As my acoustic preferences have evolved with a few years in this hobby, I have come to realize that your average PC makes for a pretty lousy source component.  It’s full of electrical noise and (for many) only has 1 output option.  The noise laden USB port.
 
I have been a USB “denier” for a long time.  I firmly believed that “bits are bits” and that it did not matter how the bits and bytes got where they needed to go.  It was the job of the DAC to deal with whatever crappy signal it was fed and convert the stream of ones and zeros into analog sweetness.
 
The Intona USB isolator changed my perception about USB “decrapifiers”.  (Thank you Schiit Audio for creating a lasting moniker for an entire range of products)  I have a DAC that just didn’t seem to live up to its potential using a “straight USB” connection.  Thanks to Tim @ Kitsune HiFi, I had a chance to evaluate an Intona.  It made a significant difference in my system.  I was impressed with the improvements in soundstage and clarity.  The veil that covered some recordings felt lifted. 
I was reminded of that famous quote, “The proof of the pudding is in the eating”.
A USB gadget that really makes an audible difference! 
 

 
What is it?
The Intona high-speed Isolator is nifty USB powered circuit board stuffed into an oversized butt ugly plastic box.   The isolator is connected to your PC via a standard USB A-to-B cable and sits between your computer and your DAC.  (or whatever USB device needs “decrapification”)  It provides galvanic isolation to the USB bus, repacketizes and provides clean power to your USB device.  It blocks the noise coming out of your computer by isolating the both the data and power.  The Isolator can operate at full USB bandwidth of 480MBit/s and works perfectly with  192kHz PCM and DSD content. 
 
There are two versions of the Intona USB Isolator – both do the exact same thing.  The “base” model retails for around $230.  The difference is that the ~$320 Industrial version is built to tighter tolerances with a “bill of materials” where component substitution is not permitted. 
 
The Intona was not designed for audio use, but quickly gained acceptance among audiophiles.  It was designed for industrial and laboratory applications requiring a low noise floor and low jitter.
 
I was particularly intrigued by the photo below.  You can see how the two PCB boards (input and output) are completely separate.  They are only joined by the digital isolator chips.

 
 
Feature List
According to the manufacturers website, the Intona provides the following features:
 

1. Provides galvanic isolation to the USB bus
2. Compliant to USB 2.0 Hi-Speed (480 MBit/s)
3. Isolation withstands up to 2.5kVRMS
4. Works out of the box, no drivers needed
5. Transparent to host and device
6. No additional round-trip latency
7. Powered by USB bus, no separate power supply
8. Ultra-quiet power rail isolation, linear regulation

 
While browsing the Intona frequently asked questions; I found two interesting graphs that compare a 1k signal with and without the Intona isolator.  You can see that the noise floor is reduced dramatically. 
Of particular interest is how the 8kHz spike commonly associated with packet noise is suppressed.  That makes an audible difference and is one of the metrics for audiophiles. 
 
 
Installation
Since the Intona does not require an external power supply and does not require a device driver on your PC - installation is about as easy as it gets.  It is truly plug-and-play.
 
The Intona will require two USB cables; a standard A-to-B cable from your computer to the Intona and another cable from the Intona to your USB device.  The “dirty side” cable from your computer to the input of the Intona can be any generic USB cable of an appropriate length.  I am currently using a 6” Belkin USB “daisy chain” cable to from my laptop to the Intona.  It was less than $5 from Amazon.

On the “clean side” of the Intona, I use a quality AudioQuest Carbon http://audioquest.com/usb-digital-audio/carbon USB cable from the Intona’s output to the input of my DAC.  Many Intona users keep the “clean side” cable as short as possible to minimize the chances of stray RF noise getting picked up enroute to their USB device. 
 
How Does It Sound?
To test the benefits of the Intona I queued up a couple of familiar tracks.  I plugged my DAC into the computer with nothing else in the signal path.  Then I compared the same track played through the Intona.
 
