Need a PEQ in-depth guide/resources for my IEMs

[settanta7ec]

Hello, I'm new here and in the audio world in general. I purchased my first IEM in march 2025 and got gradually more invested into the hobby, buying more pairs of them and a dongle DAC as well. I wasn't all that much into thoughtful music listening, but I started being into it a few months into discovering what IEMs were.

Having tried a handful of pairs of IEMs (will list later) has definitely helped me in recognizing what my signature preferences are and which pairs fit me the best in terms of overall comfort.

Although none of my IEMs were paid more than 60€, some of them have a much higher MSRP. That doesn't necessarily mean I prefer them over the cheaper pairs, or vice versa. Having put decent listening hours into each pair, I identified what I like and what I don't about each set.

Which is why I think the time has come for me to dig deeply into learning how to use PEQ. I don't want "quick" solutions, I don't want to use AutoEQ or other people's settings. I want to tune my IEMs to my preference or even creating different profiles for different genres/playlists.

However, I know it's hard, time-consuming work. I'm currently a bit overwhelmed by all the potential options I could apply through parametric EQ. I would like to have a guide or resources, as in-depth as possible, on how to utilize and get proficient at using PEQ.

I tried looking online, on YouTube or on the web in general, and most "in-depth" guides seem targeted at music producers/mixers. I would like to know how to use PEQ solely for music listening and, more specifically, for IEMs, since I don't own any headphone currently.

If anyone could provide me with guides and resources to learn this skill, I'd be very grateful. I think it's the only logical next step in improving my audio experience, since I already have very good gear and have a sufficient knowledge, at this point, about the audio world and my sound preferences. My current preferences are brightly-tuned IEMs without too much emphasis on the bass.

List of IEMs I own/owned: Tripowin Vivace, Tangzu Wan'er I, KZ Gale (ended up in the laundry machine, RIP), Campfire Audio IO (real stinker in terms of sound, but fits very well and was 30€), Kiwi Ears Aether (bought 2nd hand), Kbear Tourbillon TB-Pro (aliexpress discount), GK Kunten, AKG K3003i (bought 2nd hand). As a dongle DAC/Amp, I own the Jcally JM20 Max. I have also bought a few sets of replacement eartips

 

[kemmler3D]

PEQ is IMO more intuitive than you think.

I don't know of a definitive guide to PEQ that will help you use it effectively.

What I am sure of is that learning by doing will make the light bulb turn on.

Specifically: you should download FL Studio, load in some music you're familiar with as an audio clip, load up the wave candy visualizer, put it in spectrogram mode, then load up PEQ 2 on the same channel. Play the music (or pink noise) and play with the EQ while listening and watching for the effect it has on the signal in real time.

If that's too complicated just use EQAPO or something and listen while you tweak. After a short while you will have a working sense of what the various filters sound like and what different amounts of gain and Q do.

Probably the end goal of using PEQ for you is to get a maximally neutral sound signature. Opinions vary on the right way to do this, but it boils down to adjusting the frequency response with EQ so that all frequencies sound about equally loud to you.

Think of EQ simply as a volume knob that affects a limited frequency range at a time. You add more filters according to how many ranges you want to adjust individually.

 

[IsaacOscar]

Which is why I think the time has come for me to dig deeply into learning how to use PEQ. I don't want "quick" solutions, I don't want to use AutoEQ or other people's settings. I want to tune my IEMs to my preference or even creating different profiles for different genres/playlists

I've found very nice EQ profiles by using AutoEq to (very inaccurately) simulate other headphones. This can give you a starting point for an EQ and you can modify the generated one to your taste. (AutoEq also has many options you can fiddle with, and even more in the command line version).

 

[IsaacOscar]

Probably the end goal of using PEQ for you is to get a maximally neutral sound signature. Opinions vary on the right way to do this, but it boils down to adjusting the frequency response with EQ so that all frequencies sound about equally loud to you.

Umm, this only makes sense if you know the music your listening to is meant to sound "neutral". E.g. should the female singer sound much louder than the bass guiter? Unless you have some tracks that you know how they're supposed to sound (which I don't, I just guess).
I personally just choose whatever sounds nicer with the artists I listen to (and doesn't make the music sound too different to what I'm used to).

