[gallappago]

Someone should really take the plunge on these then cause they graph like the teas too it looks like. These are probably really good. I wish someone would compare these to the teas.


Yeah this further makes me want to get an iec711, definitely my next purchase before I get anymore iems. I mostly want it to figure out what it is I don't like so much in other iems and what I do like but it's gonna be interesting seeing what else these other types of measurements can reveal.

I would love to see how well the gt12x and st8b graph on a waterfall, going to see if I can order both in the near future

I can probably get the st8b(and teas since they should be arriving tomorrow ) done for you around this time tomorrow once I get back in from work

 

[sutosuto]

when it has so much weirdness going on in its coherency

Hi, I am interested in learning what coherency is.
After reading your posting I took my GK10 and compared to my daily driver, ISN EST50 and I still have no idea what the coherency is. LOL.
Can you please show me any specific song and part of the song which I can hear the coherency of the iem?

Thanks

 

[mndless]

Hi, I am interested in learning what coherency is.
After reading your posting I took my GK10 and compared to my daily driver, ISN EST50 and I still have no idea what the coherency is. LOL.
Can you please show me any specific song and part of the song which I can hear the coherency of the iem?

Thanks

Coherency is a complicated topic because there are several aspects described within the overarching umbrella: time, phase, attack and decay.

Time coherency is based on the difference in distance from the listener's ears to the various drivers in a speaker array. This is significantly minimized in IEMs because we're dealing with millimeters instead of feet, so time coherence is pretty much a non-issue for IEMs as long as there aren't massive disparities between the sound tube length between drivers or driver placement within an open resonance shell.

Phase coherence is related to time coherency in loudspeakers, but for IEMs it basically means that all of the individual drivers should have the same phase as long as their attack speed is close enough

Attack and decay are intricately related as they are determined largely by the mass of the driven diaphragm, the strength of the magnetic spring force (determined by the size and number of windings of the voice coil and the strength of the permanent magnet), and the strength of the mechanical spring force, which is determined mostly by the surround material of the driver which joins the diaphragm to the surface. Use a very stiff membrane material with something that has a light membrane and strong magnetic spring force and you get a driver that has high attack speed and quick decay. Use something with a very soft membrane material instead, but the same configuration for everything else and you get something with high attack speed but a slower decay because there's less force pulling the driven diaphragm back to its nominal position. There are various other configurations as well, but you'll usually see similar magnetic spring force in IEM drivers to increase efficiency, and the primary difference will be in diaphragm mass and surround stiffness. Use a heavier diaphragm with a softer surround and you get a driver suited to being a subwoofer. Use something light with a stiff surround and you get a midrange, full range, or tweeter driver, often based on size.

the best way to hear the difference between a more an less coherent driver is to listen to the same music, preferably something relatively busy that has clear sounds in various frequency bands, on two different sets: a single driver set and a multi-driver or hybrid set. In IEMs, it is typically differences in the rate of decay that result in the perception that something is less coherent, since single-driver sets have a pretty predictable rate of change for decay, wherein low frequencies decay slower than higher frequencies. In multi-driver arrays, a midrange driver or tweeter may decay more slowly at the bottom of their usable frequency range compared to the rate of decay of the previous driver at the top of its usable range. This can result in the perception that the driver is less coherent. With the GK10 as an example:

You can see that the rate of decay is slower at 5kHz, 8kHz and 15/18kHz. This is likely related to the rate of decay for those particular drivers (the BA and piezo drivers) being slower at the beginning of their crossover point compared to the upper range of the graphene midrange driver. You can also see that the titanium composite driver decays extremely slowly compared to the other drivers, which is quite different from how a single DD set decays from this over-ear headphone's plot:

you can see that there's still some breakup above 4kHz, which is something you can largely avoid in IEMs because of the driver size being so much smaller in single DD IEMs, but look at how the rate of decay occurs from 100Hz all the way through the graph. It diminishes almost linearly (due to the frequency representation in the graph's axes, the actual rate of decay is more logarithmic, but it's more about the decay being a smooth transition until it levels out when the driver reaches a certain equilibrium. In this case, it's around 2kHz that this occurs, and the decay remains largely the same.