First up was Alison Krauss’s  “Forget About It”.  Without the Intona the delicate opening guitar work is present, but felt a little distant.  The sound of the shaker was the same from one beat to the next.  Alison’s vocals were warm, but weren’t particularly separate from the rest of the band.
 
Plugging the Intona into the signal path made a significant difference.  There was a little more air in Alison’s vocals.  The soundstage widened and I got a better mental image of the placement of guitars, piano and percussion.  Detail was certainly improved.  The sound of the shaker was distinct – not monotonous.  Cymbals started to shimmer and decay into the ether correctly.  
 
I think this is a good point in the review to remind readers that these changes aren’t huge night/day differences.  One can certainly enjoy the music without an Intona.  However, as my musical preferences have evolved, my ears are keen to notice the improvements in timbre and soundstage that the Intona brought out of my system.  To me, it’s the difference between a good sounding system and a great sounding system. 
 
Rebecca Pidgeon’s performance of “You Got Me” is one of my favorite test tracks; not just for Rebecca’s amazing voice, but for the two acoustic guitars and simple percussion.  I feel that Dr. Chesky’s recordings are among some of the best mastered albums and sound amazing on a quality HeadFi system.  Without the Intona I found the acoustic guitar on the left to lack a little detail.  With the Intona plugged in, the detail that I felt was missing emerged.  The background vocals now stood apart from Rebecca’s voice – a great example of detail and separation. 
 
One of the most common USB complaints is “packet noise” which is centered around the 8kHz frequency.  I hear this anomaly in the reproduction of cymbals.  I’ve characterized this as hearing cymbals that “crash” instead of “shimmer” – where roll-off and decay sound unnatural.  I have used the Queen/David Bowie hit “Under Pressure” to test reproduction of cymbals.  I have four different versions of this song.  To me; the best sounding version comes from the 2011 remastered Greatest Hits album.  I found the Intona to really shine on this track as the decay of the cymbals felt much more realistic.
 
Your Mileage May Vary
So will the Intona work wonders on your system?  That depends on the quality of your USB source and the capabilities of your USB device.
 
I recently reviewed two DACs.  The Benchmark DAC3 and the Holo Spring DAC.  Both DAC’s were connected to my 4 year old Windoze 10 laptop.  The Benchmark DAC3 sounded amazing right out of the box.  The Holo Spring sounded “good”, but it was pretty clear to me that something was missing.  The Intona provided the extra clarity that helped me get the most out of the Holo.  I was left with the impression that the USB implementation in the Benchmark did a better job handling a poor quality, noisy USB signal from my laptop.  The Intona helped up the Holo’s game. 
 
Will an Intona make a good DAC sound better?  Maybe. 
Will an Intona make a bad DAC sound better?  Doubtful. 
Will make up on an ugly girl make her beautiful?  No
 
In my experience with the Intona, I found that it improved the soundstage on most recordings.  It improved separation between instruments and vocals.  The timbre (not timber) of acoustic guitars came through where it had been lost.  Cymbals shimmered and decayed correctly instead of just crashing.
Overall, I was very happy with the improvements that the Intona brought to my system.
 
Will you get the same benefits with an Intona in your system?  Maybe.  Maybe Not.  That will depend a lot on the quality of your source and the quality of the downstream components.  Some components will exhibit more noticeable improvements than others.  YMMV. 
That said, I would recommend that you try one out for yourself. 
 
I think the Intona can be an important tweak to help wring additional performance from your USB system.  

This is a pretty long review.  For the benefit of those with short attention spans, let’s cut right to the chase.
 

1. The “entry level” Holo Spring DAC is good – very good!  It is transparent, detailed and handled just about every format I could throw at it. 
   With the level 1 DAC retailing for $900 less than the level 3; I think it’s a bargain.
2. I found that the Level 1 Holo performs exceptionally well; especially in the non-oversampling mode (NOS). 
   I haven’t found a case where I preferred any of the oversampling modes to NOS mode. 
3. I found the performance of USB input somewhat lackluster. 
   The I²S input performed significantly better with the help of a Singxer SU-1.

