[fzman]

Question for those in the know: does IEM FR change at different volume levels? I know that's probably a silly question, but I notice bass response, for example, is more prominent with some IEMs as volume increases, which got me thinking that this must affect the overall FR (or at least FR perception). Which also means we should be able to dial in a certain volume level on an FR graph and see what the IEM is doing at that volume level - otherwise what we're looking at may not be representative of the IEM at our personal listening volume.

Great question - probably lots of factors at play here. There's the whole Fletcher-Munson thing, along with the interaction of the amplifier and the iem -- imedance curve, how a reactive load effects the amp, that kind of thing.

In the speaker world, some speakers soound better when they are pushed harder/louder

 

[chickenmoon]

Question for those in the know: does IEM FR change at different volume levels? I know that's probably a silly question, but I notice bass response, for example, is more prominent with some IEMs as volume increases, which got me thinking that this must affect the overall FR (or at least FR perception). Which also means we should be able to dial in a certain volume level on an FR graph and see what the IEM is doing at that volume level - otherwise what we're looking at may not be representative of the IEM at our personal listening volume.

https://en.wikipedia.org/wiki/Equal-loudness_contour

 

[gLer]

https://en.wikipedia.org/wiki/Equal-loudness_contour

Well that scrambled my brain without getting me any closer to an answer

 

[RTodd]

Well that scrambled my brain without getting me any closer to an answer

What I took from it was the answer is YES different frequencies are influenced by power level in a non linear and non equal way, with all sorts of other variables also influencing what we hear, any one of which could diminish or amplify the perceived difference.

The part about third level harmonics messing up the ability to measure and create a standard was interesting.

 

[EMINENT]

Question for those in the know: does IEM FR change at different volume levels? I know that's probably a silly question, but I notice bass response, for example, is more prominent with some IEMs as volume increases, which got me thinking that this must affect the overall FR (or at least FR perception). Which also means we should be able to dial in a certain volume level on an FR graph and see what the IEM is doing at that volume level - otherwise what we're looking at may not be representative of the IEM at our personal listening volume.

Well according to the latest firmware fr measurements in the XM4 thread, yes!

 

[chickenmoon]

Well that scrambled my brain without getting me any closer to an answer

Well, broadly speaking, bass and treble perception isn't as good at low volume than it is at higher volume and so increasing volume should make the extreme of the spectrum more prominent.

Loudness switches on vintage amplifiers boost bass and treble more at low volume than they do at higher volume for that very reason.

 

[mashuto]

Well that scrambled my brain without getting me any closer to an answer

I'm pretty sure the actual frequency response doesn't change at different volume levels (or shouldn't), but, how we perceive it does indeed change based on volume.

 

[cocolinho]

Question for those in the know: does IEM FR change at different volume levels? I know that's probably a silly question, but I notice bass response, for example, is more prominent with some IEMs as volume increases, which got me thinking that this must affect the overall FR (or at least FR perception). Which also means we should be able to dial in a certain volume level on an FR graph and see what the IEM is doing at that volume level - otherwise what we're looking at may not be representative of the IEM at our personal listening volume.

That's a good question and I believe you can get some feedback from all those guys measuring a bunch of IEMs like @crinacle or @Banbeucmas . They probably experienced this

 

[Deezel177]

Question for those in the know: does IEM FR change at different volume levels? I know that's probably a silly question, but I notice bass response, for example, is more prominent with some IEMs as volume increases, which got me thinking that this must affect the overall FR (or at least FR perception). Which also means we should be able to dial in a certain volume level on an FR graph and see what the IEM is doing at that volume level - otherwise what we're looking at may not be representative of the IEM at our personal listening volume.

I talked about this a while ago and it's - as @fzman mentioned - related to the Fletcher-Munson curve. Basically, the bass will drop off a cliff as you drop the volume, and it'll perceivably rise the most as you increase it. As @mashuto said, though, that's only in relation to how our ears perceive the sound. An IEM, when driven at a higher or lower volume, will still read exactly the same in a frequency response graph, because a microphone doesn't work exactly like our ears and brain do.

Below, I've taken three quick measurements of the A18t. The middle one is control (or the reference), the one above it was taken after I increased the volume of the amp, and the one below it was taken after I reduced the gain on the measurement mic:

As you can see, whether I changed the load on the amp or the gain on the mic, the curve comes out exactly the same. So, when it comes to reading graphs, those parameters shouldn't really matter.

When it comes to listening though, it does matter quite a bit. I always write my reviews with the assumption that the IEM is listened to at an average or listenable volume; when dynamics sound life-like, and when any outside noise whatsoever is completely drowned out. But, there have been scenarios where listening volume is a relevant factor. When someone says an IEM is brighter or warmer than they expected based off of reviews, then it's valid to ask that volume they heard it at. Like I mentioned when I did talk about the Fletcher-Munson curve, it's also a semi-viable way to ever-so-slightly tune your IEMs as well by driving it louder or softer.

 

[fzman]

I'm pretty sure the actual frequency response doesn't change at different volume levels (or shouldn't), but, how we perceive it does indeed change based on volume.

I think the perception changes as our hearing is not 'Flat'. I think it is possible that the frequency response of the system (ie, amp speakers aka iems and cable) may vary. FR is usually measured 1 tone at a time, then graphed. Music is more complex, and not every amp circuit responds equally to back emf and reactive loads. Qucik analogy: It's much easier to hold a calm, basically motion-less dog than if that same dog is agitateted and squirming vigorously.

