# Speaker distortion and compression: where the numbers come from

Researched 2026-10-18. Values in js/speaker-dynamics.js were read from published charts (about ±2 dB on dB-relative Klippel charts). Charts were downloaded only to a temporary folder for reading, not kept.

Each band's THD is the geometric mean of two published frequencies: sub-bass 30 and 50 Hz, bass 100 and 200 Hz, midrange 500 Hz and 1 kHz, treble 3 and 10 kHz. It is split between the 2nd and 3rd harmonics according to which one dominates in the chart. The sub-bass figure is then restated against the drive level (see below). THD over 40 % is capped, after that conversion. Compression is the output lost at 102 dB @ 1 m relative to 76 dB, read per band from Erin's "instantaneous compression" sweeps (or SoundStage linearity). Where no compression measurement exists, the typical value for that kind of speaker is used and the source says so.

## How compression acts (2026-10-25)

Erin's Audio Corner runs two compression tests. The **instantaneous** test plays 2.7-second logarithmic sweeps at 76, 86, 96 and 102 dB @ 1 m back to back and compares each with the 76 dB sweep. For some speakers a **long-term** test plays a music-like multitone at 86 or 96 dB for two minutes, then for two more, and sweeps again after each. Klippel describes the two causes these separate: the motor and suspension running out of linear range as the cone travels (following the signal), and voice-coil heating (seconds to minutes) ([Klippel, compression of fundamental components](https://www.klippel.de/know-how/measurements/nonlinear-distortion/compression-of-fundamental-components.html)).

**Long-term (heating) results at 96 dB @ 1 m**, traced from Erin's charts, band means, after 4 minutes:

| Speaker | Sub-bass | Bass | Mid | Treble | Largest at any frequency |
| --- | --- | --- | --- | --- | --- |
| Klipsch Heresy IV | 0.0 | 0.0 | 0.0 | 0.0 | 0.22 dB at 53 Hz |
| Kii Three | 0.30 | 0.30 | 0.0 | 0.0 | 0.68 dB at 95 Hz |
| Dutch & Dutch 8C | 0.08 | 0.24 | 0.0 | 0.0 | 0.35 dB at 18 kHz |
| JBL M2 | 0.0 | 0.0 | 0.0 | 0.0 | 0.04 dB at 42 Hz |

The instantaneous loss of the same speakers at the same 96 dB is several times larger (Kii Three about 1 dB in the sub-bass, from its 102 dB values and the level law below). So the simulator applies the instantaneous figures to each peak as it happens and does not model heating; the site says so.

**Level law.** Traced from Erin's instantaneous charts for six speakers (KEF LS50 Meta, Vandersteen Treo CT, Wharfedale Linton, Dynaudio Special Forty, Linkwitz LXmini, Klipsch Heresy IV) at every sixth-octave from 40 Hz where all three curves show compression and the 102 dB loss exceeds 0.4 dB (123 points): the loss at 96 dB is a median 0.39 of the 102 dB loss, and at 86 dB 0.08. Loss in dB proportional to drive amplitude to the power 1.36 fits both (6 dB down: 0.39; 16 dB down: 0.08). Proportional to power would give 0.25 and 0.025; to amplitude, 0.50 and 0.16. Above 102 dB @ 1 m, where no sweep goes, the same law is extrapolated.

**Timing (our choice).** No measurement gives how fast compression lets go after a peak. It follows the band's peak over one cycle of its lowest frequency (20, 80, 300, 3000 Hz) and lets go over one more cycle: any faster and the gain would ripple within a cycle, adding distortion the charts don't show. The bands are detected with 4th-order Butterworth filters at 80 Hz, 300 Hz and 3 kHz, and the gains are applied with a broadband gain and three low shelves, solved so each band's centre (40, 155, 949, 7746 Hz) gets its band's loss. tests/dynamics-check.js checks the calibration, the level law, a 100 ms burst and that compression adds no harmonics.

## Sub-bass against the drive level (2026-10-27)

Published THD is a share of what the speaker puts out at that frequency. The simulator (js/speaker-processor.js) adds the harmonics to the signal before the speaker's response, scaled to the drive level: the level that gives the test SPL in the midrange. Below a speaker's bass roll-off its output is 10–25 dB down, so the published share overstates the harmonics it adds several-fold. Each speaker's sub-bass THD is therefore multiplied by its measured on-axis response at 30 and 50 Hz (dB against its 300 Hz–3 kHz level, the level the tests are set by): THD × 10^((r30 + r50) / 40), the geometric mean of the two converted points. From 100 Hz up the responses are within a few dB of the midrange (the Magnepan's 6 dB down at 100 Hz is the largest), so the other bands are left as published.

