Knowledge Zenith (KZ) impressions thread
Oct 6, 2017 at 10:36 AM Post #22,906 of 63,980
They might not be for everyone, but I really do think a set of foams puts the ZS6 right where it needs to be. I'm faaaaaar from a treble head.
Good news. :call_me:
 
Oct 6, 2017 at 10:40 AM Post #22,907 of 63,980
By most accounts the ZS6 tends towards a somewhat V shaped tuning? So definitely not reference.


As for the deep cut, that's right around the sibilance zone, but at 4kHz you're already past most instrument fundamentals, I believe?

Here's a neat chart that tries to describe what different frequencies do to various instruments :wink:
http://www.independentrecording.net/irn/resources/freqchart/main_display.htm
According to the link you gave, by 4khz you have passed all the fundamentals of almost all instruments but what about harmonics. You haven't crossed the harmonics. Why would one tune a earphone to not show some of the harmonics of instruments?
 
Oct 6, 2017 at 10:43 AM Post #22,908 of 63,980
According to the link you gave, by 4khz you have passed all the fundamentals of almost all instruments but what about harmonics. You haven't crossed the harmonics. Why would one tune a earphone to not show some of the harmonics of instruments?
You can check out other graphs, they all look squiggly like that. :call_me:
 
Oct 6, 2017 at 11:06 AM Post #22,910 of 63,980
I did. Not all of them have such huge dips in higher frequencies. They do have small peaks or small dips but the zs5/ zs6 has a huge one
Compensation schemes and scale have a lot of effect there. The salient point in a comparison graph is how they differ - which is the sort of useful information you can more universally draw from a FR plot since you don't need to carefully consider the measurement setup (which hakuzen described btw, uncompensated, and you can look up the FR variances on his coupler since it's an IEC standards compliant one).

The graphs being posted here as you can see have about a 50 dB spread on the Y axis - this lets you see what's going on in better detail.

As an extreme example, a lot of graphs manufacturers like to post have a 140+ dB spread so everything looks nice and smooth.
 
Last edited:
Oct 6, 2017 at 11:37 AM Post #22,911 of 63,980
No, but it is true that overwhelmingly most of Chi-Fi IEMs have quite bad response graphs. I mean, look at this wiggly stuff:
http://reference-audio-analyzer.pro/en/report/hp/kz-zst.php
http://reference-audio-analyzer.pro/en/report/hp/kinera-bd005.php (this one is actually one of the least messy graphs I've seen, along with Urbanfun hybrids which can be found on audiobudget.com, along with graphs for a lot of other IEMs. Though audiobudget's graphs are made with much worse equipment, probably no artificial ear canal that has the same acoustic parameters as human flesh)
vs.
http://reference-audio-analyzer.pro/en/report/hp/ultimate-ears-fi-5-pro.php
http://reference-audio-analyzer.pro/en/report/hp/etymotic-er4-2x.php
http://reference-audio-analyzer.pro/en/report/hp/shure-se-535.php

Of course, of course, the price difference is gargantuan, but nonetheless it's obvious that big names actually fine tune their stuff, while the Chinese just stick random drivers into random casings.
 
Oct 6, 2017 at 11:39 AM Post #22,912 of 63,980
Compensation schemes and scale have a lot of effect there. The salient point in a comparison graph is how they differ - which is the sort of useful information you can more universally draw from a FR plot since you don't need to carefully consider the measurement setup (which hakuzen described btw, uncompensated, and you can look up the FR variances on his coupler since it's an IEC standards compliant one).

The graphs being posted here as you can see have about a 50 dB spread on the Y axis - this lets you see what's going on in better detail.

As an extreme example, a lot of graphs manufacturers like to post have a 140+ dB spread so everything looks nice and smooth.
Great exposition
 
Oct 6, 2017 at 11:46 AM Post #22,913 of 63,980
Oct 6, 2017 at 12:01 PM Post #22,915 of 63,980
Just curious to know... Is there peeled off coating those bright corners on the earphones... You made it_.. Or just came in this condition from plant.

that is the factory finish after about a weeks worth of use. They have a few little scars from being tucked in a shirt pocket.
 
Oct 6, 2017 at 12:06 PM Post #22,917 of 63,980
Ah, Midnite. Good stuff. Seen him a few times!

Here's what I do. Should be pretty good on your earholes for testing your ZS6's.

Full album here:


Damn that's good.
 
Oct 6, 2017 at 12:13 PM Post #22,918 of 63,980
that is the factory finish after about a weeks worth of use. They have a few little scars from being tucked in a shirt pocket.

Seems some coating durability problem? Or very hard conditions you tested them.
 
Oct 6, 2017 at 12:17 PM Post #22,919 of 63,980
No, but it is true that overwhelmingly most of Chi-Fi IEMs have quite bad response graphs. I mean, look at this wiggly stuff:
http://reference-audio-analyzer.pro/en/report/hp/kz-zst.php
http://reference-audio-analyzer.pro/en/report/hp/kinera-bd005.php (this one is actually one of the least messy graphs I've seen, along with Urbanfun hybrids which can be found on audiobudget.com, along with graphs for a lot of other IEMs. Though audiobudget's graphs are made with much worse equipment, probably no artificial ear canal that has the same acoustic parameters as human flesh)
vs.
http://reference-audio-analyzer.pro/en/report/hp/ultimate-ears-fi-5-pro.php
http://reference-audio-analyzer.pro/en/report/hp/etymotic-er4-2x.php
http://reference-audio-analyzer.pro/en/report/hp/shure-se-535.php

Of course, of course, the price difference is gargantuan, but nonetheless it's obvious that big names actually fine tune their stuff, while the Chinese just stick random drivers into random casings.
You're comparing graphs from different sources and different compensation schemes...

And you're comparing earphones with very different tuning orientations... :shrug:


No one is trying to say ZSTs are reference tuned (because they're not...)

You could just as easily pick through that data and find "squiggly" graphs for "big names" too... so what? People aren't allowed to use anything but reference tuning?
 

Users who are viewing this thread

Back
Top