And here's one from the Toneking T5, a 5BA per side set, to demonstrate that multi-driver sets don't necessarily come with poorer coherency:

Also note that drivers tend to have slower decay from higher amplitude, so the decay at the peaks at 2.5, 4.5, and 8.5kHz are pretty much in line with everything else. Interestingly, the rate of decay of the BA drivers they're using seems to start off slowly then progress more rapidly after about 4.13 seconds, which is quite interesting.

Edit: in case anyone wants to do more searches for the tiny number of such graphs on the interwebs, they're officially called Cumulative Spectral Decay waterfall plots.

Other things to note: when you see weirdness on the graphs like what I described as driver breakup on the plot for the Sennheister HD800, this can be caused by things like the air in the measurement chamber acting as a spring and damping the travel of the driver while also slightly distorting the surface of the driver, amongst other causes. There are just so, so many factors involved in the way that speakers behave the way that they do.

 

[Lamim Rashid]

I can probably get the st8b(and teas since they should be arriving tomorrow ) done for you around this time tomorrow once I get back in from work

Thanks looking forward to this!

Coherency is a complicated topic because there are several aspects described within the overarching umbrella: time, phase, attack and decay.

Time coherency is based on the difference in distance from the listener's ears to the various drivers in a speaker array. This is significantly minimized in IEMs because we're dealing with millimeters instead of feet, so time coherence is pretty much a non-issue for IEMs as long as there aren't massive disparities between the sound tube length between drivers or driver placement within an open resonance shell.

Phase coherence is related to time coherency in loudspeakers, but for IEMs it basically means that all of the individual drivers should have the same phase as long as their attack speed is close enough

Attack and decay are intricately related as they are determined largely by the mass of the driven diaphragm, the strength of the magnetic spring force (determined by the size and number of windings of the voice coil and the strength of the permanent magnet), and the strength of the mechanical spring force, which is determined mostly by the surround material of the driver which joins the diaphragm to the surface. Use a very stiff membrane material with something that has a light membrane and strong magnetic spring force and you get a driver that has high attack speed and quick decay. Use something with a very soft membrane material instead, but the same configuration for everything else and you get something with high attack speed but a slower decay because there's less force pulling the driven diaphragm back to its nominal position. There are various other configurations as well, but you'll usually see similar magnetic spring force in IEM drivers to increase efficiency, and the primary difference will be in diaphragm mass and surround stiffness. Use a heavier diaphragm with a softer surround and you get a driver suited to being a subwoofer. Use something light with a stiff surround and you get a midrange, full range, or tweeter driver, often based on size.

the best way to hear the difference between a more an less coherent driver is to listen to the same music, preferably something relatively busy that has clear sounds in various frequency bands, on two different sets: a single driver set and a multi-driver or hybrid set. In IEMs, it is typically differences in the rate of decay that result in the perception that something is less coherent, since single-driver sets have a pretty predictable rate of change for decay, wherein low frequencies decay slower than higher frequencies. In multi-driver arrays, a midrange driver or tweeter may decay more slowly at the bottom of their usable frequency range compared to the rate of decay of the previous driver at the top of its usable range. This can result in the perception that the driver is less coherent. With the GK10 as an example:

You can see that the rate of decay is slower at 5kHz, 8kHz and 15/18kHz. This is likely related to the rate of decay for those particular drivers (the BA and piezo drivers) being slower at the beginning of their crossover point compared to the upper range of the graphene midrange driver. You can also see that the titanium composite driver decays extremely slowly compared to the other drivers, which is quite different from how a single DD set decays from this over-ear headphone's plot:

you can see that there's still some breakup above 4kHz, which is something you can largely avoid in IEMs because of the driver size being so much smaller in single DD IEMs, but look at how the rate of decay occurs from 100Hz all the way through the graph. It diminishes almost linearly (due to the frequency representation in the graph's axes, the actual rate of decay is more logarithmic, but it's more about the decay being a smooth transition until it levels out when the driver reaches a certain equilibrium. In this case, it's around 2kHz that this occurs, and the decay remains largely the same.