 

[kemmler3D]

Umm, this only makes sense if you know the music your listening to is meant to sound "neutral"

I should have clarified that the process of EQing something by ear usually uses test tones/ filtered noise and not music. It's not impossible, but almost, to tune FR successfully with music.

 

[IsaacOscar]

I should have clarified that the process of EQing something by ear usually uses test tones/ filtered noise and not music. It's not impossible, but almost, to tune FR successfully with music.

Ahh, I use music because that's what I want to sound better!

Human hearing is naturally more sensitive to some frequencies than others though, so wouldn't musicians already take this into account (e.g. technically louder bass, so it sounds the same as the treble)?. Do you have specific test tones in mind that are universally agreed to be meant to sound equally loud?

 

[Torin Krell]

Before using PEQs in my studio I initially used graphic EQs (5 - 15 band) and then moved on to fully PEQs after using graphic EQs for a while, when I naturally started craving finer EQ control. I would recommend this path whether you are using software or hardware graphic & parametric EQs.

 

[IsaacOscar]

Before using PEQs in my studio I initially used graphic EQs (5 - 15 band) and then moved on to fully PEQs after using graphic EQs for a while, when I naturally started craving finer EQ control. I would recommend this path whether you are using software or hardware graphic & parametric EQs.

But how do you convert between the two? (I guess you could measure the frequency response and then plug that into AutoEq, but that sounds like more effort than necessary, and would be inaccurate)

 

[Torin Krell]

But how do you convert between the two? (I guess you could measure the frequency response and then plug that into AutoEq, but that sounds like more effort than necessary, and would be inaccurate)

If you use my suggestion to learn how to use parametric EQs by first using graphic EQs you would likely not want to "convert" from the graphic to the parametric EQ because of all the added control and nuance that a parametric EQ provides. However the general shape of the graphic EQ could be your starting point when switching to the superior control possible with a parametric EQ. Having the ability to choose the exact frequency and width (Q) of the boost / attenuation at several frequencies by using a parametric EQ is vastly superior to a graphic EQ which has controls set to specific frequencies and Q, however it is exactly the graphic EQ's relative simplicity which will help you surmount the learning curves of both.

 

[settanta7ec]

PEQ is IMO more intuitive than you think.

I don't know of a definitive guide to PEQ that will help you use it effectively.

What I am sure of is that learning by doing will make the light bulb turn on.

Specifically: you should download FL Studio, load in some music you're familiar with as an audio clip, load up the wave candy visualizer, put it in spectrogram mode, then load up PEQ 2 on the same channel. Play the music (or pink noise) and play with the EQ while listening and watching for the effect it has on the signal in real time.

If that's too complicated just use EQAPO or something and listen while you tweak. After a short while you will have a working sense of what the various filters sound like and what different amounts of gain and Q do.

Probably the end goal of using PEQ for you is to get a maximally neutral sound signature. Opinions vary on the right way to do this, but it boils down to adjusting the frequency response with EQ so that all frequencies sound about equally loud to you.

Think of EQ simply as a volume knob that affects a limited frequency range at a time. You add more filters according to how many ranges you want to adjust individually.

Thank you for your reply.

I think the "goal" for me is mostly to explore new listening experiences and "fixing" my IEMs. All my pairs can definitely be improved, but some of them really need to be fixed in some parts. I'll bring some examples:

The AKG K3003i is a fantastic IEM. It retailed for a whopping 1500€ back when it came out. It fits well, nozzle size is good for me, metal body is very nice on the touch and overall shape is very comfortable, even for long sessions. I bought it 2nd hand for 60€ at the end of December. Sadly, it misses its reference filter, so I'm stuck with the treble and bass ones. The bass filter makes the in-ear sound awful. The treble filter is very good, but needs some fixing in its sound signature. In some tracks, I experience harshness with vocals and cymbals. Particularly, any time a singer pronounces the "sh" sound or the end of a cymbal vibration are extremely harsh, unpleasant and fatiguing, to the point it takes out joy from listening to the track.