Found while adding the Neumann KH 120 II: read as a share of its filtered-out 30 Hz output (over 100 %, capped at 40 %), its harmonics made it 9.3 dB louder than the Revel F328Be at 79 Hz in the verdict and graded its distortion audible, though its measured response is flat there. Its sub-bass is read directly against the drive level from ASR's absolute chart instead (its limiter also holds the bass back at 96 dB, which the quiet response doesn't show). The typical profiles (TYPICAL_DYNAMICS) and the O/96's sub-bass, a typical estimate, are not converted.

| Speaker | Published sub-bass THD % | Response at 30 / 50 Hz, dB | Against the drive level, % |
| --- | --- | --- | --- |
| KEF LS50 Meta (neutral coaxial) | 36 | -27.2 / -5.5 | 5.4 |
| Revel F328Be (Harman-tradition floorstander) | 1.4 | -9.8 / -5.8 | 0.57 |
| KEF Blade 2 Meta (coaxial point-source tower) | 6.3 | -4.5 / -1.4 | 4.5 |
| Polk Audio Monitor 40 Series II (budget bookshelf) | 22 | -22.8 / -9.8 | 3.4 |
| Rogers LS3/5A (BBC monitor) | 10 | -18.6 / -10.9 | 1.9 |
| Harbeth Monitor 30.0 (BBC lineage) | 9.0 | -19.7 / -8.6 | 1.8 |
| Bowers & Wilkins 805S (tweeter-on-top) | 14 | -14.8 / -1.7 | 5.4 |
| Dynaudio Special Forty (Danish bookshelf) | 10 | -19.3 / -4.0 | 2.6 |
| Wilson Audio TuneTot (luxury compact) | 89 | -22.5 / -14.4 | 11 |
| Vandersteen Treo CT (time-coherent) | 8.7 | -5.8 / -1.9 | 5.6 |
| Wharfedale Linton 85th Anniversary (heritage three-way) | 5.2 | -15.7 / -3.3 | 1.7 |
| Audio First Designs Cadentia 3 (kit three-way) | 2.5 | -8.4 / -1.1 | 1.5 |
| Audio First Designs Fidelia (kit two-way) | 21 | -17.0 / -4.1 | 6.1 |
| JBL 4349 (horn studio monitor) | 1.1 | -10.2 / -2.8 | 0.55 |
| JBL 4367 (big-box horn studio monitor) | 0.86 | -10.8 / -1.7 | 0.42 |
| Klipsch Heresy IV (horn-loaded heritage) | 9.4 | -16.6 / -9.8 | 2.1 |
| DeVore Orangutan O/96 (wide-baffle high-efficiency) | 1.1 | -13.6 / 0.6 | 1.1 (a typical estimate, not converted) |
| Magnepan LRS (planar dipole) | 10 | -25.8 / -9.6 | 1.3 |
| Bose 901 Series V (direct/reflecting) | 22 | -9.3 / 0.6 | 13 |
| Linkwitz LXmini (omnidirectional DIY) | 32 | -11.3 / -3.1 | 14 |
| Kii Three (active cardioid) | 15 | -2.3 / 0.4 | 14 |
| Dutch & Dutch 8C (active cardioid) | 13 | 0.3 / -0.4 | 13 |
| Genelec 8351B (active coaxial studio monitor) | 4.4 | -7.6 / -1.3 | 2.6 |
| Neumann KH 120 II (compact active studio monitor) | 75 | -21.2 / -0.6 | 11 (read directly from ASR's absolute levels) |
| Arendal 1610 Slim 8 (sealed on-wall three-way) | 8.6 | -16.0 / -7.5 | 1.76 (read directly from ASR's absolute levels; the conversion gives 2.2) |
| JBL M2 (reference studio monitor) | 0.90 | -3.6 / -1.4 | 0.67 |