And here's one from the Toneking T5, a 5BA per side set, to demonstrate that multi-driver sets don't necessarily come with poorer coherency:

Also note that drivers tend to have slower decay from higher amplitude, so the decay at the peaks at 2.5, 4.5, and 8.5kHz are pretty much in line with everything else. Interestingly, the rate of decay of the BA drivers they're using seems to start off slowly then progress more rapidly after about 4.13 seconds, which is quite interesting.

I remember seeing some transient response graphs over in another forum comparing ba driver to DD driver. The ba driver has a faster and decay but the DD driver had a more even looking attack decay. I'll attach the link and graphs if I can find it.

On another note I'm looking for a cheap coupler to mess around with. Does anyone have know a good cheap iec 60318-4 spec coupler off Ali to play around with? I found this one https://m.aliexpress.com/item/1005001844592992.html?spm=a2g0n.shopcart-amp.item.1005001844592992& and was thinking of getting it. it's pretty cheap but I'm wondering if it's missing anything like the microphone cause I see a lot of other sellers selling the same one for more than twice the cost and I can't tell if it's missing anything from the product description page or review.

 

[sutosuto]

Coherency is a complicated topic because there are several aspects described within the overarching umbrella: time, phase, attack and decay.

No simple way to describe how you test the coherency with some of sample songs?

 

[mndless]

Thanks looking forward to this!

I remember seeing some transient response graphs over in another forum comparing ba driver to DD driver. The ba driver has a faster and decay but the DD driver had a more even looking attack decay. I'll attach the link and graphs if I can find it.

On another note I'm looking for a cheap coupler to mess around with. Does anyone have know a good cheap iec 60318-4 spec coupler off Ali to play around with? I found this one https://m.aliexpress.com/item/1005001844592992.html?spm=a2g0n.shopcart-amp.item.1005001844592992& and was thinking of getting it. it's pretty cheap but I'm wondering if it's missing anything like the microphone cause I see a lot of other sellers selling the same one for more than twice the cost and I can't tell if it's missing anything from the product description page or review.

That link appears to go to some pneumatic coupler fittings, not an IEM frequency measurement coupler.

 

[mndless]

No simple way to describe how you test the coherency with some of sample songs?

The best way is to listen to some songs with several sounds that occur simultaneously on two or more IEMs, one that is a single dynamic driver and another that is a multi-driver array, preferably a hybrid. Because single-DD sets typically have better coherency compared to multi-driver arrays, you may be able to notice it more easily in that the sounds decay more naturally and, thus, sound more like how they would in real life. I'd recommend some songs with sounds from the real world that you might be familiar with. This album is basically Foley art transformed into music:

 

[Lamim Rashid]

That link appears to go to some pneumatic coupler fittings, not an IEM frequency measurement coupler.

Well that explains a lot

I remember seeing some transient response graphs over in another forum comparing ba driver to DD driver. The ba driver has a faster and decay but the DD driver had a more even looking attack decay. I'll attach the link and graphs if I can find it.

Here are the graphs I was talking about, this was posted by purr1n here: https://www.superbestaudiofriends.o...nced-armature-drivers-suck.11485/#post-361554

 

[sutosuto]

The best way is to listen to some songs with several sounds that occur simultaneously on two or more IEMs, one that is a single dynamic driver and another that is a multi-driver array, preferably a hybrid. Because single-DD sets typically have better coherency compared to multi-driver arrays, you may be able to notice it more easily in that the sounds decay more naturally and, thus, sound more like how they would in real life. I'd recommend some songs with sounds from the real world that you might be familiar with. This album is basically Foley art transformed into music:

Thanks

 

[mndless]

Thanks looking forward to this!