The GK Kunten, on the other hand, has an extremely relaxed treble extension, to the point that I only really like listening to it late night, when I don't really want to focus too much on music listening, but rather just relax. The treble extension of the Kunten makes it sound a bit congested to me on some tracks. Particularly, any wind instrument (but mostly brass ones) sounds awful, to the point of not reaching fidelity of how they sound in reality.

And so on, for the other pairs I own. I was asking for a "technical guide" because I know what has to be fixed or what I would like to change about a sound signature, I just don't have the knowledge on how to actually do it. I am sure that I will get better by just "learning by doing", however, what I've learnt studying in university is that having a starting point, something like a blueprint, really helps me learning more effectively. I will start by following your advice and instructions and see what's going to happen. I know there are some videos on YT about PEQ, but they're always very short and focus on the very basic stuff. I know out there that there are some people really invested into EQing their gear into whatever is their preference and I wanted to know if any of those people has ever shared some of their knowledge on the subject.

 

[kemmler3D]

Particularly, any time a singer pronounces the "sh" sound or the end of a cymbal vibration are extremely harsh, unpleasant and fatiguing, to the point it takes out joy from listening to the track.

This is a pretty good scenario to practice on. If you want to get straight to work on this, create a peaking filter in whatever software you are using with gain = -6dB (fairly large adjustment), Q= 2.0 (medium-narrow, higher is narrower) and frequency = 3500hz. The frequency is a random guess at where you are getting sibilance, could be a little lower or much higher.

Repeat a harsh section of music on loop, and move the frequency around until it sounds better. Once you think you've found the center of the problem, move on.

Next adjust the gain. Most likely 6dB is too much correction, but it's possible that you will need more. Adjust the gain up and down until it sounds just right, probably a hair past the point where it just stops sounding harsh.

Lastly, try adjusting the Q of the filter. This is often the hardest to judge by ear, but if there is some residual harshness across a given frequency range, you can mitigate it with either lower Q or another high Q filter nearby.

This is how I would approach dialing in a single filter for an easily audible issue. You can do this with music when it happens to reveal a specific problem.

For the Kunten you can pick a frequency where you think the lacking treble starts, and add a high shelf filter there. Start with +2dB gain and see how much you think makes it sound natural. Alternatively you can use a low shelf filter at the same frequency and lower it by the same amount.

Finding problem frequencies and dialing in gain and Q will make more sense if you go through the process of watching the effects on a visualizer as I initially suggested, but you can jump right in doing it by ear, too.

Identifying problems by ear, that are smaller than several dB, is pretty hard. It's much easier if you can take a measurement, compare it to a reference curve, and simply see where the filters need to go and what the gain should be. This is why EQ presets are so popular, most people don't have even a cheap measurement rig, let alone a professional one.

That said, I think your goal of learning how to use PEQ properly is a good one. Target curves are never really 100% perfect for a given individual, so being able to make adjustments for yourself and eventually establish a target curve you know is better for you - and why- is the path to a next level listening experience IMO.

Something you will also want to do is reading about different target curves, how they were developed, and what decisions led to the specific features of the curve. RTINGS has an article about that and there is also a lot of stuff on the Harman curves out there.

From there you can compare your listening experiences to how your IEMs compare to a given curve, and start to develop informed opinions about how your hearing differs from that curve and finally what EQ adjustments may get you closer to a truly neutral sound.

Hope this all makes sense!

 

[settanta7ec]

This is a pretty good scenario to practice on. If you want to get straight to work on this, create a peaking filter in whatever software you are using with gain = -6dB (fairly large adjustment), Q= 2.0 (medium-narrow, higher is narrower) and frequency = 3500hz. The frequency is a random guess at where you are getting sibilance, could be a little lower or much higher.

Repeat a harsh section of music on loop, and move the frequency around until it sounds better. Once you think you've found the center of the problem, move on.

Next adjust the gain. Most likely 6dB is too much correction, but it's possible that you will need more. Adjust the gain up and down until it sounds just right, probably a hair past the point where it just stops sounding harsh.

