Both are dongles, $299 against $210. This page puts them side by side on the only three things a maker can be held to — power, output impedance and noise — and then asks the question that actually matters: which headphones do they treat differently?
The EarMen Colibri reaches 100 dB on 563 of the 567 models here, the EPZ TP50 on 561.
| Specification | EarMen Colibri | EPZ TP50 |
|---|---|---|
| Power into 32Ω | 75 mW | 121 mW |
| Power into 150Ω | 16 mW | 26 mW |
| Power into 300Ω | 8.0 mW | 13 mW |
| Power into 600Ω | 4.0 mW | 6.5 mW |
| Output impedance | not published | not published |
| Output noise | not published | 1.0 µV |
| Price | $299 | $210 |
| Headphones it reaches 100 dB on out of 567 | 563 | 561 |
A figure marked not published is missing from the maker’s own material, which is not the same as being good or bad. Where one of these two publishes a number and the other does not, the comparison on that axis cannot be made at all — and saying so is more useful than guessing.
| Load | EarMen Colibri | EPZ TP50 |
|---|---|---|
| 32 Ω | Published | Published |
| 150 Ω | Derived | Derived |
| 300 Ω | Derived | Derived |
| 600 Ω | Derived | Derived |
Both makers publish a power rating, which is less common than the table above makes it look: of the 280 sources here, 209 publish a figure at at least one load and only 2 publish one at all four. EarMen Colibri publishes at 32 Ω; EPZ TP50 at 32 Ω. Every other row is worked out from those — between two published loads, or beyond them holding voltage constant, which understates rather than flatters.
The two publish at the same loads, which is the one case where the rows of this table can be read straight across.
What each of these labels means, and why a derived figure is still worth printing, is set out in how this data was built.
4 headphones reach 100 dB on the EarMen Colibri but not on the EPZ TP50 — Moondrop COSMO, Symphonium Crimson, Symphonium Helios, 1 more.
2 headphones reach 100 dB on the EPZ TP50 but not on the EarMen Colibri — HiFiMAN Susvara, Moondrop SKYLAND.
559 of the 567 models here reach 100 dB on both. That is a power count and nothing more; whether either source suits a given headphone also depends on the two lines above.
The counts above say how many models the two treat differently. This says by how much. The same 8 headphones appear on every page of this kind, ordered from the easiest to drive to the hardest, so a row can be read across pages as well as across columns. Headroom is at a 100 dB SPL peak.
The widest gap on the list is the Campfire Audio Andromeda 2020: EarMen Colibri has +13 dB of headroom there against +9 dB. 4 dB is about 2.5 times the power, which matters for peaks rather than for how loud the music sits.
The hiss columns are empty for EarMen Colibri because that maker publishes no output noise figure. That is a gap in the specification, not a quiet amplifier, and it is left blank here rather than filled with a plausible number.
| Headphone | Impedance | Needs | EarMen Colibri headroom | EPZ TP50 headroom | EarMen Colibri hiss | EPZ TP50 hiss |
|---|---|---|---|---|---|---|
| Shure SE846 | 9 Ω | 0.04 mW | +33 dB | +29 dB | not published | 14 dB |
| Campfire Audio Andromeda 2020 | 9 Ω | 4.0 mW | +13 dB | +9 dB | not published | -6 dB |
| Audeze LCD-2 Classic | 70 Ω | 0.79 mW | +16 dB | +18 dB | not published | -7 dB |
| Beyerdynamic DT 1990 Pro (250Ω) | 250 Ω | 0.63 mW | +12 dB | +14 dB | not published | -12 dB |
| HiFiMAN HE400se | 25 Ω | 7.9 mW | +11 dB | +11 dB | not published | -13 dB |
| Audio-Technica ATH-R70X | 470 Ω | 1.6 mW | +5 dB | +7 dB | not published | -19 dB |
| Sennheiser HD 600 | 300 Ω | 6.6 mW | +1 dB | +3 dB | not published | -23 dB |
| HiFiMAN Susvara | 60 Ω | 50 mW | -1 dB | +1 dB | not published | -25 dB |
The EarMen Colibri costs $89 more than the EPZ TP50. Out of the 567 models here, that difference buys 4 more models driven to 100 dB.
It does not run one way only: the cheaper EPZ TP50 drives 2 models to 100 dB that the EarMen Colibri does not, which is the part worth checking against your own headphones before paying the difference.
Everything in this paragraph counts only figures the two makers published. An advantage on an axis one of them leaves blank cannot be counted, and is not counted — how often that happens is its own page.
Each has a page of its own with every power figure traced to what the maker published and a full headphone list: EarMen Colibri and EPZ TP50. Or put either against your own headphones in the source comparison tool.