Review & Bench Test: Aphex HeadPod 454 Headphone Amp

While audio pros are moving towards multi-channel monitoring systems that offer user “mixing” at each station, a whole generation of desktop recordists often deal with having only one headphone output … and still a whole band’s worth of performers needing headphones.
Author:
Publish date:

While audio pros are moving towards multi-channel monitoring systems that offer user “mixing” at each station, a whole generation of desktop recordists often deal with having only one headphone output … and still a whole band’s worth of performers needing headphones. Filling that need is the new Aphex HeadPod 454.

Image placeholder title

Features

With a list price of $249, the HeadPod is designed for tabletop use, measuring a mere five-and-a-half inches across, four-and-a-half inches deep, and two-inches high. Please note that this compact design is not “budget”: the unit has an all-steel chassis with metal jacks and firm pots.

Inputs are provided twice: first on a pair of balanced 1/4-inch TRS jacks that accept up to +4 dB with a 20 kHz ohm nominal impedance, and also on a TRS unbalanced stereo input (at 10 kHz ohms). This input provides easy connection to a headphone output with only a TRS cable, and is selected by a back-panel switch.

Four headphone outputs are provided (from four independent stereo amplifiers), each 1/4-inch TRS with their own unstepped volume control. A master input level control is provided alongside the unit’s only indicator — a single white LED. Power is provided by a 12 VAC wall-wart. Frequency response is within one dB from 10 Hz to 120 kHz (!) with -80 dB crosstalk, 35 dB max gain, and <.001-percent THD at 100 Mw at 25 ohms.

In Use

Upon powering up, the first thing I noticed was the striking white LED against the glossy white background; it’s attractive, but the LED is actually so bright it made the input level hard to see! Using some popular Fostex T20RPs with my console’s headphone output and the HeadPod’s stereo input, I received ample gain with all controls in reasonable positions. I then connected a pair of console aux sends to the balanced inputs and, again, received a good gain structure with no discernible sound quality difference. Using that same pair of auxes, I substituted both a Fostex and a Behringer headphone amp; all offered about the same sonic accuracy, although the HeadPod seemed a little cleaner and less muddy in the bottom end. A test with all four outputs cranked up really loud revealed the HeadPod’s nice headroom, maintaining lots of punch and no loss of bottom.

A test with some high impedance (600 ohm) AKG K240 headphones required the unit to run nearly wide open, right on the verge of distortion: a fine test of a headphone amp’s abilities. The HeadPod did get the job done, remaining full and clear. The bottom of the unit has input wiring graphics, as well as a handy chart showing output in milliwatts versus headphone impedance. Rated from 8 ohm to 1 kHz ohm input, the unit delivers a range of about 1200 Mw – 100 Mw, respectively. As the HeadPod is devoid of metering, settings of the master input and headphone outputs are approached “by ear,” although I was able to achieve a balance between radically divergent headphones and personal monitoring earpieces rather easily.

Summary

The HeadPod may not offer any fancy features (or many standard ones, for that matter), but it does deliver on its promise of four loud and clear headphone outputs. Rugged, attractive, and built to last, the HeadPod is the perfect solution for laptop recordists who are on the go and have access to AC power. Users of desktop computer interfaces will appreciate the additional headroom and lack of any “Y” splitters cluttering their setup. Users requiring multiple inputs, aux inputs (“mo’ me” features) or metering will have to look to more complicated systems.

Rob Tavaglione is the owner and operator of Charlotte NC's Catalyst Recording. catalystrecording.com 

----------

Bench Test

Bench Measurement Data

(Unless otherwise specified, all tests reflect signals applied to the balanced input and taken from output #1, with outputs 1 and 3 loaded on both channels, measurement bandwidth of 22 Hz to 30 kHz, or <10 Hz->500 kHz, and worst-case results.)

Input Impedance (1 kHz)
Bal: 21kø; unbal: 10 kø

Output Impedance
10ø approx.

Output Polarity
Non-inverting

Max. Input (1 kHz) into
(28 dBu), 150ø (28 dBu), 600ø (28 dBu)

Max. Output (1 kHz) into
(750mW), 150ø (540 mW), 600ø (152 mW)

Frequency Response

20 Hz-20 kHz +0.1, -0.2 dB;
10 Hz-100 kHz +0, -0.6 dB

THD+noise (20 Hz–20 kHz)
<0.03% at 1mW into 8ø, or 10 mW into 600ø

S/N (A-wtd.)
92.8 dB re: 10 mW into 8ø
90.3 dB re: 1 mW into 600ø

Crosstalk
>71 dB, 10 Hz to 20 kHz

Bench Measurement Commentary

I measured the Aphex HeadPod Model 454 with its stereo balanced inputs driving two headphone outputs (mostly because I only had fixtures readily available for four channels of load). I looked at loads of 8ø, 150ø, 600ø, and 1kø, but since there was not a great deal of meaningful difference among the latter three other than output before clipping, I stuck with the 8ø and 600ø as representative.

For the most part the graphs will speak for themselves; the HeadPod performs very well indeed, but 8ø seems pretty much to define the lower limit of its usable loads (probably due to power-supply and output-stage current limits). Since the overwhelming majority of headphones, especially in the pro environment, present loads of 100ø or more, this is not much of an issue, but does suggest that some cans may sound quite different than others. 

Figure 1 shows frequency response of a typical output and level configuration: very wide and flat with eminently useful response from 10 Hz to 40 kHz or so.

Figure 2 shows a “detail” frequency response view—please note the high-rez, half-dB vertical scale—of response driving 8 (purple/green), and 600ø (yellow/blue). Are these miniscule differences evoked by different loads audible? I very much doubt it, and anyway in the real world they’re liable to be swamped by the far greater response changes imposed by reactive-load actual headphones.

An array of THD+noise results were returned, with measurement bandwidth limited to the audio band, for 8ø and 600ø loads. The 454 is near the boundary of clipping with 8ø loads at 100 mW and even 10 mW in the lower frequencies, but with 8ø phones I would imagine this to be so hellaciously loud that an extra 10 dB or so of harmonics would hardly seem to matter.

In regards to THD+noise versus amplitude, the input signal-amplitude swept the Aphex from about 1 mW to 1w into 600 ø, and from about 0.1 mW to around 10 mW into 8ø, each at 100 Hz, 1kHz, and 20 kHz.

— D. Kumin