Lastly, try adjusting the Q of the filter. This is often the hardest to judge by ear, but if there is some residual harshness across a given frequency range, you can mitigate it with either lower Q or another high Q filter nearby.

This is how I would approach dialing in a single filter for an easily audible issue. You can do this with music when it happens to reveal a specific problem.

For the Kunten you can pick a frequency where you think the lacking treble starts, and add a high shelf filter there. Start with +2dB gain and see how much you think makes it sound natural. Alternatively you can use a low shelf filter at the same frequency and lower it by the same amount.

Finding problem frequencies and dialing in gain and Q will make more sense if you go through the process of watching the effects on a visualizer as I initially suggested, but you can jump right in doing it by ear, too.

Identifying problems by ear, that are smaller than several dB, is pretty hard. It's much easier if you can take a measurement, compare it to a reference curve, and simply see where the filters need to go and what the gain should be. This is why EQ presets are so popular, most people don't have even a cheap measurement rig, let alone a professional one.

That said, I think your goal of learning how to use PEQ properly is a good one. Target curves are never really 100% perfect for a given individual, so being able to make adjustments for yourself and eventually establish a target curve you know is better for you - and why- is the path to a next level listening experience IMO.

Something you will also want to do is reading about different target curves, how they were developed, and what decisions led to the specific features of the curve. RTINGS has an article about that and there is also a lot of stuff on the Harman curves out there.

From there you can compare your listening experiences to how your IEMs compare to a given curve, and start to develop informed opinions about how your hearing differs from that curve and finally what EQ adjustments may get you closer to a truly neutral sound.

Hope this all makes sense!

Thank you very much for the detailed answer!

 

[Chromatischism]

I should have clarified that the process of EQing something by ear usually uses test tones/ filtered noise and not music. It's not impossible, but almost, to tune FR successfully with music.

Right.

Use a tone sweep from Youtube.

Adjust up/down in the areas that sound low or high vs the rest of the range.

Repeat until sound levels are mostly even.

Now go through your musical references to confirm/fine tune.

 

[IsaacOscar]

Right.

Use a tone sweep from Youtube.

Adjust up/down in the areas that sound low or high vs the rest of the range.

Repeat until sound levels are mostly even.

Now go through your musical references to confirm/fine tune.

Right, but we naturally hear different frequencies at different volumes. So I think what you need is one matched to an "equal loudness curve"?

(I found a "hearing" test thing on Peace that takes this into account and generates an PEQ for you, I didn't get good results with it, but I'll try again)

 

[Chromatischism]

Right, but we naturally hear different frequencies at different volumes. So I think what you need is one matched to an "equal loudness curve"?

For you personally, that correction will be "what you hear" at that volume level.

Equal Loudness adjusts for changes in level, which you can still use after the fact.

So probably pick one level to use for testing/tweaking and always use that same level each time you come back to it.

 

[10PSPMemSticksForMiku2007]

A reasonable level to calibrate with would be roughly ~85db average. The ear's frequency response is decent there. You would play noise of that volume on speakers for reference, using a spl meter or calibrated microphone for example to know how loud it is. Then you can work with comparable volume on your IEMs. Once you have them equalized at ~85db, you'll have the proper starting point for loudness equalization.

 

[Klonatans]

My advice would be to to use a B&K5128 graph as a reference, if available, and to patientlly change PEQ settings until you are satisfied with the final result. Adherence to specific target is not needed as there's a high likelihood that it won't coincide with your preference and that you will not hear what's shown in the graph, especially in the treble area.

 

[settanta7ec]

My advice would be to to use a B&K5128 graph as a reference, if available, and to patientlly change PEQ settings until you are satisfied with the final result. Adherence to specific target is not needed as there's a high likelihood that it won't coincide with your preference and that you will not hear what's shown in the graph, especially in the treble area.

This is a pretty good scenario to practice on. If you want to get straight to work on this, create a peaking filter in whatever software you are using with gain = -6dB (fairly large adjustment), Q= 2.0 (medium-narrow, higher is narrower) and frequency = 3500hz. The frequency is a random guess at where you are getting sibilance, could be a little lower or much higher.

Repeat a harsh section of music on loop, and move the frequency around until it sounds better. Once you think you've found the center of the problem, move on.