 
 
 
Features
The Holo Spring is a very interesting DAC.  Its primary draw is that it is an R2R DAC with a non-oversampling mode. 
 
The Spring uses a patented linear compensated R2R network.  Not being an engineer, I understand this to mean that the Spring uses two R2R “ladders”.  The second R2R ladder essentially provides error correction or “compensation” for resistor tolerance in the first.  This results in a very precise and linear output.  The Spring has one R2R network for PCM and a completely separate R2R network for DSD.  The R2R circuit board has the same resistor ladder on the top and the bottom as illustrated in the pictures below.
 
  
 
The 3x oversampling modes are enabled using an AK4137 sample rate converter. 
In NOS mode, this circuit is bypassed completely. 
 
The Spring offers RCA and XLR outputs and a full complement of inputs including: USB, Optical, SPDIF, AES and I²S. 

 
The coax, optical and AES inputs can handle PCM 24/192kHz and DSD64 content.
The USB and I²S input are capable to handling higher PCM at rates up to 384khz and up to quad-rate DSD512.
 
The I²S input is particularly interesting.  This input bypasses the internal AK4118A receiver and pretty much sends the signal straight to the digital/analog conversion.  This is one of the reasons that so many Holo owners have paired their Holo DACs with the Singxer SU-1 USB bridge.
 
Finding the Right Level
The Spring DAC is available in 3-different configurations; Level 1 through 3.  There is a $900 difference in retail pricing between the “entry level” Spring DAC (being reviewed here) and fully pimped-out version.  The “core” of the DAC – the R2R network – is identical across all levels.  To help me understand the different configurations, I created a chart of the upgraded components in each level and the corresponding price increase. 
 
 
Being a man of modest means (a polite synonym for “cheap b*stard”)  I opted for the Level 1 DAC.
 
I did not know if I could cost justify the small differences that the upgraded components would make.  …but primarily – I’m cheap! 
 
Delivery
I placed my order with Kitsune HiFi here in the US and the DAC shipped directly from China.  If you are unfamiliar with this process know that shipping takes a little longer than Amazon Prime; you will need some patience.  In my case, I was pleasantly surprised to have my unit delivered in just 2 weeks.  Not bad at all!
 
The DAC came “double-boxed”.  The outer box arrived fully wrapped in orange shipping tape.  The inner box was unlabeled and devoid of any branding.  Inside the DAC was wrapped in a plastic bag and securely held in place with sturdy molded foam.  There was no manual, no power cord, no cabling - just a DAC in a bag. 
 
Out-of-the-box Experience
After removing the DAC from the packaging, the first thing I noticed was the casework.  The chassis is very solid and is deceptively heavy.  The copper colored side panels are interesting and attractive, but invisible in the location where I put my DAC.  Overall, I am very impressed with the quality of the chassis. 
 
Since the Spring DAC doesn’t ship with a power cord, I sourced a Shunyata Venom 3 (~$125) to connect the DAC to my PS Audio P3 Power Plant AC regenerator.
 
…oh, and that Singxer SU-1 I mentioned - it doesn’t ship with a power cable either.  If you don’t already own all of the necessary cabling – both signal and power - you will want to budget accordingly and plan to see a lot of your friendly UPS driver. 
 
The Spring’s power switch is on the rear of the chassis.  That might not be a problem depending on your installation.  Since my DAC lives on a shelf in my office, the power switch is completely inaccessible.  Therefore the Spring stays powered-on 24x7 and I use the standby mode when not in use.
 
I connected the DAC to my HeadAmp GS-X mk2 amplifier with a pair of WireWorld XLR interconnects and connected the DAC to my laptop with a non-descript USB cable.
 
