EQing Headphones Video, does Auto EQing IEMs based on measurements, have all the same challenges?

IEMs / In-Ear-Monitors General IEM Discussion
1

Just when I think I understand, I watch this video again, and then it hits me. AutoEQ on headphones is just a starting point. And I get that. Albeit these concepts are not the easiest to grasp, for anyone who is not an audio scientist.

I’m still struggling with the two causes of deviation that could make AutoEQ not the most effective.

  1. Headphone measurements depend on the “ear” on which they are measured
  2. We also have our individual HRTF

I once thought this was all the same variable - i.e a single factor, but the video indicates they are separate variables. That was something completely new to me, and I know it may take me a while to get my head round this. I must admit I still do not understand the difference, and maybe google AI will be my next port of call, to explain it to me further.

Now on to IEMs, which is the listening domain where I live, over two years, and all my personal listening, except auditioning over ear headphones, has been on IEMs.

Question. Since IEMs are not measured with an “ear” i.e a physical emulation of a human ear, does that imply that AutoEQ may be more relevant to IEMs, or more effective on IEMs, compared to applying AutoEQ on headphone measurements, since IEMs and IEM measurements do not have to contend with a variance from an “ear”,

Of course I can imagine that AutoEQ applied to IEM measurements, will have its own issues, discussed in the video, such as :

  1. Unit Variances, between our own copy of an IEM, and the unit measured
  2. Hearing differences between our own ears (left different from right) due to ear canal and our own biology - different frequency response at each ear.
  3. Variances between whatever rig was used to measure the IEM, and our own HRTFs(for each ear)

Based on this video, I have adopted in addition to using AutoEQ as a starting point, for IEM EQ correction, further small adjustments by ear in the 2K to 9K region using the tone generator GoldenSound showed in the video. Subjectively I have been extremely pleased with the end result. All my critical listening is via routing all audio on a desktop through a DAW, where I can apply several layers of parametric EQ. Layer 1 is the AutoEQ correction filters, as generated by squig.link, and Layer 2 includes any further manually defined filters based on listening for loudness of frequencies via the tone generator.

But if I understand the video well enough, does this imply that AutoEQ applied to IEM measurements, does not need us to consider variances in how the measurement rig responds to the outer ear, cos in this case, there is no outer ear affecting the measurement, and because when we also use IEMs, we do not have an outer ear which has an impact on the sound of the IEM?

Merely wishing to double check if my thinking on the subject, as it applies to IEMs, is correct., and happy to be corrected, if I’ve missed something, and got it wrong.

2

It’s arguably worse for IEMs because you’re bypassing pinna effects, which need to be assumed by the response. So you’re essentially flying blind. We try to account for this by using generic population average pinna effects in the compensation (JM-1), but inevitably you’ll have some disconnect between the graph and your experience.

Also, crucially, with IEMs you need to identify your length mode resonance with the product. Unless you get lucky and it’s in the perfect spot for you, there’s typically a peak somewhere between 5-9khz, caused by the air volume between the iem and the eardrum (the canal).

1 Like
3

Thanks. Your clarification has helped immensely.

Indeed any AutoEQ, must have the caveats you have described. These caveats would also apply to any IEM at the design/manufacture in their aim to achieve a target, whatever the justification supporting that target. i.e there will be deviations based on a variance between the target and individual experiences.

So the result of the AutoEQ, or rather the caveats post EQ, are identical to the caveats of any IEM FR, before any attempt to correct that FR via EQ.

Anyone attempting an AutoEQ or a process largely based on that, MUST accept these caveats, and I certainly do.

All caveats accepted, and this is only anecdotal, i.e listening experience of ONE person, even when AutoEQ’d with minor corrections to deal with the most significant resonances and nulls, between 2 and 9khz, it’s informative to hear how each IEM, supposedly “tuned” to the same target, as best as one can, still sounds different. Similar yet different.

If FR was the significantly predominant factor in how an IEM is perceived, the assumption would be that if all were tuned to the same FR, with all caveats accepted, they would sound pretty identical. I’m going through a few IEM’s AutoEQing to the same target - JM-1(+Tilt), and I accept this is imperfect cos the measurements I am relying on were all done on 711 couplers. It’s early days in this experiment. Each IEM sounds better, significantly better, to my ears, but the improvement via AutoEQ+manually created filters to counter the individual resonances, has not been equivalent. One IEM, in my case, which was probably one of the worst sounding, before EQ - and I would normally not listen to this without EQ, now sounds to me really superb, accurate, etc, etc, deep, wide, all the superlative adjectives apply. It would take a study, to examine if this is a repeatable observation.

Whatever the case, I’m extremely pleased with the results, and have begun to apply the process, to each IEM I own, and now MUST include parametric EQ capability in any DAC I buy going forward, so that the EQ correction, is not limited to implementation in software, and I can enjoy this improvement on any IEM, on any source with a digital output - via an EQ enabled DAC.

I have added a few things to what was discussed in the video, these are the steps, in entirety, that have given me the best results.