| Speaker | THD % at test level (sub / bass / mid / treble) | Test level | Compression dB at 102 dB | Max SPL (active) | Source |
| --- | --- | --- | --- | --- | --- |
| KEF LS50 Meta (neutral coaxial) | 36 / 3.8 / 0.32 / 0.13 | 96 dB | 3 / 1.5 / 0.5 / 0.2 |  | Erin's Audio Corner (96 dB THD, 102 dB compression); agrees with Audio Science Review and SoundStage |
| Revel F328Be (Harman-tradition floorstander) | 1.4 / 0.21 / 0.15 / 0.21 | 96 dB | 1.5 / 1 / 0.7 / 0.5 |  | Audio Science Review THD; no compression measurement published (typical tower values) |
| KEF Blade 2 Meta (coaxial point-source tower) | 6.3 / 0.59 / 0.13 / 0.13 | 96 dB | 0.35 / 0.2 / 0.2 / 0 |  | Erin's Audio Corner ("negligible" compression) |
| Polk Audio Monitor 40 Series II (budget bookshelf) | 22 / 1.1 / 0.32 / 0.84 | 96 dB | 4 / 2 / 1 / 0.8 |  | Audio Science Review THD; no compression measurement published (typical bookshelf values) |
| Rogers LS3/5A (BBC monitor) | 10 / 0.63 / 0.39 / 0.28 | 86 dB | 4 / 2 / 1 / 0.8 |  | Audio Science Review THD at 86 dB (the owner asked that it not be pushed to 96 dB); no compression measurement published (typical bookshelf values) |
| Harbeth Monitor 30.0 (BBC lineage) | 9.0 / 0.95 / 0.40 / 0.30 | 96 dB | 2.7 / 2.2 / 0.5 / 0 |  | SoundStage/NRC Model 30 Domestic (2009): THD+N at 90 dB @ 2 m (about 96 dB @ 1 m), compression 95 vs 70 dB @ 2 m; mid and treble at the noise floor |
| Bowers & Wilkins 805S (tweeter-on-top) | 14 / 0.67 / 0.80 / 0.27 | 96 dB | 4 / 2 / 1 / 0.8 |  | Audio Science Review THD (1 kHz spike annotated as an artifact, ignored); no compression measurement published (typical bookshelf values) |
| Dynaudio Special Forty (Danish bookshelf) | 10 / 1.3 / 0.82 / 0.18 | 96 dB | 1 / 0.6 / 0.7 / 0.05 |  | Erin's Audio Corner (96 dB THD, 102 dB compression); agrees with SoundStage |
| Wilson Audio TuneTot (luxury compact) | 40 / 1.7 / 0.26 / 0.28 | 96 dB | 4 / 2 / 1 / 0.8 |  | Audio Science Review THD, port open (over 80 % at 30–50 Hz, capped at 40 %); no compression measurement published (typical bookshelf values) |
| Vandersteen Treo CT (time-coherent) | 8.7 / 3.5 / 0.76 / 0.22 | 96 dB | 1.7 / 3 / 1.5 / 0.4 |  | Erin's Audio Corner (96 dB THD, 102 dB compression) |
| JBL 4349 (horn studio monitor) | 1.2 / 0.49 / 0.21 / 0.89 | 96 dB | 1 / 0.5 / 0.3 / 0.3 |  | Audio Science Review THD, agrees with Hi-Fi News at 90 dB; no compression measurement published (typical horn values) |
| Klipsch Heresy IV (horn-loaded heritage) | 9.4 / 2.1 / 1.9 / 1.1 | 96 dB | 0.8 / 0.4 / 0.3 / 0 |  | Erin's Audio Corner (96 dB THD, 102 dB compression) |
| Magnepan LRS (planar dipole) | 10 / 2.5 / 0.32 / 0.12 | 90 dB | 3 / 1.4 / 0.8 / 1.5 |  | Audio Science Review THD (true level below the labeled 96 dB; about 90 dB assumed); compression from Erin's LRS+ measurement as a proxy |
| Quad ESL-57 (original electrostatic) | 7.1 / 1.4 / 0.42 / 0.28 | 96 dB | 3 / 2 / 1 / 1 | 100 | No measurements; typical panel values with a 100 dB limit from Quad's own guidance (the panels arc if overdriven) |
| Bose 901 Series V (direct/reflecting) | 22 / 8.7 / 2.2 / 0.70 | 99 dB | 2 / 0.8 / 0.5 / 1.7 |  | Erin's Audio Corner, with the active equalizer at 14 V (about 99 dB at 1 m) |
| Linkwitz LXmini (omnidirectional DIY) | 32 / 2.2 / 1.3 / 0.16 | 96 dB | 4 / 1.5 / 1.5 / 1.3 | 104 | Erin's Audio Corner; output-limited in its default DSP configuration |
| Kii Three (active cardioid) | 15 / 2.7 / 0.16 / 0.13 | 96 dB | 3 / 2.4 / 0.3 / 0.6 | 112 | Erin's Audio Corner; built-in excursion limiter above about 96 dB; max SPL from Kii's specification (110 dB short-term, 115 dB peak) |
| Dutch & Dutch 8C (active cardioid) | 13 / 4.7 / 1.0 / 0.15 | 96 dB | 1.45 / 0.1 / 0.7 / 0.6 | 110 | Erin's Audio Corner; max SPL from the 106 dB long-term specification |
| Genelec 8351B (active coaxial studio monitor) | 4.4 / 0.69 / 0.070 / 0.21 | 96 dB | 1 / 0.5 / 0.3 / 0.3 | 113 | Audio Science Review THD; no compression test published (typical values); max SPL from Genelec (113 dB short-term, driver protection) |