Usability & First Impressions
Whenever I get new gear, I like to pay special attention to my initial impressions.  Over time you can get used to a product’s unique quirks.  I like to capture those first impressions before they fade into familiarity.
 
There aren't many features on the front of the Spring DAC.  You have a giant a digital display and four attractive copper colored buttons.  The switches engage positively but noisily. 
 
The display is large and very bright.  If you live in an apartment complex your neighbors will be able to see if you’re playing “44.1K” or “DSD64” content.  …yeah, it’s that big. 
 
The “Display Intensity” button steps through several brightness levels with each press.  At first I didn’t think the button worked at all.  In reality there is very little brightness difference between the first several button presses.  I counted 16 different illumination levels between full bright and off.  That’s right – if you want to shut off the overly bright display – press the button 16 times!  You will find yourself hitting the display button with all the speed and fury of a 14-year old banging on a Playstation controller.
 
Pressing the “over sampling” button naturally cycles the DAC through all four of its sampling modes.   The mode flashes on the display for about 5 seconds and then returns to displaying the sample rate.  …and that can be a bit of problem.  There is no indication on the display unit telling you what sampling mode the DAC is currently operating in.  Did you leave the unit in “NOS” mode 2 hours ago or was that “OS PCM” mode?  Now you have to cycle through all four modes to make sure the correct mode is selected.  I wish the first press of the over sample button would display current mode instead of changing it.
 
Since we’re on the topic of usability, I would really like to see a power lamp on the front panel.  It would be a nice touch.  If you completely dim the display, there is no indication that the unit is powered up. 
 
Initial Listening
I began my critical listening tests trying to assess straight USB performance.  I found that playback was good but in a word - lackluster.  Most notably, I found a fairly narrow soundstage.  Cymbals crashed instead of shimmered.  Crowd noise in live recordings like Eric Clapton’s Unplugged album felt two dimensional. The acoustic guitar in R.E.M.’s  “Drive” from “Automatic for the People” just didn’t feel realistic.  Michael Stipe’s voice sounded a little wooden.  It was pretty clear that straight USB was not going to cut it.  …at least not in my system.  I could tell that I was leaving something on the table. 
 
If you stopped reading this review at this point you would be left with an incorrect negative perception of Level 1 Spring DAC. 
 
Improving Performance
Undeterred, I reached out to Tim @bimmer100 at Kitsune HiFi for his thoughts.  Tim graciously agreed to send me a Singxer SU-1 and an Intona USB isolator to evaluate. 
(crack dealers employ the same business strategy.  “Hey kid, try a hit of this new DDC convertor and USB isolator”.  Next thing you know you’re slinking around the Kitsune website at o-dark-thirty trying to get another Head-Fi hit.) 
 
Each device made a significant improvement to the sound of the DAC.  Both the Intona and the Spring broadened the soundstage and increased detail.  Used together, the veil that I felt covered the music was finally lifted. 
 
I think it is important to note that the majority of the lackluster USB performance I encountered is due to the quality of my source.  A 4-year old Windoze 10 laptop isn’t exactly an audiophile grade source.  Could the Spring’s USB implementation be improved?  Most certainly.  By way of comparison, I recently reviewed a Benchmark DAC3 using the same laptop source.  The Benchmark dealt with the whatever nastiness came out of my USB port and performed admirably whereas the Spring struggled.  I would wager an uneducated guess that Benchmark’s USB implementation is superior to the Holo’s.
 
I believe that a different source, say a quality CD transport, streaming server or a MicroRendu, would have yielded different results.  Being a cheap bast*rd, I did not have an alternative source with SPDIF or AES output to compare USB performance against.  I had to make the most of my PC source.  The combination of an Intona (~$300) to clean up my USB signal and a Singxer SU-1 ($400) to convert it to I²S format worked for me.  …so much for my so called budget.
 
