Why 384 kHz? Eclipse manufacturer Antelope is committed to building technology to high standards and those who can afford it.

 

When I first looked over the specs and features of the Antelope Eclipse, two words came to mind. The first word was “Wow.” This unit has so much packed into a two rack-space box that it is shocking. I could spend this entire review just detailing the feature set, which includes three main functions: converter, monitor controller and clock. 

 

The Eclipse includes a mastering quality AD/DA, supporting sampling rates up to 384 kHz at 24-bits, a monitor control section with precision stepped attenuator, outputs for three sets of monitors, two main headphone outputs, two Cue headphone outputs, bass management with LFE output, talkback switching, and a super-stable clock all in a very attractive and comprehensive package.

 

This is a no-compromise design and the list of superlative features goes on. From the relay-controlled, stepped precision attenuator (L/R matched to within 0.05 dB) ranging from 0 to -91 in single dB steps, to the dynamic range of the AD/DA (123 and 129 dB, respectively, at 0.0004% THD+N), the oven-controlled ultra-low jitter clock (there are two independent sample rates offered for dual outputs), it is clear that no expense was spared in making this a top-of-the-line offering.

 

When pulling the Eclipse out of its glossy, four-color box, the first thing I notice is its weight. At 17 pounds, this unit feels more like a power amp than a converter. The ultra-clean design features a glass and black metal front panel, accented by two brushed aluminum knobs, ten buttons and two large 32-segment, stereo LED meters. The back panel is wall-to-wall connectors: XLR combo jacks (18 of them), RCAs (x3), ¼” (x10), TOSLINK (x3), BNC (x6), a USB port and IEC power connection. No D-sub connectors here! I did discover that all the WC OUT connectors were slightly recessed which made it impossible to lock down a BNC connector to them. 

 

The I/O complement is impressive as well. Three balanced analog inputs (XLR/TRS ¼” combos), eight digital inputs (3 AES/XLR, 2 SPDIF, 2 TOSLINK), three analog monitor outs, two DAC outs, 1 LFE out and two main and two Cue headphone outs, along with four Digital Outs (two converter out/two de-jittered out) and SPDIF and TOSLINK. One Word Clock (WC) In and four WC Outs along with an Atomic Clock Input round out the clocking I/O. Lots of connectivity.

 

For a unit this powerful to have such a minimal set of front-panel controls (only five) seems counter-intuitive. The unit requires the control software on a separate computer to access all the parameters. The software is written for either PC or Mac, is downloadable from the internet and is painlessly installed. The computer connects via a single USB 2.0 cable, also capable of carrying a 384 kHz data stream. With the software, there are a plethora of options for inputs and outputs, far too many to describe here, along with exhaustive monitoring options, I/O metering and a built-in oscillator. Inquisitive minds can find the manual online.

 

Hooking up the Eclipse was a breeze, with all connections clearly marked. The bright meters (-40 to 0 dB) are wonderful and the software control is very clear and intuitive, even self-explanatory. Hookup and navigation was accomplished without the manual. The volume knob, calibrated in 1 dB steps with numeric readout, is a delight. The clicking of the relays, an audible confirmation of the accurate stepped resistors, however is very loud. If the unit is racked or in a different room (which I’d suggest) then that noise will not be a problem. Having the ability to monitor Stereo (Normal, Flip [L>R, R>L], Invert L, Invert R or Invert Both) or Mono (as L, R, L+R, L-R, R-L) is a wonderful feature. Everything I could imagine needing in a monitor controller is here. The only criticism is that the software volume control seemed far too coarse in response to mouse movements, unlike the smooth response of the stepped volume knob on the unit. The varispeed function worked perfectly, offering +-200 cents or over 12% up or down.

 

When it came time to listen, I went to the mastering room of renowned mastering engineer Glenn Meadows, formerly of Masterfonics and now with Mayfield Mastering, whose credits include Shania Twain, Taylor Swift, Steely Dan, Isaac Hayes, Vince Gill, Alabama and hundreds of others. We auditioned the Eclipse on his PMC IB-1 monitors with Bryston 7B ST power amps and a Sadie v6 DAW. We utilized an analog remix master tape (1/4” at 15 ips) sourced from an analog recording played back on a Studer machine, so there were no converters ahead of the playback source. This was perfect because visual analysis via CEDAR Retouch revealed sonic information in the analog recording up past 30 kHz, clearly beyond the cutoff filters in a 44.1/48 kHz converter.

