TRUTHEAR x Crinacle Zero IEM Review

[whazzup]

I wrote an article about the Harman Target Curve in Acoustics Today:

The Perception and Measurement of Headphone Sound Quality- What Do Listeners Prefer?​

The Perception and Measurement of Headphone Sound Quality- What Do Listeners Prefer? - Sean E. Olive

acousticstoday.org

I doesn't really deal with the IEM Target but much of the same rationale and testing methodology was used for both.

Please please please enable comments in your page. It will be so entertaining to watch the commentary shit storm unfold
And thanks for the work!!

 

[GaryH]

I wonder how long it will last. I'm still sticking with Apple earpod 3.5m since I bought iPhone SE. I have owned two IEMs that cannot out last Apple earpod.

You must not like bass.

 

[MayaTlab]

You must not like bass.

The EarPods / AirPods 1 & 2 response will vary a lot depending on fit between individuals, and other 711 results show quite a bit of a difference. For some individuals it's possible that they may output more bass than the AirPods 3 / Pro (both of which have a feedback mechanism to limit these variations) if the front volume is somewhat sealed because they have a small concha / narrow ear canal opening.

 

[charleski]

Any recommendations for a Bluetooth adapter for IEM's? Fiio? TRN?

If you’re looking (like me) for an ear hook adapter to turn these into TWS the choices are:
TRN - case hasn’t got enough space to hold the iems while charging, and uses a proprietary usb-c connector. But it’s cheap.
KZ AZ09 - seems workable, but unfortunately the pro model with a better Qualcomm chip is only sold with a shrouded 2-pin connector which won’t fit these.
Shure - MMCX only, so won’t fit. Also, expensive.
Fiio - seems the best bet, but you might want to wait for the utws5 which is coming soon and features an AK4332 DAC/amp rather than just using the one in the Qualcomm chip. Priced around £100.

 

[jae]

I wonder how long it will last. I'm still sticking with Apple earpod 3.5m since I bought iPhone SE. I have owned two IEMs that cannot out last Apple earpod.

Interesting to hear this. It's been known for a while that (at least the wired) apple earbuds were designed with materials specifically designed to fail after regular use after a certain period of time (don't remember the specifics, but it was in the 1 or 2 year range). Not sure if they are still using those practices or that philosophy. In the earing aid space, the industry is moving away from those small watch-style batteries and moving to rechargeable lithium cells for convenience, but the cells chosen are meant to lose significant holding capacity after ~3-4 years, which is coincidentally the exact period of time most insurance companies allow between coverage of hearing aids and getting a new pair. I don't doubt that manufacturers of new wireless buds, including apple, will probably be factoring something like into future products to encourage upgrading.

 

[MayaTlab]

Another point in the discussions about EQ: is the assumption that two IEMs EQ’d to show the exact same FR on a given test fixture must ‘sound’ the same on your own ears correct?

The question would be what does "the same" mean, ie what's the "different difference" and how does it compare with minimum thresholds of audible difference.
But IMO : that assumption is probably erroneous, just like it already demonstrably is for over-ears, albeit depending a lot on the type of over-ears, ie most large fully open dynamic headphones probably translate a lot better between ear simulators and individuals than most passive closed back or possibly some ANC headphones above the feedback range (maybe the same applies to different types of designs for IEMs ?).

I thought the correlation between two test fixtures, or a test fixture and your own ears, was… non linear at best… or worse case, not even a (mathematical) function. Two IEMs EQ’d to show perfectly identical FR on a given test fixture may show (minor?) FR differences if the insertion depth varies, others FR differences when measured on a different test fixture model, and unknown FR differences on your own ears.

We have very little info for in-ears when it comes to in-situ response.

A while back Rtings compared the 5128 to their HMS, this shows the difference between the two fixtures for IEMs only, for both R and L channels :

Like with the over-ears you tend to get an identifiable trend with some noise around it and some outliers.
This is just me but the "different difference" shown here would definitely go into the "audible" category.