I was not thrilled at having to throw an additional ~$700 at a $1,700 DAC.  It certainly felt like throwing money at a problem.  (This is headfi after all – it’s what we do!)  That being said, the combination of supporting hardware seems to have been the right solution for my personal setup.  Whether it made economic sense or not is entirely debatable.
 
Do you need a Singxer SU-1 and a USB decrapifier to get the most out of your Holo?  That will depend on the quality of your source.  YMMV. 
 
DSD Playback
Over the years I have dabbled with the SACD format and have collected a fair number of DSD encoded titles.  Therefore, DSD playback and performance was an important consideration in my DAC choice.  It’s the primary reason that I did not buy the well regarded Schiit Yggdrasil.
 
The Holo Spring provides native support for DSD.  That means that you do not need to use “DSD-Over-PCM” or “DoP”.  When describing DoP; the analogy I like to use is that DoP is like stuffing a DSD “letter” into a PCM “envelope”.  The receiving device has to unpack the PCM stream to extract the original DSD content.  You can find a more detailed explanation here.
 
The benefit of native DSD is that your media player can skip the resource intensive process of packaging DSD into a PCM stream. 
 
Since DSD playback was a key feature for me, I was keenly interested in the Spring’s dual R2R network.  Many DACs handle DSD by first converting the DSD signal into a PCM stream before performing the digital-to-analog conversion.  The Spring DAC is unique in that it has a completely separate R2R network exclusively for DSD playback.  DSD is never converted to PCM - they are kept separate.
The Ghostbusters said it best – “Don’t cross the streams”.
 
I played DSD64 encoded tracks from Pink Floyd, Queen, Peter Gabriel, Dire Straits, Michael Jackson and David Bowie.  (guess who grew up in the 80’s?)  I have been very pleased with all of the DSD content I’ve tried.  Nothing has sounded artificial or unnatural to my ears. 
 
I found one notable, irritating quirk during DSD playback.  There is an audible click present at the start of most DSD encoded tracks.  This noise is also present when one DSD track ends and another begins.  I did not notice this behavior during PCM playback.  It’s more of annoyance than anything.
 
Oversampling Modes
If I were at a Head-Fi meet and someone asked me to demonstrate the difference that the oversampling modes make; I would queue up the first minute of Alison Krauss & Union Station’s track “Let Me Touch You For A While”. 
 
I found that in OS PCM mode; Alison’s voice had a certain unrealistic wooden edge to it.  Switching to NOS mode seemed to lift the veil (to my ears at least) from her vocals.  I found that same characteristic present in Rebecca Pidgeon’s “You Got Me”. The acoustic guitar in this track seems to present an unnatural resonance in OS PCM mode that is absent in the NOS mode.  I lack the eloquent vocabulary to describe it better than my inner redneck who just says “that ain’t right son”. 
 
After much listening and experimentation with the other modes, I am thoroughly convinced that I prefer the non-oversampling mode.  None of the oversampling modes did anything for me. 
 
More Listening Impressions
Having finally cleaned up my USB signal path and hidden the AMEX bill from my spouse, I was ready for some more critical listening.
 
One of my favorite test albums is Doug MacLeod’s “Exactly Like This” published by Reference Recordings.  The album is available on DVD as 176.4kHz/24bit uncompressed WAV files.   
 
On the track “Serious Doin’ Woman”  you can hear Doug’s chair creaking while he plays.  There are a couple of finger snaps early in the track that have that realistic, fleshy sound.  The Spring does an admirable job of rendering minute details like that. 
 
I have always been a big fan of live performances.  I am captivated by the sound of the crowd, the musicians talking between songs and overall “imperfection” that is a live performance.  The first track from Eric Clapton’s Unplugged album is called Signe.  The Spring is able to present a wide soundstage on this track giving the giving the crowd noise a realistic 3D quality.  The highs ring true and roll off naturally and the guitars are upfront and rich.  It’s everything that you want from a live performance – this DAC takes you there.
 