Next adjust the gain. Most likely 6dB is too much correction, but it's possible that you will need more. Adjust the gain up and down until it sounds just right, probably a hair past the point where it just stops sounding harsh.

Lastly, try adjusting the Q of the filter. This is often the hardest to judge by ear, but if there is some residual harshness across a given frequency range, you can mitigate it with either lower Q or another high Q filter nearby.

This is how I would approach dialing in a single filter for an easily audible issue. You can do this with music when it happens to reveal a specific problem.

For the Kunten you can pick a frequency where you think the lacking treble starts, and add a high shelf filter there. Start with +2dB gain and see how much you think makes it sound natural. Alternatively you can use a low shelf filter at the same frequency and lower it by the same amount.

Finding problem frequencies and dialing in gain and Q will make more sense if you go through the process of watching the effects on a visualizer as I initially suggested, but you can jump right in doing it by ear, too.

Identifying problems by ear, that are smaller than several dB, is pretty hard. It's much easier if you can take a measurement, compare it to a reference curve, and simply see where the filters need to go and what the gain should be. This is why EQ presets are so popular, most people don't have even a cheap measurement rig, let alone a professional one.

That said, I think your goal of learning how to use PEQ properly is a good one. Target curves are never really 100% perfect for a given individual, so being able to make adjustments for yourself and eventually establish a target curve you know is better for you - and why- is the path to a next level listening experience IMO.

Something you will also want to do is reading about different target curves, how they were developed, and what decisions led to the specific features of the curve. RTINGS has an article about that and there is also a lot of stuff on the Harman curves out there.

From there you can compare your listening experiences to how your IEMs compare to a given curve, and start to develop informed opinions about how your hearing differs from that curve and finally what EQ adjustments may get you closer to a truly neutral sound.

Hope this all makes sense!

For the AKG K3003, I ended up trying kemmler's suggestion for fixing the harshness. It didn't really work though. So I picked up an autoEQ setup on the equalizer APO database for my IEM and tried using that. While the harshness and sibilance of vocals and cymbals went away, I was really not satisfied with the new sound signature.

So, I used the autoEQ settings as reference on how to proceed. I knew I had to take away rather than adding, so I looked in the profile where they took away instead of adding. Turns out, there's a harsh peak at 5400 Hz on the IEM graph and that's also what corrected the sibilance. I ended up making the -6dB gain and Q = 2.0 at 5400 Hz and it solved most problems, but not all the sibilance. So I ended up making another adjustment at 11 kHz (where I noticed another peak in the graph) with -3dB gain and Q = 1.5.

I would say I'm currently satisfyied with the result and, switching between the autoEQ profile and the one I created, I like the tonality of mine way, way more. It keeps the IEM bright, as I like it to be, without the annoying harshness.

I think for other IEMs it will be much harder to get similar results, because the "problems" are less evident, especially by ear alone. I will surely give it a shot when I'm going to have more time in my hands. I will also try and see if there are more potential scenarios I can explore in terms of sound signatures. This is mostly because I like different tunings with different genres, or even artists/groups, for the matter. Listening to a lot of jazz, a lot of tracks are mixed poorly. Newer mixes are also lackluster, in my opinion. I have found that some IEMs are a real hit on certain tracks and artists, but a miss on others. I could try to experiment to take out the best listening experience possible out of these poorly produced tracks. For newer stuff or other genres, this is not relevant. But jazz and its variants are still my favourites and the ones that give me the best emotions while listening, so I have to work around these obstacles.

 

[kemmler3D]

Not mentioned in this thread yet but IMO very useful, a tone generator where you can sweep in real time: https://www.szynalski.com/tone-generator/ very good for identifying problem frequencies.

 

[IsaacOscar]

Not mentioned in this thread yet but IMO very useful, a tone generator where you can sweep in real time: https://www.szynalski.com/tone-generator/ very good for identifying problem frequencies.

My problem with such tools is human hearing is naturally more sensitive to some frequencies, so recordings will already account for this. Equalizer APO's Peace Add-on has a "hearing test" feature that takes this into account and can generate EQ's.