I’ve purchased a couple of iems recently which have resulted in a massive discrepancy in the low end between what I see in the measurements and what I hear. Basically some iems sound like someone put a subwoofer into my ear. Tip rolling doesn’t affect this.
For reference (pun intended) the Crinear Reference is a little below my preference for quantity of low end. The Truthear Pure is closer to my preference overall although I wouldn’t mind a little more low end for certain genres of music.
So given the praise I decided to get the Daybreak which in the graphs is not that far off from the Pure. I assumed I would get a little more deep low end and rumble. When I got them the low end was so exaggerated that made them simply unlistenable to me. I got a similar experience with the Thieaudio Hype 4 mk2.
Anyone has any idea what is going on?
BTW I’ve never experienced this with headphones. Typically what I see in the graphs aligns somewhat with what I hear.
It’s because the FR at your eardrum is not the same as the FR at the ‘eardrum’ of the measurement fixture. Remember the graph just gives you the picture of the sound in the condition of the device being on (or in) that head and ears, not yours.
Regarding the specific squigs, you’re seeing how acoustic impedance really matters when it comes to understanding how these devices perform. The 5128 (and 4620) will reveal more information about IEM performance, especially at low and high frequencies, than the 711 datasets, because it has the more accurate acoustic impedance.
But even still, there’s more to it than that. With IEMs, there are some very specific challenges to do with in-situ response variation, where the graph becomes less predictable than we’d like. This is why:
The frustrating aspect is it appears there’s no way for me to predict when this “subwoofer in my ear” effect will happen other than buying the IEM and trying it (no CanJam happening in Mexico).
Obviously the FR is not a good predictor but apparently neither are the drivers.
For the low end, the Hype 4 mk2 has 2x 8mm DD and the Reference has 2x 10mm DD. One triggers this and the other doesn’t. Same with the Pure and the Daybreak. Both have 1x 10mm DD. One triggers the effect and the other doesn’t. There’s probably some spec like travel excursion or something that explains this… but this is not something that anyone publishes.
We assume you are using an equivalent testing/auditioning approach, but its best one double checks.
Apologies if I am stating the obvious. I have had to become extremely disciplined in my own listening comparisons, between IEMs, to rule out anything which could be introducing variations, leaving the only changes, to a change in the IEM.
Same song list.
Same DAC/Dongle DAC/Headphone Amp, or whatever the chain one listens through, up till the output of the headphone amp
Same IEM cable. I found this important cos different cables have different resistance which changes the loudness, and as much as people think all cables sound the same, I beg respectfully to disagree - I easily hear differences in cables, and I am not alone in this. Whatever the case, taking out any variation in cables, helps to have a more equivalent comparison.
Same Eartips (at least the same product, if not the same physical eartip instance - to avoid time taken to move the eartip from one IEM to another)
And as one swaps between IEM to IEM, is the loudness level the same. Unless one is managing the gain individually, to set the loudness level about the same, so easy to be deceived we are hearing a difference. Each IEM has its own sensitivity and impedance which uniquely affects the loudness. In my playback/streaming software, I make an effort, to setup recallable volume settings for each IEM, so that as much as possible, one is hearing each one, at the same loudness This is done by gain presets which I can switch between, when switching between IEMs.
On another note, in many IEM measurements, the details like what cables were used or tips used, are not disclosed, and these (especially the eartips) are responsible for possible variances with what we hear, in the case we are listening via different eartips, from those used to measure.
Then there is insertion depth, which does change how an IEM sounds. The combination of IEM and tips, as well as our own habits, or mood, can have its own impact on how deep one inserts the IEM. Using the same tip does provide some level of equivalence, so any differences we hear can be attributed more to the IEMs.
Yes, I can’t really measure the response of IEMs in my ears.
I am disciplined and careful about testing headphones and IEMs. That said nothing in your list explains what I would estimate is like 5-7db of extra low end below 80hz. This is not a slight variation.
And Jay Audio on Youtube has done an outstanding job measuring this variance of many tips vs a
specific tip - the KBear 07.Just shows that if we are swapping between IEMs, and not listening on the same tips, we have two moving parts - the tips and the IEMs, and best we limit the variance to only one thing, the IEM itself,. and nothing else. So we can really hear the difference between them.
His google spreadsheet has links to his squig database where one can examine the different measurements for each eartip.
Phenomenal effort, and huge objective contribution, on his part, to the community. Actual measurements, not just opinions.
The measurements are somewhat useful, though I would argue if he’s not doing positional variation measurements, you’re not getting the full picture of how the tips mentioned would actually compare against one another.
Any of the individual impressions in the list re: the sonic effects of the tips can be discarded entirely without losing anything useful, though.
Only reason I shared the info, on the variation that tips can contribute to FR, was as supporting evidence to recommend to the OP, that when comparing IEMs, it helps to keep everything else consistent, and that includes the eartips. To avoid a change in eartips, further skewing our impressions of each IEM.
Can imagine that adding positional variation, would have been so much more of an effort.
Furthermore from a practical viewpoint, an end user inserting tips into their ear, would gain nothing from that additional information, cos its almost impossible to replicate positional variation, in a set of human real ears. So while informational this added infor, serves no true life benefit.
I had already highlighted, in a previous post, the opportunity for variation in insertion depth, to introduce variances in FR, even when one is using the same eartip, on different IEMs. Not sure how anyone could aim to achieve any consistency in insertion depth, even with the same IEM and eartip. There would be some random minute change each time.
We can also assume that in an audition, with multiple insertions of each IEM, the insertion depth variances, would average out over many auditions, i.e these variations would average/cancel out, leaving behind the most consistent and median impression of each IEM, assuming all other aspects of the chain remain consistent. So unless the evaluator was using varying extremes of insertion depth, this issue would not be a significant one, with the IEM itself being the main contributor to the subjective average impression across multiple insertions.
