEST, which is a little warmer and laid back then the other ones that are more or less some sort of modern Etymotic-curve based IEM with bass-shelf. 


For all of these measurements, I used custom versions of the IEMs because that's what I own and prefer. So the basic target generated was really meant for CIEMs. That said, for the most part, these IEMs measured the same for the universal as the custom model up until about 6KHz or so in the treble region, so much of the target can be used for universals as well. On top of that, I smoothed out and reduced the upper treble to factor in the resonance peaks and universal fit so that this could work with either customs or universals more effectively.

Like the headphone target, my predictions for the IEM target was very spot-on to what I suspected it would be and I am pretty happy with how it turned out. The results are below with the 2018 In-Ear Target from Harman International in gray.



Further Discussion

Now that I have both the headphones and in-ear targets for my own personal preferences, I decided to compare them head-to-head and with the Harman International research preference curves. The results are pretty interesting.

First, I figured that IEMs had a larger bass shelf gain than headphones, but it is also interesting to see that the Harman research also measured similar gains in the bass shelf for their IEM curves versus the headphone curve. Both Harman targets had more bass gain than my personal liking, but the differences of their OE and IE targets trend similarly.




So, the question that comes to mind is why do I, and most other people (it seems), need more bass gain in an in-ear monitor than a larger over-the-ear headphone? Is it because of the driver size and physics? Is it the ability to feel the vibration of the low end rumble of bass/sub-bass?

I feel it has to do with the "feels," but this is an area I'd like to learn a little bit more about later.

Until then, here's the target curves for your own use if you'd like. They are raw text files which can be imported into Excel, Room EQ Wizard, ARTA, or your favorite data manipulation software tool.

Downloads

Antdroid's Headphone Target Curve v1.1 [UPDATED]


Antdroid's IEM Target Curve v1.2


Basic Usage Instructions

In Room EQ Wizard, you can import this into its own MDAT file by opening REW, and then clicking File -> Import -> Frequency Response  and then locate and choose the target file.



If you want to use it as a Target Curve for using REQ's EQ Generation Tool, click Preferences -> Preferences and choose the House Curve tab


In the Curve Data input, click Clear Curve, and then Browse and choose the txt file.

Then click the EQ Button on the main screen, and it'll automatically load the Antdroid target as the default house curve for EQing the current measurement data to. You can then click Target Settings, and click Calculate Target Level from Response to automatically align the House Curve to the measurement data.


Make your filters and enjoy!

Comments

  1. Hey, I actually think your sketched curve gives a more neutral representation than your final target, I was fooling around with my own target to bring it as close to a neutral speaker in front of me and my own curve looks more similar to your sketch, although I think 4k needs a bit more boost, also bass and upper treble is very "vague". I noticed that different IEMs equalized towards exactly the same target still will differ a lot with bass and treble.

    ReplyDelete
    Replies
    1. I have modified both IEM and headphone curves very slightly since I wrote this. My headphone curve no longer has a bass bump, and is a flat line. The IEM curve was smoothed out a little bit, but generally looks the same. I have contemplated adding a little more 4K actually as I do find eq'ing to the target can create a little bit of a boring sounding IEM. I end up losing a little bit of a feeling of dynamics and excitement. But that's not a bad thing for a neutral target.

      https://iems.audiodiscourse.com/?share=Antdroid_Target
      https://headphones.audiodiscourse.com/?share=Antdroid_Target

      Delete

Post a Comment