You then have to add the issue of properly measuring active headphones in the first place, which is a whole another pitfall riddled with issues such as :
- the ratio between the feedback range and the passive range above it, which will vary - here it's crucial that the fixture is an adequate representation of the average human ear, and this may also add additional variance between individuals in addition to the upper treble resonances variation
- making sure that you've fooled the wear sensors to activate the proper mode (the ones relying on capacitance can be particularly tricky)
- making sure that you understand very well how the feedback mechanisms work, some of them are not real-time and won't give proper results with sweeps unless you go through the proper procedure (ex : AirPods 3 and AirPods Pro and Max when ANC is off).
- making sure that you understand how some of their individualisation features work - ex the newer Bose QuietComfort Earbuds II.

Will two IEMs with identical FR measured on the same test fixture sound the same? Assuming their distortion is below threshold. they seal properly and are inserted in your ears at the same depth -- they should sound pretty darn close.

If I may, the idea that two IEMs can be inserted at the same depth is a bit academical given that individuals will be forced to position most IEMs in a way that is in accordance to their own anatomy. I just physically cannot insert a pair of AirPods Pro quite as deep as a pair of Etymotic for example.

It's probably a more practical consideration to have a good database on human anatomy and analyse how the IEM being developed interacts with various individuals to statistically determine the range where this particular design falls in terms of resonances, and tune it in accordance to these findings.

And in extenso probably more realistic as well to measure IEMs on ear simulators with a realistic pinnae / concha / ear canal entrance rather than just a coupler and try to match the resonance.

We published a paper where we equalized several AE headphones to the same frequency response leaving only distortion differences as the primary cue -- and the headphones with low distortion sounded remarkably similar.

Is that a reference to "The Correlation between Distortion Audibility and Listener Preference in Headphones", 2014 ?
Since that article deliberately and for good reasons (since FR had to be a controlled variable) eschewed the question of how different headphones translate more or less differently between ear simulators and individuals, by playing recordings of the headphones on a virtual pair of headphones, it stands to reason that the headphones with low distortion sounded similar.

The way I understand it, @CedarX question was primarily interested in the issues of ear simulator to individuals translation, something I believe Harman is currently evaluating for over-ears and with blocked ear canal entrance mics.

 

[jae]

The question would be what does "the same" mean, ie what's the "different difference" and how does it compare with minimum thresholds of audible difference.
But IMO : that assumption is probably erroneous, just like it already demonstrably is for over-ears, albeit depending a lot on the type of over-ears, ie most large fully open dynamic headphones probably translate a lot better between ear simulators and individuals than most passive closed back or possibly some ANC headphones above the feedback range (maybe the same applies to different types of designs for IEMs ?).



We have very little info for in-ears when it comes to in-situ response.

A while back Rtings compared the 5128 to their HMS, this shows the difference between the two fixtures for IEMs only, for both R and L channels :
View attachment 231071
Like with the over-ears you tend to get an identifiable trend with some noise around it and some outliers.
This is just me but the "different difference" showed here would definitely go into the "audible" category.

You then have to add the issue of properly measuring active headphones in the first place, which is a whole another pitfall riddled with issues such as :
- the ratio between the feedback range and the passive range above it, which will vary - here it's crucial that the fixture is an adequate representation of the average human ear, and this may also add additional variance between individuals in addition to the upper treble resonances variation
- making sure that you've fooled the wear sensors to activate the proper mode (the ones relying on capacitance can be particularly tricky)
- making sure that you understand very well how the feedback mechanisms work, some of them are not real-time and won't give proper results with sweeps unless you go through the proper procedure (ex : AirPods 3 and AirPods Pro and Max when ANC is off).



If I may, the idea that two IEMs can be inserted at the same depth is a bit academical given that individuals will be forced to position most IEMs in a way that is in accordance to their own anatomy. I just physically cannot insert a pair of AirPods Pro quite as deep as a pair of Etymotic for example.

It's probably a more practical consideration to have a good database on human anatomy and analyse how the IEM being developed interacts with various individuals to statistically determine the range where this particular design falls in terms of resonances, and tune it in accordance to these findings.

And in extenso probably more realistic as well to measure IEMs on ear simulators with a realistic pinnae / concha / ear canal entrance rather than just a coupler and try to match the resonance.