| JBL M2 (reference studio monitor) | 0.90 / 0.37 / 0.19 / 0.54 | 96 dB | 0.3 / 0.3 / 0.3 / 0 | 123 | Erin's Audio Corner ("practically no compression"); max SPL 123 dB from JBL |
| Wharfedale Linton 85th Anniversary (heritage three-way) | 5.2 / 1.7 / 0.58 / 0.48 | 96 dB | 2.2 / 0.75 / 0.24 / 0.66 |  | Erin's Audio Corner; the 102 dB curve runs off the bottom of the chart (−3 dB) from about 45 to 80 Hz, counted as 3 dB, so the sub-bass figure is a lower bound |
| JBL 4367 (big-box horn studio monitor) | 0.86 / 0.32 / 0.20 / 0.65 | 96 dB | 0.2 / 0.05 / 0.05 / 0 |  | Erin's Audio Corner (almost no compression at 102 dB) |
| Audio First Designs Cadentia 3 (kit three-way) | 2.55 / 0.49 / 0.23 / 0.40 | 96 dB | 0.17 / 0.28 / 0.12 / 0.12 |  | Erin's Audio Corner (Klippel NFS). THD traced by colour at 30, 50, 100, 200, 500 Hz and 1 kHz; at 3 and 10 kHz THD equals the 2nd harmonic (the THD trace is hidden under it). Compression traced as band means of the 102 dB curve (sub-bass from 41 Hz, where the chart starts) |
| Audio First Designs Fidelia (kit two-way) | 21 / 6.7 / 0.19 / 0.18 | 96 dB | 1.6 / 2.0 / 0.35 / 0.8 |  | Erin's Audio Corner (Klippel NFS), traced 2026-10-27. THD by colour at 30 Hz (88 %), 50 Hz (4.8 %), 100 Hz (15 %), 200 Hz (3.0 %), 500 Hz (0.37 %), 1 kHz (0.094 %) and 3 kHz (0.089 %); at 10 kHz the THD trace is hidden under the 2nd harmonic, read there (0.36 %). The 2nd harmonic dominates except near 50 Hz. Compression as band means of the 102 dB curve (sub-bass from 40 Hz, where the chart starts); the curve runs off the chart at −3 dB from about 120 to 200 Hz, counted as 3 dB, so the bass figure is a lower bound. Erin sums it up as "~3–4 dB" of loss at high volume |
| Neumann KH 120 II (compact active studio monitor) | 75 / 0.70 / 0.18 / 0.31 | 96 dB | 1 / 0.5 / 0.3 / 0.3 | 108 | Audio Science Review (Klippel NFS), traced 2026-10-27 from the 96 dB chart (absolute levels, cross-checked against the percentage chart). 100 Hz 1.3 %, 200 Hz 0.37 %, 500 Hz 0.29 %, 1 kHz 0.12 %, 3 kHz 0.24 %, 10 kHz 0.41 % of the fundamental. Sub-bass read differently: Neumann's 40 Hz, 12 dB/octave bass filter cuts the fundamental to about 70 dB at 30 Hz and 88 dB at 50 Hz, so as a share of it the THD is over 100 % and 15 %; but the simulator adds harmonics to the full-level signal before the speaker's response, so what it needs is the harmonic level against the 96 dB drive level: 81.5 dB at 30 Hz (19 %) and 72 dB at 50 Hz (6.3 %). Read as a share of the cut fundamental (40 %, the cap), it added harmonics around 80 Hz: against the Revel F328Be the verdict put the biggest gap at 79 Hz (9.3 dB) and graded the distortion audible; read against the drive level, the biggest gap is 5.1 dB at 63 Hz, as the two measured responses show, and the distortion isn't flagged. Elsewhere the fundamental is within about 3 dB of 96 dB, so the two readings agree. Amir notes the clip light came on 1 dB above this level. No compression test published (typical active values). Max SPL from Neumann's specification: 108 dB(C) short-term with IEC 60268-5 noise at 1 m |
| Arendal 1610 Slim 8 (sealed on-wall three-way) | 8.6 / 0.91 / 0.43 / 0.10 | 96 dB | 1.5 / 1 / 0.7 / 0.5 |  | Audio Science Review (Klippel NFS), read 2026-11-01 pixel by pixel from the 96 dB charts. 30 and 50 Hz from the absolute chart (harmonics 56 and 66 dB, 7.7 % and 9.7 % of the output there, mostly 2nd harmonic): the percentage chart runs off its 5 % scale below 60 Hz. 100 Hz–10 kHz from the percentage chart (0.86, about 1.0, 0.65, 0.29, 0.07 and 0.15 %). No compression test published: typical tower values (our choice, for an 8-inch woofer with a separate midrange) |
| DeVore Orangutan O/96 (wide-baffle high-efficiency) | 1.1 / 0.5 / 0.08 / 0.5 | 90 dB | 1.5 / 1 / 0.7 / 0.5 |  | Hi-Fi News THD at 90 dB, one frequency per band (100 Hz, 1 kHz, 10 kHz); sub-bass is the typical tower value halved for the 6 dB lower level, and compression is the typical tower value (neither published) |