I remember seeing some transient response graphs over in another forum comparing ba driver to DD driver. The ba driver has a faster and decay but the DD driver had a more even looking attack decay. I'll attach the link and graphs if I can find it.

On another note I'm looking for a cheap coupler to mess around with. Does anyone have know a good cheap iec 60318-4 spec coupler off Ali to play around with? I found this one https://m.aliexpress.com/item/1005001844592992.html?spm=a2g0n.shopcart-amp.item.1005001844592992& and was thinking of getting it. it's pretty cheap but I'm wondering if it's missing anything like the microphone cause I see a lot of other sellers selling the same one for more than twice the cost and I can't tell if it's missing anything from the product description page or review.

Looks like the cheapest one I can find with a cursory dig through AE is $81.88 USD. It doesn't come with a flat pedestal mount, so you'd need to find something to strap it to or securely lean it against while testing.

https://www.aliexpress.com/item/400...6-23&pdp_ext_f={"sku_id":"10000002796398525"}

 

[Lamim Rashid]

Looks like the cheapest one I can find with a cursory dig through AE is $81.88 USD. It doesn't come with a flat pedestal mount, so you'd need to find something to strap it to or securely lean it against while testing.

https://www.aliexpress.com/item/4000544425172.html?spm=a2g0o.productlist.0.0.366250dfkQJczI&algo_pvid=db37ba82-614d-4572-a2c2-61c9f1303ce6&algo_exp_id=db37ba82-614d-4572-a2c2-61c9f1303ce6-23&pdp_ext_f={"sku_id":"10000002796398525"}

https://m.aliexpress.com/item/40009...er_id=de0811157ec74ef09dc95ee766ce6646&is_c=N

This is the cheapest one I could find that looked about the same as all the other ones going for $90~
$74 plus another 5 for shipping. Not bad.

From what I understand these are pretty accurate, it's just the calibration files they come with usually suck.

I just placed my order, can't wait to get it. Estimated date is a month a way.. oh well.

 

[sutosuto]

The best way is to listen to some songs with several sounds that occur simultaneously on two or more IEMs, one that is a single dynamic driver and another that is a multi-driver array, preferably a hybrid. Because single-DD sets typically have better coherency compared to multi-driver arrays, you may be able to notice it more easily in that the sounds decay more naturally and, thus, sound more like how they would in real life.

LOL, I think I give it up. I tried to listen that bubbles song using my DD (ISN D10) and GK10. I heard the sound pretty similar from both
At least GK10 is not a bad iem for my ears and I can still enjoy it.

BTW, isn't the natural sound more refer to timbre instead?
No need to answer if this is not the right thread.

Thanks again for your previous response.

 

[mndless]

Well that explains a lot

Here are the graphs I was talking about, this was posted by purr1n here: https://www.superbestaudiofriends.o...nced-armature-drivers-suck.11485/#post-361554

That was definitely an interesting and enlightening read.

To casually speak about a topic brought up in that thread: for loudspeaker tweeters, I am a huge fan of soft dome tweeters for their very pleasant and natural sounding timbre. A good silk dome tweeter is very nice to listen to, and a dual ring radiator is even better because of its ridiculously impressive dispersion. Vifa made some XC25 series dual ring radiators that Scanspeak has adopted and some chinese companies are remanufacturing them as well. If you want to pick some up, MeloDavid Hifi on AliExpress has some nice ones. I actually helped my friend build some PVC tweeter pods for his car to install them alongside some rather lovely Silver Flute 6.5" 4ohm woofers. They're an excellent match, though I'd recommend running with the 8Ohm tweeters and 4Ohm woofers because of how crazy sensitive those tweeters are. As for metal dome tweeters, every example I've heard that wasn't tuned by KEF have been generally unpleasant, though I've not had much experience with really exotic dome materials like DLC, beryllium, or ceramic domes, so I can't really comment on them. I just know that titanium and aluminum domes have failed to earn my seal of approval in general (KEF being the sole exception thanks to whatever witchcraft and wizardry their driver engineers managed to pull off with their Uni-Q driver designs). And I loathe plastic and piezo tweeters in general, since their timbre sounds awful. I know there's a lot of people that love what LZ has managed to pull off with their IEMs that feature them, so I won't fully write them off in that application until I can hear them myself, but for loudspeakers, they're a hard pass from me. I have enjoyed some of the AMT tweeters in the past, so I would be interested to see what their burst measurements, harmonic distortion graphs, and decay graphs look like, but due to the impracticality of implementing them where I'd be looking to do so in the relatively near future, it's certainly not a pressing matter.