Is that a reference to "The Correlation between Distortion Audibility and Listener Preference in Headphones", 2014 ?
Since that article deliberately and for good reasons (since FR had to be a controlled variable) eschewed the question of how different headphones translate more or less differently between ear simulators and individuals, by playing recordings of the headphones on a virtual pair of headphones, it stands to reason that the headphones with low distortion sounded similar.

The way I understand it, @CedarX question was primarily interested in the issues of ear simulator to individuals translation, something I believe Harman is currently evaluating for over-ears and with blocked ear canal entrance mics.

There is some data for Etymotics, and I suspect the response for many shallower models with more leakage could perhaps be even more variable. That is also one partial explanation of the wider occluded bass preference skewing results.

 

[whazzup]

And now we have the ability to compare measurement rigs! Not versed in the finer details of the different types of dB scales, just curious if anyone can help me understand the differences between Amir's and Crinacle's measurements in the above image (Crins' measurements in faded blue/red). I've tried to adjust the images so that their axes match. The 'zero line' is aligned at 95dB of both images.

*Used Crin's 'sample 10' image for the comparison

Truthear x crinacle Zero – In-Ear Fidelity

Sample 3 Sample 4 Sample 5 Sample 6 Sample 7 Sample 8 Sample 9 Sample 10 Sample 1 (Pre-production)

crinacle.com

 

[PeteL]

I have no final opinion on this question. But he did definitely not answer my exact question, which stated the condition that two different IEMs could both be EQed perfectly to a target curve and he said something about resonances that might prevent that. But nevertheless, it seems that for headphones, as compared to amps, DACs and even speakers, the percentage what can be measured to determine the sound is the lowest. There was the case of the Mark Levison which closely followed the Harman curve but still did not sound all that great. For IEMs this is probably a bit easier as they anyway do not have a huge soundstage and so detailed imaging. But I tend more and more to believe that most, if not all, terms that are usually called technicalities, like speed, detail, air, et. are either non-existent or a consequence of the FR.

Well, air is a term to describe frequency response, so on that one it's quite agreed upon, speed may be a flawed term but it's still true that two headphone with the same frequency response (to a degree of tolerance or readability) can have different impulse response, how it translate subjectively is where there are grey zone. Detail is quite simple, it means that you hear more low level far in the mix elements on headphone A than headphone B. It is true that some headphones will let you hear things that you will miss with others like reverb tails, or even mumbling far away from a microphones or all sort of things that are on the recording but not audible on some, audible on others. Does it strictly means more highs, maybe, maybe not.

 

[MayaTlab]

just curious if anyone can help me understand the differences between Amir's and Crinacle's measurements in the above image (Crins' measurements in faded blue/red). I've tried to adjust the images so that their axes match. The 'zero line' is aligned at 95dB of both images.

The blue/red lines for Crin's measurements are compensated curves, Amir's curves are the raw curves. The raw curves from Crin are the light grey ones below the blue/red ones.

 

[MayaTlab]

There is some data for Etymotics, and I suspect the response for many shallower models with more leakage could perhaps be even more variable. That is also one partial explanation of the wider occluded bass preference skewing results.

Is that https://acta-acustica.edpsciences.org/articles/aacus/full_html/2021/01/aacus210058/aacus210058.html ?

I notice that the measurements were performed with probe tubes, for IEMs this can have a significant impact on the results as it may compromise seal. Also, re-positioning the IEM may slightly shift the tube fore and aft in the ear canal which may introduce variation at higher frequencies.

One thing I like a lot about Harman's research with IEMs is that they were quite careful to design their test to minimise the impact of installing a small MEMS mic in the bore of their replicator IEMs (with the caveat that the measurements probably aren't useful above somewhere around 500-1kHz)

 

[whazzup]

The blue/red lines for Crin's measurements are compensated curves, Amir's curves are the raw curves. The raw curves from Crin are the light grey ones below the blue/red ones.

Aha! Thanks, I understand now. Reposting image. Very close and consistent graphs now (aligned by eye to 200Hz, Y-axis units now a bit off).

 

[DavidEdwinAston]

Can I ask please, what should I look for in my Quad Artera peamp spec. to find the sensitivity requirements for these iem,s?
It does say, " headphone out, 6.35 mm". And nothing else, that I can see.