Macy Gray’s “I Tried” from her Stripped album on HDTracks is another great track for evaluating a headphone system.  It is uses Dr. Chesky’s binaural recording process.  The Spring delivered an excellent sense of realism; especially from the double bass.
 
Conclusions
The entry level Spring DAC works well, but in a narrow capacity.  The consensus of Spring owners here on HeadFi is that the NOS mode sounds best.  After a few weeks with my Spring and plenty of my own critical listening, I completely agree.  It’s the only mode that I use now.
 
If the Holo Spring won an Oscar for best actor, then the Singxer SU-1 would surely win the award for best supporting actor.  I found that the SU-1 helped take the Holo’s performance up a notch to reveal its true capabilities.
 
Do I feel like I “left something on the table” by choosing the entry level Spring over the fully pimped out level 3 DAC?  In a word, “no”.  In my particular system, I found that I got a significant performance boost by improving the quality of the USB signal path.  I do not think that the upgraded capacitors and hand wound silver transformer in the level 3 would have made anywhere near as much difference as cleaning up my USB signal path.  The $900 cost “savings” of the level 1 vs level 3 provided the budget for a DDC convertor, USB filtering and new cabling.  Had I “sprung” for costlier Spring (I purposely saved that bad pun for the end of the review), I would have had the same experience with sub-optimal USB performance and still found myself in need of a better source.
 
My personal recommendation is that one should have everything “in front of and behind the DAC” completely locked down before investing in a level 3 Holo Spring.  The level 1 DAC is an admirable performer.  I found that I could wring a lot more performance out of the entry level Holo Spring DAC by tweaking my USB signal path.  

I’ll cut right to the chase:
The BenchMark DAC 3 HGC is a sonic marvel and an engineer’s wet dream.
 
Overview & Comparison
The Benchmark DAC3 is based on the ESS Technologies ES9028PRO chipset and is the younger brother to Benchmark’s well regarded ES9018 based DAC2.  ES9018 based DAC’s have been common for about 10 years now.  The ES9028PRO is still relatively new (as of Q4 2016) with just a few commercially available implementations.  (Notable implementations include the Exasound E32 DAC at  ~$3,500USD, the Resonessence Veritas at ~$2,900USD and the Gustard X20 Pro at ~$900USD)
I was eager to audition the DAC3 to see what kind of magic Benchmark could wring from ESS Technology’s latest chip. 
 
Quoting directly from Benchmark:

The DAC3 builds upon Benchmark’s highly successful DAC2 product family. The DAC3 maintains the familiar DAC2 form factor, but adds the higher performance available from the new ES9028PRO D/A converter. The DAC3 offers the following improvements over the DAC2:
• Active 2nd Harmonic Compensation
• Active 3rd Harmonic Compensation
• Lower THD
• Lower passband ripple
• Improved frequency response
• Increased Dynamic Range
• Faster PLL lock times
• Faster switching between inputs

Click to expand...

 
In short – it's everything you like about the DAC2 only mo’ better.
 
The Benchmark DAC3 is available in two different configurations:

1. DAC3 L – This model offers analog only output, without headphone amplification and retails for $1,895
2. DAC3 HGC – An extra $300 ($2,195) gets you the same DAC plus an integrated HPA2 headphone amplifier. 

 
This review focuses on the DAC3 HGC model with the integrated HPA2 headphone amplifier.
 
My goal was to evaluate the DAC3 HGC serving in two roles – both as a standalone DAC and as an integrated all-in-on DAC/AMP combo.
 
The older Benchmark DAC2 and the new DAC3 are practically indistinguishable visually.   They use the same chassis/casework, but with different labels.  They have the same controls, features and operation.  The HPA2 amplifier section is the same design that debuted in the DAC1 circa 2009.   The underlying ES9028PRO chipset is the big difference in the new DAC3 – everything else is pretty much the same.  …even the USB drivers are the same.  (Though it is a bit odd to have a DAC3 on your desk and drivers & dialog boxes that say a DAC2 is connected.)
 
