Denafrips Hades 12th Review: The R2R Preamp That Changes Things
A pure Class A, fully balanced, true discrete preamp with a 60-step relay-based resistor ladder volume control - perfect channel balance, 0.00045% THD, 122dB SNR.
Amplifiers
Headphone amp & integrated amplifier reviews
Amplifier reviews covering headphone amps, integrated amplifiers, and standalone power amps. Tube, solid-state, Class A, Class AB, and Class D - tested against hard-to-drive headphones, low-sensitivity speakers, and matched against reference amps in the same price tier so the verdict reflects sound, not spec sheets.
6 amplifiers reviews so far - newest first.
All amplifiers reviews
Tonewinner AD-1PA+ Review: A $3,000 Class A/AB Beast
A 43kg switchable Class A/AB power amplifier paired with a fully balanced preamp - 100W Class A or 500W Class AB into 4Ω with serious bass control and golden-hued mids.
HiFiMan EF500 Review: The Cheapest Himalaya R2R Ever
The most affordable HiFiMan unit to feature their proprietary Himalaya R2R DAC chip - 4.5W per channel balanced, network streaming, and a vertical tower form factor.
HiFiMan EF499 Review: A Real R2R DAC/Amp Under $300
A balanced R2R DAC/headphone amp with built-in network streaming for under $300 - warm, musical, and a serious one-box answer to the separates-or-not question.
Tonewinner AD-2PRO+ Review: Heavy and Powerful Class A/AB Integrated Amp
A 40 kg, 240W integrated amp with switchable Class A and Class A/B modes, neutral tonality, and the kind of dynamic headroom that pairs well with virtually any speaker.
HIFIMAN EF400 Review: Entry-Level R2R DAC + Class AB Amp Combo
An entry-level Class AB amp/DAC with HIFIMAN's Himalaya R2R DAC, fun subjective sound, and a slightly soft character - underpowered for the hardest planar loads.
Buying primer
How amplifiers reviews work on this site.
Every amplifier review on this page was run on the catalogue's reference loads - the HIFIMAN Arya Organic for headphone amps, the Diora Acoustics Chors 5 for speaker amps - fed from the Denafrips Enyo 15th DAC and Hermes 12th transport. Output is measured for clipping headroom into the actual load, not just the spec-sheet nominal impedance, so the amp's real behaviour at musical levels is what we score.
What we score in amplifiers
Power and drive
Continuous output into the actual load, peak transient capability, behaviour at the impedance dip rather than the nominal spec. Whether the amp can drive the catalogue's hardest-to-drive references with crest-factor headroom.
Tonal character
How neutrally the amp delivers the signal: solid-state transparency, the harmonic texture of Class A, the upper-harmonic spice of well-implemented tubes. Both transparent and characterful approaches can earn high scores; both have to be coherent.
Dynamics and noise floor
Quiet between notes, dynamic swing on demand, no soft compression on transients. Output stage noise relative to the load's sensitivity is part of the score for headphone amps especially.
Build and connectivity
Inputs (XLR, RCA, headphone outs of various types), output impedance behaviour, gain options, build quality of switches and knobs, daily-use ergonomics over multi-month listening windows.
Value and synergy
Performance against reference amps in the same bracket, and whether the amp's character pairs well across a range of headphones or speakers rather than being a single-pairing specialist.
How to read the scores: A 9.0 amplifier is a category benchmark - usually a piece that combines reference-tier power, low-distortion electronics, and a coherent voicing. An 8.0 amp is the top of its price bracket without being category-defining. The lower the score, the more specific the flaw: insufficient power for the load, audible coloration that does not serve the music, or build quality below the price bracket.
We do not score amplifiers by published wattage, by topology marketing, or by tube nostalgia. The rubric is built around real measurement into real loads and weeks of listening with the actual reference transducers downstream.
FAQ
Amplifiers reviews: common questions.
Buying advice, terminology, and how the amplifiers category is reviewed on The Audio Stuff.
How much power does my amplifier need?
For headphones, depends heavily on impedance and sensitivity - a 600 ohm Sennheiser HD 800 needs orders of magnitude more voltage than a 32 ohm Audeze. The headphone power calculator gives the exact requirement. For speakers, use the SPL distance calculator with sensitivity, listening distance, and target SPL.
Tube amps vs solid-state: what is the difference?
Tubes typically add a small amount of even-order harmonic distortion that many listeners describe as warmer or more dimensional, especially with vocals and acoustic instruments. Solid-state aims for transparency - lower distortion, flatter frequency response, more current on tap. Some of the best sound systems use tubes for vocals and SS for bass; many top-tier amps blur the line.
Class A vs Class AB vs Class D amplifiers: which is best?
Class A runs hot, draws constant current, and historically sounded best at the cost of efficiency. Class AB is the mainstream high-end choice - clean enough, efficient enough. Class D used to be a compromise; modern implementations (Hypex Ncore, Purifi) measure better than most Class AB and run cool. The class is no longer a quality marker; the implementation is.
Do I need a balanced amplifier for headphones?
For most headphones - no. Single-ended is fine and often electrically simpler. Balanced amplifiers double the voltage swing on tap, which helps with low-sensitivity flagships (Susvara-class) that need extreme drive. If your headphone is easy to drive, balanced topology buys you nothing audible.
How do I match an amplifier to my headphones?
Three numbers: headphone impedance, headphone sensitivity, and amplifier output impedance. Output impedance should be at least 8x lower than headphone impedance (the "rule of eighths") to avoid frequency response shifts. Power should exceed the calculator-derived requirement at the headphone impedance. Match these and the amp will not be the bottleneck.