## With a subwoofer (2026-10-28)

With subwoofers on (Build's room setup), the speakers no longer play the sub-bass band: the simulator's band split (below 80 Hz, js/speaker-processor.js) is the same as the bass management's crossover (js/room.js). So engine.speakerParams sets the speaker's sub-bass distortion and compression to zero, and an active speaker's limiter, and the meter, follow only what the speaker still plays (through the bass management's 80 Hz high-pass). When the limiter acts it turns the subwoofers down too, since one output feeds both (our approximation). The subwoofers play clean: there are no measurements to model one. The other bands keep the values above, measured without a subwoofer; with the deep bass taken off, a woofer moves less, so they may be a little pessimistic. Before this, a speaker with a subwoofer still added its sub-bass harmonics (which pass the 80 Hz high-pass) and compression.

Verdicts of the three subwoofer comparisons, before and after (same tracks, 85 dB):

| Comparison | Before | After |
| --- | --- | --- |
| Bookshelf speakers with and without a subwoofer (KEF LS50 Meta, Trio) | Obvious, 7.92 steps | Obvious, 8.20 steps |
| Subwoofer in the corner vs. middle of the front wall (Einimali) | Clear, 3.93 steps | Obvious, 4.02 steps |
| One subwoofer vs. two (Trio) | Clear, 2.55 steps | Clear, 2.55 steps |

No distortion was audible either way at 85 dB. The corner vs. front-wall change is a threshold crossing: the average tone difference grew by 0.03 dB (1.18 to 1.21 dB) once the LS50 Meta's own sub-bass distortion and compression, the same on both sides, dropped out. A repeat run gave identical figures.

## Typical values

Zu Audio Method: no level-dependent measurements published. It uses the typical profile for its kind and is labeled typical.

Pi Speakers four π: no distortion or compression measurements published (Pi gives only a maximum of 123 dB, from a half-space measurement). It uses the typical horn profile and is labeled typical.

Klipsch La Scala AL5: no distortion or compression measurements found (Stereophile's report has none; Hi-Fi News's lab report, which gives distortion at 90 dB, could not be opened). It uses the typical horn profile and is labeled typical. Klipsch rates its maximum output at 121 dB.

Aperion Novus N6SR: Erin's Audio Corner measured its distortion and compression, but his data policy forbids digitizing his graphs, so it uses the typical bookshelf profile and is labeled typical (permission pending).

Typical compression with measured distortion (`compTypical` in js/speaker-dynamics.js): Revel F328Be, Polk Monitor 40 II, Rogers LS3/5A, B&W 805S, Wilson TuneTot, JBL 4349, Genelec 8351B, Neumann KH 120 II, Arendal 1610 Slim 8 and DeVore O/96. Audio Science Review and Hi-Fi News publish distortion but no compression test.