The front of the DAC3 HGC offers dual ¼” single-ended headphone outputs.
Initially, the absence of a balanced output dismayed me.  I have a balanced Moon Silver Dragon cable for my Audeze LCD-3’s that I’m rather found of.  John Siau, Benchmark's VP & Director of Engineering, makes a compelling argument why you won’t find balanced headphone connectors on their products.  You can read it here.  After reading his article I better understood the approach he used in Benchmark’s HPA2 amplifier design. 
 
Sound
I’m not one for burning in new components for 100 hours – I like to tear open boxes in a fit of instant gratification.   I unboxed the DAC, tossed the manual and remote into the corner and connected my laptop to the USB input.  (Note that ALL of my testing was done via the USB input, I did not use the other digital inputs or the analog inputs.)  After configuring the correct USB mode (more on that later) and swapping to the OEM single-ended cable on my LCD-3’s I jumped into my evaluation playlist. 
 
I don’t go for the typical audiophile fare of classical, chamber music or slow jazz.  I’m more into guitars and drums. 
With that in mind, I started my evaluation with the Black Keys’ El Camino album and their track “Sister”.
The big bass drum kick that anchors this track has left me disappointed on other systems and pairings.  The DAC3 HGC nailed it and kicked with the authority that I've often found lacking.
 
Immediately impressed with the “grunt” of the HPA2 amp section, I moved on to my all-time favorite Red Hot Chili Peppers song – their cover of Stevie Wonder’s “Higher Ground”.  I love Flea’s bass guitar in this track. The benchmark had me toe tapping to the “chicka-wow-wow” rhythm. 
 
I have a special affinity for acoustic tracks, live recordings and concert performances.  I want my head-fi gear to transport me to the show and make me feel like I’m the crowd.  I like to hear the ambiance of the venue and the crowd noise in live recordings.  John Mayer’s performance of “Vulters” live in Los Angeles is a great example of what I want in a live track.  
The DAC3 HGC gave me the wide soundstage you would expect from a concert recording.  The vocals were tight and centered and the crowd noise wide. 
 
I have some Florence + The Machine tracks that sounded dull and lifeless on other equipment, but popped on the DAC3 with new clarity and energy.  You know you’re onto something special when a piece of gear can breathe new life into flat tracks. 
 
Switching from the all-in-one role, I next evaluated the DAC3 as a standalone DAC; bypassing the HPA2 amplifier section.   I put the Benchmark in home theater mode and connected it to my HeadAmp GSX-mk2 amplifier.  This also gave me an opportunity to compare the Benchmark’s DAC against another DAC head-to-head (albeit with some cable swapping). 
 
I queued up one of Dr. Chesky’s evaluation discs and got to listening.  Perhaps the most interesting test (to me at least) was Rebeca Pidgeon’s cover of “Spanish Harlem”.  Each shake of the shaker was indeed different.  The DAC3 did an excellent job of resolving minute details in the material.  Compared back-to-back against the DAC section of the ES9018 based Oppo HA-1, the Benchmark nailed it.   It wasn’t one of those subtle differences that makes you go “hmmm” – it was a “wow” moment. 
 
I had a similar experience with the “Brushes & Brass” track on Chesky’s Audiophile test disc.  There’s some subtle vocalization in that track that had me looking over my shoulder to see who was talking or wondering if I left a TV on in another room.  It was very disconcerting the first time I heard it.  Played through the Oppo HA-1’s DAC the sound was more or less two-dimensional.  Played through the DAC3, I got a completely different sense of space and depth.  I really felt like someone was in the room with me. 
 
Another favorite test track of mine is John Hammond’s “Get Behind the Mule” from his Rough & Tough album.  It’s a simple track consisting of vocals, guitar and foot stomping.   The DAC3 was startling in its clarity and realism.  With such a simple sound you can really focus on the details of Hammond’s guitar work.  The DAC3 did not disappoint!
 