Because I brought it up, have some tweeter pod build pictures:

They started off as some PVC pipe end caps with some PVC pipe jammed inside with some of the chemical weld solvent spread on the mating surfaces. The excess pipe was cut off with a coping saw and the surfaces filed and then sanded smooth.

Several coats of dupont flat black automotive engine enamel were applied, then holes were pre-drilled for mounting screws for the speaker, the mounting fixture, and a large hole bored through to allow the wire to be fed through. We used some leftover 16 gauge multi-strand copper speaker wire

Spade terminals were used to terminate the connections and the speaker was installed. The fixture has pivot and angle adjustment and was harvested from a pair of very inexpensive black painted track lights.

Fiber reinforced bondo was used to reinforce the A-pillar trim before holes were drilled for the threaded end of the mounting fixture to pass through and the matching rear plate for the mounting fixture was epoxied down and then bondo-ed over

The final product installed into a 2014 VW Jetta. I think they look rather fantastic and the adjustment means that it is easily possible to turn them more or less on-axis depending on the need. Once they're arranged properly, though, additional adjustments aren't really that necessary. Because of how ridiculous the dispersion is on these things, you don't really lose much performance at all even 60 degrees off-axis. They're quite impressive and they sound wonderful. Exceedingly resolving.

 

[mndless]

LOL, I think I give it up. I tried to listen that bubbles song using my DD (ISN D10) and GK10. I heard the sound pretty similar from both
At least GK10 is not a bad iem for my ears and I can still enjoy it.

BTW, isn't the natural sound more refer to timbre instead?
No need to answer if this is not the right thread.

Thanks again for your previous response.

It's my thread anyway, so I don't really mind if it goes a bit off track if it helps the people watching the thread or provides some amusement. I'm glad the GK10 have worked out well for you. I just didn't like the set I received. Given how much QC variation we've seen in the community, it's entirely possible I got a dud. Regardless, what matters is whether or not you like the sound.

Sounds in nature will have the default timbre, because you are hearing the sound directly instead of hearing an imitation of the sound through speakers. If you have guitars or orchestral instruments available that you can hear them unadulturated, that's also a great way to determine timbral accuracy of your IEMs. Timbre is made up of things like attack speed and accuracy, decay speed and accuracy, harmonic distortion, etc. There's a lot that goes into what comprises a sound and what makes a sound "natural". Whether or not that matters is entirely up to the individual listener. Some people actually prefer a less natural timbre for certain genres of music, since it suits them well. Some people love the more metallic timbre of KZ's BAs for grunge, metal, certain types of rock music, etc., because it compliments the genre. Many of these genre already make use of a lot of synthesizers and distortion during production and mastering, so it makes sense not to try to "fix" that sound, but emphasize it if that's what you like.

 

[sutosuto]

Whether or not that matters is entirely up to the individual listener. Some people actually prefer a less natural timbre for certain genres of music, since it suits them well.

Good point, I concur.
I have piano and guitars.
I prefer natural sound. I hear unnatural from GK10, my BAs and hybrid sets actually. Few of thems annoy me a bit especially if I swap from my favorite set, ISN EST50, but some are okay.
Usually, I let my ears rest awhile before rotating to other sets.

Cheers