 

Now for the ultimate question: How does the Eclipse sound?  My second question about the Eclipse (which was echoed by other engineers): “Why 384 kHz?” Does the recording world really need 384 kHz sampling? Many of us have hashed and rehashed the reasons for higher sampling rates for years and yet there still seems to be no consensus, apart from “We’re stuck with a 44.1 kHz release format (CD) and most people are listening to MP3s. So do we need 180 kHz of bandwidth in our recording format?” 

 

To answer that question, I did a simple test. Could I hear a difference between 44.1, 48, 96, 192 and 384 kHz? (Sample rates of 32, 88.2, 176.4 and 352.8 kHz are supported as well.) For me, there was an audible difference between 44.1 and 48 kHz, which is why, if given the option, I prefer to record at 48 kHz. The high end feels more relaxed and open, less restricted at 48 kHz. Next was 48 kHz vs. 96 kHz. Again, to me the difference was clear. When I say “clear,” I mean hearing a digitized sample that I can repeatedly identify with 100% accuracy in a blind test. Between 48 kHz and 96 kHz, it was not difficult to hear the difference, but to my surprise it had nothing to do with frequency response. To me, the imaging at 96 kHz was the decider. The image became more stable and instruments seemed to have a more precise location. At 48 kHz, the center image seemed to come closer, step forward, with more presence and a corresponding narrowing of the stereo field. At 96 kHz, the center image merged into the stereo field and everything felt more integrated, more of a cohesive stereo whole. I was fortunate to have three other engineers listen as well and everyone drew the same conclusion about the imaging. But that’s where the agreement ended. When asked for a preference, two of the four engineers actually preferred the presence of the 48 kHz conversion to the 96 kHz, Glenn had no clear preference and I strongly preferred the 96 kHz. It felt more natural to me. 

 

Next I compared 96 kHz to 192 kHz. Try as I might, I was unable to perceive a difference in imaging or frequency response or any other sonically significant measure. Then I compared 192 kHz to 384 kHz with the same results: no perceivable difference to me, remembering that this test was using a pre-recorded source as opposed to a live source. Switching back and forth between differing sampling frequencies, ranging from 32 kHz to 384 kHz and several in between, confirmed my first impressions. To make sure I was not pre-biased, I even let Glenn do the switching from A to B to A to B while I listened with my eyes closed. The blind test and my correct assessments of what I was hearing confirmed that I was not imagining things or biased by foreknowledge. 

 

At 48 kHz, I thought the Eclipse sounded very good. However, even at 96 kHz, when I A/B’d it with the source tape, I was still able to reliably discern which was which between the two. I have been auditioning converters for 25 years now hoping to find an AD/DA converter chain that will sound like the proverbial “straight wire,” where the output is indistinguishable from the input. I’ve listened to over 50 different converters in the intervening years, ranging in price from hundreds of dollars to tens of thousands, hoping to find the Holy Grail of converters. I’ve not found it yet.

 

The Antelope Eclipse sounds very good. It has a very smooth quality with wonderful integrity of imaging and just the slightest accent on the high end, almost sounding a bit “excited” in low-level detail. With the combination of clock, converter and monitor control in a single unit, it’s not easy to tell exactly which aspect is having the greatest impact on the sound, but we recorded the output of the Eclipse into SADIE at 96 kHz and played it back via a different DAC, different clock and monitor controller to confirm that I was hearing the effect of the Antelope ADC. I was.

 

I guess the ultimate question for an expensive, high-end piece of gear like this is one of value. If you consider it just as a mastering converter, its price is equivalent to the some of the world’s finest converters. At $6,995 street, that’s a lot for a stereo converter, but considering the added value of a monitor controller with talkback (which might cost $3K) and the clock (another $1.5K), then it seems much more reasonable. However, in this day when a fully functioning DAW studio can be assembled for under $6000, that outlay would be quite a stretch for most. Antelope’s main market is upscale mastering houses and recording studios and, even in our current troubled economy, there are those with caviar taste who can afford and appreciate a fine piece of equipment like this. I, for one, am glad there are companies like Antelope who are committed to building technology to such high standards and that there are those who can afford it. 

 

Lynn Fuston is the technical editor for PAR, an accomplished audio engineer and owner of 3D Audio Inc. 3daudioinc.com

 

Price: $6,995 street

 

Contact: Antelope Audio | antelopeaudio.com