  1. Choose measurement of IEM on squig.
  2. Generate EQ correction filters, via squig to whatever is the desired target FR, here I choose 800hz as the normalize frequency.
  3. Option - add additional filters, via squig, if not satisfied with filters generated in step 2.
  4. Implement filters defined in Step 2 and 3. listen via a tone generator to sweep from 1khz to about 10Khz, and define additional filters in real time, while listening, to even out any significant dips and peaks. I try to limit this to no more than the top 2 or 3 dips and peaks that are quite audible. This is a manual attempt to correct any significant actual dips and peaks, from the combination of the IEM itself, and one’s own ear canal.
  5. Final polish adds 2 more gentle slope filters(i.e low Q values below 0.3), a low shelf @ 250hz, to boost or cut the low end to taste, and a similar a high shelf @ 7Khz. This is intended to address any errors of measurement in the rigs, i.e differences between what the human being hears, and how the rigs measure, for each IEM, and also address personal taste.
  6. By listening to a few tracks, you’ll be able to estimate how much overall gain compensation (plus or minus) is needed, post EQ, and include that in the EQ, if that is possible, so that engaging the EQ, does not have a noticeable change in loudness.

I do not have the scientific explanation, but it’s only fair to state, in my experience, foam tips tend to reveal a lot more of the benefit of EQ, to me, and even without EQ, also seem to be more revealing of the sonic signature of the IEM. Silicone tips are convenient, but do not seem to allow the improvement via EQ to be as audible as with foams. Indeed getting the right size and fit for foams can be a bother, especially with any variances in ear canals. I’ve had to modify mine to get them to fit better with some IEMs.

4

One more caveat, our eartips are not assured to be the same as the eartips used to create the measurements on squigs.

So one should think of this as a best endeavour, rather than an effort to achieve perfection. Hopefully in spite of the caveats, if we can reduce the most significant deviations of the stock IEM FR, from our desired target, that’s good enough.

5

EQ for headphones seems like a bloody crap shoot. I gave up on IEM’s years ago. I find them to be quite uncomfortable, regardless of design. Some are worse than others.

I found the sound changes dramatically when one shifts the fit slightly in the ear. Same when you change out the ear pads. Additionally, I found the connectors to many of them to be too fragile.

In summary, it seems to me that trying to EQ IEM’s is all trial and error.

6

I do some work as an audio engineer, and am also a vocalist and musician. In those pursuits, IEMs have become essential for personal monitoring, and are the norm in our world today. Which is why I considered it important a few years ago, to take a deep interest in these products, which we must use, for live performances.

Indeed there are many variables, which would skew each person’s own experience and opinion of IEMs. Because it was important for me to get to the bottom of how best to use IEMs, as I had no choice, what I can say is that if one searches, there is a lot more information out there that can assist one in choosing and using IEMs to maximum advantage, but one has to search in many places, and there is also the risk of misinformation, sifting through some false or wrong information to get at the truth.

Even amongst professional audio people, when I started this journey 2 and a half years ago, there was so much ignorance, and very little information available. As we speak, the information is out there thankfully, but its all over the place, a bit here, a bit there. So it takes effort to make sense of it all, and it will most likely take buying a few items, to have a better understanding of the segment.

I am not surprised that anyone would give up on IEMs, if they attempted to do this a few years ago, when the resources were limited, and it was more like the dark ages of alchemy, or the wild wild west.

But things are much better now. Some of the reviewers on headphones.com such as Resolve, do use IEMs, thankfully, and there are some IEM related reviews, and threads on the various headphones.com portals, such as the main web site, this forum, and the Headphone Show channel on Youtube. Some others like Crinacle, have invested a huge amount of research and also share a ton of info on IEMs, and there are a host of decent IEM reviewers out there like Akros, Audio Amigo, AndyAudioVault, and many more, then there are sites like head-fi.org.

The proliferation of information, does not make it any easier, cos there are also contradictions from diverse opinion leaders, but at least if one makes the effort, there is a lot of knowledge available publicly which was not the case a few years ago.

The availability of products in this segment has improved a whole lot, and the quality continues to improve. Moreso, the cost has come down a whole lot, in excellent DACs, Headphone Amps, Dongle DACS which combine the function of a DAC and a headphone Amp, and IEMs also can be bought for more reasonable prices, than before.

Yes there is a steep learning curve, and many surprises. Learning how to optimise every single component in the audio chain, DAC+Headphone Amp, cables (indeed in my experience they do not sound the same cos of inconsistencies in cable design and manufacture), IEMs and eartips, does require an investment in time and money, and patience. Learning what IEMs and eartips provide the best fit or preferred frequency response, for each of us, takes some work, to optimise our purchasing decisions. But when one has all these choices implemented optimally, the result is spectacular sound, but of course even that description is subjective, cos its my opinion.

Yes it took a long while to get this far, but I feel that with IEMs, I have better control over the variables, and the outcome is well worth it, in the end. But it has many frustrations along the way, cos there is not much help with grasping a proper understanding of IEMs.

