After narrowing down possible contenders using our series of criteria, we chose 10 of the most popular, well-reviewed DAC/amps to call in for testing. They ranged in price from $40 to $300. It was now time for our expert panel to do some serious listening. The two of us were joined by professional musician, composer, audio mixer, and Wirecutter freelance reviewer John Higgins, who has a master’s in music from USC and a bachelor’s in music and audio recording from Ithaca College.

This process became the most complex testing our audio team has conducted to date. People make a lot of claims about DAC/amps and their ability to improve the sound quality of headphones. We wondered how much of that was objectively measurable, and how much was placebo or confirmation bias. So we devised a testing plan that was as blind as possible so that we could take price, brand, and design out of the equation.

To do this, we relied on Brent to come up with a game plan. First, he built a switcher that enabled us to have four DAC/amps set up with four computers as sources. The switcher contained only one component per audio channel: a relay of the same general type commonly used in many of the best power amplifiers and preamps. Electrically, the effect of the switcher was approximately the same as connecting a high-quality 3-foot extension to a headphone’s cord. All four computers ran the same software (foobar2000) and had the same digital music files (AAC, MP3, and hi-res). The proctor matched the levels by running a 1k tone through each DAC/amp and measuring the output in mV using a Neutrik ML1 audio analyzer. We adjusted until we had a matching variable of +/-0.15 dB. This process ensured that our panelist wouldn’t be swayed by one option sounding louder or softer than the others. Each panelist sat wearing headphones with their back to a screen that blocked them from viewing which, if any, DAC/amps were currently in use.

We devised a testing plan that was as blind as possible so that we could take price, brand, and design out of the equation.

Each panelist listened to the same track through all four DAC/amps and switched between sources at will, identifying each only by the number on the dial. They could listen to as many tracks as they needed to make a determination, the average being three to five tracks. These included a mix of AAC, MP3, and hi-res selections; in most cases the tracks were test files that the panelists supplied, chosen mostly from the tracks we commonly use to test headphones and speakers. Each panelist would take notes, and the proctor would then swap in another set of four DAC/amps. The proctor also had the option of including a standard laptop headphone jack without telling the panelist. We repeated this process until each panelist had heard all the contenders.

After reviewing their notes, each panelist could then request (by grouping and number) certain DAC/amps that they wanted to hear again and compare directly. At this point the panelist would pick favorites for that set of headphones.

Because we know that not all headphones are created equal, and may not respond in the same way to a DAC/amp, we tested our contenders using specifically chosen, universally well-reviewed headphones:

Sony MDR-7506 : Our under-$150 pick, these basic dynamic headphones measure with a nice, smooth frequency-response curve. We wanted to hear how the DAC/amps affected them.

: Our under-$150 pick, these basic dynamic headphones measure with a nice, smooth frequency-response curve. We wanted to hear how the DAC/amps affected them. HiFiMan HE1000 V2 : With these high-end planar magnetic headphones known for their detail, we hoped to hear subtle differences, especially on hi-res files.

: With these high-end planar magnetic headphones known for their detail, we hoped to hear subtle differences, especially on hi-res files. Shure SE535: This pair offers well-reviewed balanced armature drivers with a massive impedance dip and spike. We hoped to hear how switching amps might affect the response.

With each new pair of headphones, the proctor changed the groupings and the switch number assignments each time to avoid confirmation bias.

We also had one Lightning-specific DAC/amp, so we performed a similar test using a matched iPhone 7, iPad Mini, and MacBook Pro as sources with a +/-0.08 dB matching variable. The proctor could choose one of the Lightning-based devices to play through the DAC/amp, and have the others use the headphone jack or the Apple Lightning headphone adapter dongle. Again, each panelist did not know which devices they were hearing, which one had the DAC/amp attached, or which one was assigned to what number on the switcher. In this final test we used AAC music files and iTunes as our playback software.

After the proctor compiled all the panelists’ listening notes, they finally revealed to the panelists which DAC/amp corresponded to the group and switch numbers. From there, we discussed our favorites, our least favorites, the competitors’ design, and their prices. The results were surprising to all of us.

Brent later ran some audio tests (including output impedance measurements) using his Audiomatica CLIO 10 FW audio analyzer to make sure none of our top picks had any performance anomalies that our ears didn’t pick up, and to get a precise idea of how loud they could play and what types of headphones they could be trusted to drive.