**Loud-level labels (since 2026-10-31).** The "Clean when loud" and "Compresses when loud" labels, and the taste test's punch-at-loud-levels preference when it suggests speakers, compare only the speakers whose compression was measured (15 speakers, plus the Magnepan through its LRS+ proxy). Each is scored against those alone; the others get no label and count as neutral. Before, the typical values took part: they put "Compresses when loud" on the Polk, LS3/5A, B&W 805S and TuneTot and "Clean when loud" on the JBL 4349 and Genelec, and shifted the measured speakers' scores (the LS50 Meta and Treo now read "Compresses when loud"; the Special Forty no longer reads "Clean when loud").

## Pages

- KEF LS50 Meta: https://erinsaudiocorner.com/loudspeakers/kef_ls50_meta/, https://www.audiosciencereview.com/forum/index.php?threads/kef-ls50-meta-review-speaker.25574/, https://assets.kef.com/pdf_doc/LS50Meta/HFN_June_KEF_LS50_Meta.pdf
- Revel F328Be: https://www.audiosciencereview.com/forum/index.php?threads/revel-f328be-speaker-review.17443/
- KEF Blade 2 Meta: https://erinsaudiocorner.com/loudspeakers/kef_blade2_meta/
- Polk Monitor 40 Series II: https://www.audiosciencereview.com/forum/index.php?threads/polk-monitor-40-series-ii-review-speaker.24146/
- Rogers LS3/5A: https://www.audiosciencereview.com/forum/index.php?threads/rogers-ls3-5a-bbc-speaker-review.49746/
- Harbeth Model 30 Domestic: https://www.soundstagenetwork.com/measurements/speakers/harbeth_30_domestic/
- B&W 805S: https://www.audiosciencereview.com/forum/index.php?threads/bowers-wilkins-805s-bookshelf-speaker-review.16523/
- Dynaudio Special Forty: https://erinsaudiocorner.com/loudspeakers/dynaudio_special_forty/
- Wilson TuneTot: https://www.audiosciencereview.com/forum/index.php?threads/wilson-audio-tunetot-review-high-end-bookshelf-speaker.29219/
- Vandersteen Treo CT: https://erinsaudiocorner.com/loudspeakers/vandersteen_treo_ct/
- JBL 4349: https://www.audiosciencereview.com/forum/index.php?threads/jbl-4349-review-studio-monitor-speaker.22686/, https://www.technologyfactory.eu/data/mediablocks/JBL-4349-HiFi-News.pdf
- Klipsch Heresy IV: https://erinsaudiocorner.com/loudspeakers/klipsch_heresy_iv/
- Magnepan LRS / LRS+: https://www.audiosciencereview.com/forum/index.php?threads/magnepan-lrs-speaker-review.16068/, https://erinsaudiocorner.com/loudspeakers/magnepan_lrs_plus/
- Bose 901 Series V: https://erinsaudiocorner.com/loudspeakers/bose_901_series_v/
- Linkwitz LXmini: https://erinsaudiocorner.com/loudspeakers/linkwitz_lx_mini/
- Kii Three: https://erinsaudiocorner.com/loudspeakers/kii_three/
- Dutch & Dutch 8C: https://erinsaudiocorner.com/loudspeakers/dutch_dutch_8c/
- Genelec 8351B: https://www.audiosciencereview.com/forum/index.php?threads/genelec-8351b-review-studio-monitor.23432/, https://www.genelec.com/8351b
- JBL M2: https://erinsaudiocorner.com/loudspeakers/jbl_m2/, https://jblpro.com/en/products/m2
- Wharfedale Linton 85th Anniversary: https://erinsaudiocorner.com/loudspeakers/wharfedale_linton_85/
- JBL 4367: https://erinsaudiocorner.com/loudspeakers/jbl_4367/
- Audio First Designs Cadentia 3: https://erinsaudiocorner.com/loudspeakers/audio_first_cadentia3/
- Audio First Designs Fidelia: https://erinsaudiocorner.com/loudspeakers/audio_first_fidelia/
- Neumann KH 120 II: https://www.audiosciencereview.com/forum/index.php?threads/neumann-kh120-ii-monitor-review.46362/, https://www.neumann.com/en-en/products/monitors/kh-120-ii/
- Arendal 1610 Slim 8: https://www.audiosciencereview.com/forum/index.php?threads/arendal-1610-slim-8-speaker-review.74101/
- DeVore Orangutan O/96: https://hifinews.com/content/jadis-jpl-mkiise300b-devore-fidelity-o96-tube-ampspeaker-system-lab-report-devore-orangutan
- Quad ESL-57: no measurements; 100 dB limit from Quad guidance (panels arc if overdriven)