 

[julian_hughes]

Well, air is a term to describe frequency response, so on that one it's quite agreed upon, speed may be a flawed term but it's still true that two headphone with the same frequency response (to a degree of tolerance or readability) can have different impulse response, how it translate subjectively is where there are grey zone. Detail is quite simple, it means that you hear more low level far in the mix elements on headphone A than headphone B. It is true that some headphones will let you hear things that you will miss with others like reverb tails, or even mumbling far away from a microphones or all sort of things that are on the recording but not audible on some, audible on others. Does it strictly means more highs, maybe, maybe not.

Also worth considering: if you use both dynamic and BA driver IEMs then you'll know that they have their own qualities which are not shown in a FR graph. It would be interesting to see a BA and a dynamic compared, eq'd to supposedly be the same, and people asked to blindly identify which is which. Has bass made by a BA ever sounded exactly like bass from a loudspeaker or dynamic IEM? And do dynamic IEMs ever achieve that incredible sense of precision and clarity of something like acoustic guitar heard on BA IEMs?

 

[MZKM]

Can I ask please, what should I look for in my Quad Artera peamp spec. to find the sensitivity requirements for these iem,s?
It does say, " headphone out, 6.35 mm". And nothing else, that I can see.

No such thing as sensitivity requirements.

For IEMs that need power, you look at power output (voltage or wattage), and hopefully it shows at difference impedances.

For IEMs that don't need a lot of power, you want a very low noise floor (high SNR) at lower power (SNR at 1W is useless for these IEMs, Amir tests at 5mW for instance).

 

[DanielT]

How well an IEM couples to your ear makes a massive difference in bass. If you put that aside, two IEMs with similar frequency response will statistically be equally good or bad. You can rule them out, or in that way. Without the measurement, it would just be a game of chance.

That sounds reasonable. A speaker with a badly installed bass driver, where it leaks, does not produce good bass. Even fairly horrendous bad weak bass. If you can now make that kind of comparison, IEM - speakers that is.

The TRUTHEAR x Crinacle Zero in themselves seems good, but the big challenge, as it seems in this thread, is to find the right kind of tips.

It's been a long time since I had any IEM. The only thing I remember is that they had a tendency to pop out of the ears, but I didn't care at the time.
I just plopped in them, again if they fell out no major annoyance (it could have been annoying if it happened all the time of course, but it didn't).Didn't listen to anything but talk radio with them then.

With The TRUTHEAR x Crinacle Zero on the other hand I mean "HiFi" listening to music. I'll see if I like it/them. Exciting.

 

[DanielT]

I mean a dedicated external headphone amp like those reviewed on here.

I have a generic smart phone, one plus 8.

Aha, answer here, below from Jimbob54.

Plus if you use Android this might be worth considering:

High quality media player for Android

Has this been tested? Is what is said true? Anyone have this player? Experiences? HIGH QUALITY MEDIA PLAYER FOR ANDROID Supporting USB audio DACs to play audio files up to 32-bit resolution at any sample rate! Will play any popular format including wav, flac, wavpack, ape, mp3 and DSD files...

audiosciencereview.com

The OP8 doesnt have a headphone jack, does it?

But if you had a decent dongle, yes that would be fine.

 

[thin bLue]

The Point

This product has relatively high fidelity of Otaku factor(sighted)

FR... Check
THD... Check
Waifu material... Check

 

[Tovarich007]

The comment was regarding tonality, not total perception.

Better to precise it, indeed ! But anyway your initial statement was a source of confusion.
When we listen to music (or to any sounds in general) we don't artificially distinguish "tonality" and "total perception".
Every sound parameter is in relation with each other, the whole being integrated by our brain to form a global perception.

As for me, it's impossible to assume that a headphone's sound is "like the best speaker in best room with no room modes", just because there is no room at all, obviously ! Consequently, the global perception, including tonality, can't be felt the same for a headphone, as good as it is, and a speaker, as good as it is too..

 

[DanielT]

The Point

This product has relatively high fidelity of Otaku factor(sighted)

View attachment 231085View attachment 231084
FR... Check
THD... Check
Waifu material... Check

Waifu..Otaku factor..?

I feel like a great-grandfather, completely off.

Good thing you can google. So I learned something new today as well.