A frequency-response graph shows how strongly a measured IEM reproduces frequencies under a specific test setup. It is one of the most useful tools for estimating tonal balance, but it is not a complete picture of sound quality. The graph must be read with its measurement rig, insertion method, compensation or target, smoothing, and channel traces in mind. A curve without context can create more certainty than the data supports.
Use frequency response to predict tonal relationships and identify likely peaks or dips. Do not use it as a direct score for resolution, soundstage, speed, or comfort.
Understand the axes first
The horizontal axis is frequency, usually displayed logarithmically from roughly 20 Hz to 20 kHz. Equal visual distance therefore represents multiplication rather than equal numbers of hertz. The vertical axis is sound-pressure level in decibels. A few decibels can be audible when the change spans a broad region, while a narrow high-frequency feature may be more sensitive to measurement position.
Raw IEM response is not expected to be a flat line. The ear canal and measurement coupler shape the response, and preferred targets normally include bass support and an upper-mid rise. Compare against an appropriate target or against another IEM measured on the same rig.
Read broad regions before individual peaks
Begin with sub-bass, mid-bass, lower mids, upper mids, treble, and air. Broad relationships usually affect the overall character more than a tiny local deviation. Ask where the bass shelf begins, how cleanly it returns into the mids, how strong and broad the upper-mid rise is, and whether lower treble contains obvious excess energy.
Use consistent band definitions within the site. One practical editorial system is 20-60 Hz sub-bass, 60-250 Hz mid-bass, 250 Hz-1 kHz lower mids, 1-4 kHz upper mids, 4-10 kHz treble, and 10-20 kHz air. The boundaries are approximate and should not be treated as acoustic walls.
| Region | Approximate range | Common perceptual effect |
|---|---|---|
| Sub-bass | 20-60 Hz | Rumble, depth, low-frequency scale |
| Mid-bass | 60-250 Hz | Punch, warmth, rhythm, potential masking |
| Lower mids | 250 Hz-1 kHz | Body, fullness, instrument fundamentals |
| Upper mids | 1-4 kHz | Vocal presence, articulation, attack |
| Treble | 4-10 kHz | Definition, consonants, cymbal energy, sharpness |
| Air | 10-20 kHz | Shimmer, openness, upper-harmonic extension |
A target is a reference, not a universal truth
A target curve represents a chosen reference under a specific measurement method. Preference research has shown that many listeners favor responses with meaningful bass and an ear-related upper-mid shape, but individual preference varies. Different rigs and target systems are not directly interchangeable.
Use the target to identify relative deviations, then describe those deviations in audible language. Avoid calling every departure incorrect. A bass-focused or warm product may intentionally deviate from a general preference target to serve a particular listener.
Check rig, insertion, and high-frequency uncertainty
IEM measurements depend on seal and insertion depth. A leak can reduce bass; a different insertion can move resonances in the upper range. Standard ear simulators also have resonances that make interpretation above the lower treble more difficult. Very high-frequency traces should therefore be read more cautiously than broad bass and midrange trends.
Look for left and right traces, repeated insertions, smoothing information, and consistent test level. Channel matching matters because a large mismatch can affect center image and tonal balance.
Know what the graph cannot show directly
Frequency response does not directly show soundstage size, imaging precision, transient speed, comfort, shell pressure, cable noise, or the texture of a recording. Distortion and impedance are separate measurements. Spatial presentation also depends on the recording, channel relationships, HRTF cues, fit, and listener anatomy.
The graph can still explain part of those experiences. Excess mid-bass may increase masking, and upper-frequency extension may support openness. The mistake is turning correlation into a complete causal claim.
Use a repeatable graph-reading sequence
First confirm the rig and target. Second inspect seal and bass extension. Third compare bass-to-mid transition. Fourth read the broad upper-mid shape. Fifth identify lower-treble peaks and dips. Sixth treat the highest frequencies cautiously. Finally, connect the graph to listening notes and fit.
When comparing two products, use measurements from the same source whenever possible. A few decibels of vertical offset or different compensation can make unrelated graphs look more different than they are.
- Same rig and compensation for comparisons
- Left and right channels visible
- Seal and insertion conditions understood
- Broad regions read before narrow peaks
- High-frequency uncertainty acknowledged
- Listening notes do not claim more than the graph supports
Kiwi Ears product references
These products are used only as concrete examples of architectures, tuning goals, and ownership considerations. They are not a ranked recommendation list. Verify current specifications, pricing, availability, and approved product language before publishing.
Product highlights
- KARS 2.0 — 8dB sub-bass · clean 200Hz cut
- 10mm Titanium — Titanium-coated diaphragm
- Warm-neutral — Smooth, non-offensive
- Big Soundstage — Clean, good imaging
- Ultra-light 4.45g — Comfort for hours
- Accessible benchmark — Gateway to hi-fi
Product highlights
- DLC Dynamic — Fast, impactful sub-bass
- Dual Knowles BA — Natural, resolving mids
- Planar Driver — Speed and detail
- PZT Bone Conductor — Sparkle and air
- Coherent U — Lively, non-fatiguing
- Studio Resolution — Punches above its price
Product highlights
- 8 Balanced Armatures — Orchestra pedigree
- Warm-neutral — Mature, non-fatiguing
- BA Subwoofers — Tight, fast, textured bass
- Sweet Mids — Pronounced, delicate, lush
- Detail Retrieval — Rivals double the price
- Wide Stage — Airy, pinpoint imaging
Five-driver tribrid IEM with 10 mm dynamic, two Knowles BAs
$449.00
Product highlights
- Tuned by HBB — Signature bass tuning
- Tribrid · 5 Drivers — 1DD + 2BA + 2EST
- 15dB Sub-bass — Deep, dense, physical
- Knowles Mids — Dual 32257 BAs
- Sonion EST Treble — Smooth, extended
- Controlled Power — Balanced much of the time
Frequently asked questions
Why does a neutral IEM graph not look flat?
The ear canal and measurement reference shape the response. A useful in-ear target includes an upper-mid rise and often bass support rather than a horizontal raw line.
Can I predict soundstage from frequency response?
Only partially. Tonal balance affects masking and openness, but spatial perception also depends on recording cues, acoustic design, fit, and individual anatomy.
Why do graphs disagree above 8-10 kHz?
Insertion depth, coupler resonance, rig differences, and small positioning changes make upper-treble measurements more variable.
Sources and editorial notes
This is a base editorial draft prepared for internal revision. General principles are supported by the sources below. Product examples link to current official Kiwi Ears pages and should be rechecked immediately before publication.
- AES: Statistical Model of In-Ear Headphone Preference (opens in a new tab)
- AES: Listener Preferences for Different Headphone Target Response Curves (opens in a new tab)
- Knowles: Measuring Balanced Armature Drivers for Hi-Res Earphones (opens in a new tab)
- Audio Precision: Headphone Electroacoustic Measurements (opens in a new tab)
- Official product page: Kiwi Ears Cadenza II (opens in a new tab)
- Official product page: Kiwi Ears Quintet (opens in a new tab)
- Official product page: Kiwi Ears Orchestra Lite (opens in a new tab)
- Official product page: Kiwi Ears x HBB Punch (opens in a new tab)