I agree with you, minor positional variations can have quite an impact on how an IEM sounds. The answer for me was an unavoidable and deep dive into eartips, and nozzle lengths, with personal customisation of commercially available eartips, Lots of listening and comparing to appreciate the factors that cause an eartip/nozzle combination, to sound different from any other combination.

The current appeal of today’s IEMs, is one can have a fair number of different ones, and experience and maintain in one’s collection, a variety of sonic signatures.Achieving this with headphones or speakers, would most likely cost a lot more.

An important fairly recent development has been the coming of age of decent sounding more affordable planar magnetic IEMs, and then we have competition between especially the Chinese manufacturers of IEMs, and related accessories, which further drives down the costs.

1 Like
7

Thanks for this insightful response. Always good to hear other perspectives.

Glad to hear the IEM world is improving. As for me, I’ll likely stick with over ear headphones for the foreseeable future. I could see where IEM’s would be ideal for travel, as that was how I used them. For now, I’ll just stick with the Sennheiser Momentum 4 for travel.

9

For the benefit of anyone interested in IEMs

One of the constraints with discussing IEMs, is the variability. With headphones, the vast majority of users use the stock pads, and even where one may change the pads, to achieve a different tone, it does not vary drastically from the stock pads that were delivered with the headphone.

With IEMs, it is so easy to change the eartips, and in the hobby, or for those who use these professionally, every different eartip sounds different, and the change in tone can be very significant. So when discussing the same IEM, two different reviewers can come to completely different opinions simply because they could have been listening with very different eartips on the same IEM.

There are literally over one hundred different brands/models of eartips, that one could use on most IEMs. This makes the end result so variable, and comparing opinions.

If I may be direct. In my experience, the biggest “mistake” that makes most IEMs sound not as good as they could, is because people are listening with silicone or liquid silicone eartips. Such eartips, in my personal experience severely limit the performance of the IEM, making them sound more dull than they could.

So in my experience there are a few rules, for pairing eartips to an IEM, to enable it deliver its best result.

  1. Use a foam eartip, where possible.
  2. An eartip inner bore diameter, that is as similar as possible to the inner diameter of the nozzle exit. This will enable the proper propagation of all the frequencies.
  3. An outer diameter that fits snugly in one’s ear canal. This assures a proper seal, and is important to get the bass sounding really wonderful.

There is a fourth criteria for the correct eartips, in my own experience.

  1. A combination of the eartip length and nozzle length, when the eartip is fitted on the nozzle, that matches one’s own ear canal. If this combination length is too long, the IEM will be difficult to fit properly. If too short, this combination length causes the IEM to be inserted too deep in the ear, and sounds too bright.

The moment one gets foam tips with these criteria met, the sound of the IEM reveals itself properly, and has a clarity that has to be heard to be believed. They begin to sound really good.

I discovered these criteria, by accident, well some was by accident, but some was via an extensive study of what makes an eartip suitable or not.

There is a directness and accuracy of an IEM, with a properly fitted set of foam eartips, which is impossible to deliver via headphones or speakers, cos the IEM, has less of an interaction with anything else, such as our outer ear or reflections from the headphone housing or reflections from walls in a room.

10

This might be the case for you but for the vast majority of IEM users, silicone ear tips are superior to foam. Doesn’t mean foam isn’t better for some people, or with certain designs, but I don’t think it’s fair blanket recommendation to use foam.

All of this is variable depending on the IEM and the person’s anatomy. So sometimes narrow bore will work better, other times wide bore, but there’s no rule that says “this style tip and bore size is best”, it just doesn’t work that way. You can look at IEM tuning trends and consider how they typically perform relative to your HRTF, so for example I tend to look for narrow bore tips more often than not. But even then, length modes are heavily influenced by the shell and the way it actually fits in your ear.

What I suggest is… look at how different bore sizes impact the FR of the IEM - see if you can find sources that publish additional insertion depths and bore size results for IEM measurements.

11

Please re-read your response. You may wish to revise it. You may have been a bit trigger happy to critique me, and unfairly so.

Why?

Contrary to your assertions, nowhere have I I have made no attempt to give any general rules, for anyone, or everyone, but always qualifying, with phrases like “in my own experience”, or “in my experience”.

Hopefully it’s ok to share my own experience, for the benefit of others, to consider, without having to suffer for it, via the usual online critiques.

12

No it’s fine to talk about your experience, as you have done. I’m giving the counterpoint to this:

“So in my experience there are a few rules, for pairing eartips to an IEM, to enable it deliver its best result.”

This may be the case for you, for the IEMs and tips you’ve tried, but I don’t think that makes it correct as a rule. So in other words, I’m voicing my disagreement with your perspective on this, not saying you can’t have that perspective.

1 Like
13

Minor nitpick, but this has not been my experience. See Ear Pads can make for a great upgrade

1 Like
14

Agreed. Pad changes make a significant difference to the sound. Rarely for the better compared to stock pads in my experience, but certainly a noticeable change.

1 Like
15

Depends. The original stock pads for the Abyss V2 are certainly not as good as the revised version(s) of stock pads currently available from the manufacturer.

I think many of the aftermarket ear pads for a given headphone are not as good as the stock pads. There are a few exceptions.