 

[gLer]

I'm pretty sure the actual frequency response doesn't change at different volume levels (or shouldn't), but, how we perceive it does indeed change based on volume.

Is that because sine sweeps are used for FR curves as opposed to actual music, and therefore can't simulate how we would hear actual sounds at different volume levels?

When someone says an IEM is brighter or warmer than they expected based off of reviews, then it's valid to ask that volume they heard it at. Like I mentioned when I did talk about the Fletcher-Munson curve, it's also a semi-viable way to ever-so-slightly tune your IEMs as well by driving it louder or softer.

This gets much closer to the crux of my question. In actual fact then, listening volume is even more important, in many ways, than what you see on a FR graph, because changes to volume necessarily affect how we hear the different frequencies irrespective of what we see on the graph. Take the Legend X, for example. It can legitimately sound like three different IEMs at three different volume levels because the perception of bass kicks in significantly more at higher volumes, interacting with and affecting the other frequencies in different ways. Similarly, if someone says an IEM sounds too bright for them, increasing the volume (and increasing the bass response) can (depending on how the treble is tuned, type of drivers etc) actually balance out (effectively reduce) that brightness compared to listening to at lower levels.

Add to this the fact that different types of music emphasise different parts of the FR curve, and you suddenly have a minefield of very specific criteria that affect a person's perception - and hence their reality - of an IEM.

 

[gLer]

Well according to the latest firmware fr measurements in the XM4 thread, yes!

This suggestion on the XM4 thread is actually what got me thinking about it and asking the question here. I don't believe FR should change - unless an IEM (or source) is applying DSP that affects the FR at different volume levels. Using bitperfect output this should never be the case, but I've heard of powered speakers, for example, that have DSP that reduces certain frequencies at higher volumes to reduce distortion, which could be what's happening with the XM4 (speculation on my part).

The fact that we hear different FR (and the resulting FR interactions) at different volumes, however, blows the whole idea of using standardised measurements to get a sense of an IEM's sound at higher or lower volume levels out the water...something I've never (or perhaps very rarely) seen explicitly mentioned anywhere.

 

[mashuto]

Is that because sine sweeps are used for FR curves as opposed to actual music, and therefore can't simulate how we would hear actual sounds at different volume levels?

I'm not sure of the actual science but I don't think so. It's the difference between actual real volume and perceived loudness. So bass playing at the same sound pressure level or volume might be perceived at a different loudness level than mids or treble, and the perceived loudness of different parts of the frequency spectrum does change based on volume. Though I guess it could in part change how instruments are perceived since they aren't a single frequency and certain parts could be perceived at different loudness levels.

The way I always look at it and part of the reason I tend to prefer bassier iems is that I like to listen at lower volume levels. So the perceived loudness of bass is less, requiring more actual bass presence for it to be perceived the way I want. Also likely part of the reason why I wasn't a huge fan of the ie900 and why I felt it was bass light. At the volume levels I prefer, I perceived the bass to be light comparatively. Turning the volume up brought the bass more in line, but at that point it was just louder overall than I would have liked.

Again, I could be off base, that's just how I have understood the topic.

 

[Deezel177]

Is that because sine sweeps are used for FR curves as opposed to actual music, and therefore can't simulate how we would hear actual sounds at different volume levels?

This gets much closer to the crux of my question. In actual fact then, listening volume is even more important, in many ways, than what you see on a FR graph, because changes to volume necessarily affect how we hear the different frequencies irrespective of what we see on the graph. Take the Legend X, for example. It can legitimately sound like three different IEMs at three different volume levels because the perception of bass kicks in significantly more at higher volumes, interacting with and affecting the other frequencies in different ways. Similarly, if someone says an IEM sounds too bright for them, increasing the volume (and increasing the bass response) can (depending on how the treble is tuned, type of drivers etc) actually balance out (effectively reduce) that brightness compared to listening to at lower levels.

Add to this the fact that different types of music emphasise different parts of the FR curve, and you suddenly have a minefield of very specific criteria that affect a person's perception - and hence their reality - of an IEM.

That's exactly the reason why IEM measurements can't be looked at on their own. You would indeed be looking at a plethora of possible sounds an IEM could have. "It looks like the lows and highs are pretty even, but will my ears perceive it that same way?" "Will this be the tone I hear at the listening volume I prefer?" Etcetera, etcetera.

This is why it's often said that an IEM's FR graph is best viewed as a reference against another graph; one of an IEM that you've preferably already heard or - better yet - already know. You don't imagine what an IEM sounds like just by looking at its graph. You extrapolate what it sounds like by comparing it against graphs of other IEMs. And, that extrapolation process is you applying your "standard" listening volume, your biases, etc. to the graph and turning it into sound you can imagine.

For example, I want to imagine what the A18s sounds like compared to the A18t. I'd compare the two graphs, I'd look at all the differences, then I'd apply those differences to the sound of the A18t I have in my head, which I've perceived and internalised at my standard listening volume.

So, for me, graphs are a completely viable way of imagining what an IEM will sound like, but only if it's done by comparing that graph against one of an IEM I already know. The scenarios you've outlined are exactly why I think this way, and why I think imagining an IEM's graph without any other form of context is completely impossible.

 

[548184]

Nic (you know, the shootout guy) actually wore all his IEMs outside, often paired with white-sleeved, 12-wire, flat-braided, can't miss those unless you're blind, SilverFi cables. And he lives in Amsterdam...