DSD Support
I have some great tracks in my collection that are in DSD format.  So Benchmark’s support for DSD was “must have” feature for me.  It’s one of the reasons that I had to rule out the similarly priced and well regarded Schiit Yggdrasil which does not support DSD in any flavor. 
 
The Benchmark DAC3 supports the most common DSD format - DSD64 (2.8224 MHz) - which is sufficient to meet my needs. 
It is interesting to note that other DAC’s based on the same ES9028PRO chip support higher rate DSD formats. 
The Gustard X20 supports “Double-Rate” DSD128 (5.6448 Mhz) and the Exasound e32 supports “Quad-Rate” DSD256 (11.2 Mhz).  
While these higher rates are interesting, the only tracks I have in these higher resolutions are demo and evaluation recordings.  I don't have any commercial music in these higher rates.
 
Out-of-the-box, the DAC3 uses USB 1.1 and is "driverless", but limited to just 24/96kHz playback.  (Does anybody still use USB 1.1?)   You cannot play DSD material in this mode.  You need to install the USB 2.0 driver to enable playback of 24/192kHz PCM and DSD streams (in DoP 1.1 format).  
 
Here's where I started to struggle with the DAC3.  I had to break out the manual to find out that you have to press two button simultaneously and wait for some lights to blink to enable USB 2.0 mode.  The first time I tried to do this, I managed to screw it up because my hand covered the indicator lights.  Once I had the unit in the right mode I could play high bit rate PCM music.  Getting DSD to work right required some tinkering in my media player, JRiver Media Center 22.
 
Usablility
I’d like to explain why I chose to characterize the DAC3 as an “engineer’s wet dream”.  
 
A full 16 pages of the 71 page manual are dedicated to performance charts & graphs that only an engineer would love that scientifically prove beyond any shadow of a doubt that you’ve purchased an engineering masterpiece.   The bulk of the remaining 55 pages school you on unnecessarily complex tasks like changing the gain settings.
 
A word of advice for DAC3 buyers – DO NOT LOSE YOUR MANUAL!  You are going to need it – often.
 
As good as the DAC3 HGC sounds, I have a have a couple of gripes that (IMO) detract from its usability as an all-in-one DAC/AMP combo. 
This is a key point to remember.  My gripes about the DAC3 HGC are not about its sound quality.  It's about trying to live with it as an all-in-one AMP/DAC combo.  
 
Gripe #1 – Jumpers
Do you like taking things apart?  Do you like tiny screws?  If so, you’re in luck with the DAC3! 
The top casework is fastened to the chassis with eight screws and you’ll have frequent reasons to remove as four different features of the DAC3 HGC are controlled by jumpers on the circuit board. 

1. Headphone gain – 2x jumpers
2. XLR Gain – 4x jumpers
3. Headphone switch – 2x jumpers
4. Digital Pass Through – 1 jumper

 
I’ll be blunt, at a price point north of $2k I don’t expect to have to break out a screwdriver and fat-finger 9-different jumpers. 
 
If you have multiple headphones or widely varied sources unplugging the unit, unscrewing the case to fiddle with jumpers is a complete pain in the *ss. 
 
Gripe #2 – Overloaded Buttons
The DAC3 has 5 buttons on the front panel; many of which serve dual purposes.

1. Want to toggle between USB 1.1 and USB 2.0 modes?
   Press & hold both input buttons simultaneously for 3x seconds
2. Want to activate the HT mode?  Don’t push the button nearest to the HT indicator lamp. 
   Nope, hold the MUTE button for 3x seconds.  Intuitive huh? 
3. External 12V trigger can be defeated by holding the power button 3 seconds.

 
Gripe #3 – Overloaded Lamps
I freely admit that I am old-school geek.  I love blinkly lamps and indicator lights.  I’d be perfectly happy if my car’s dashboard looked like a jet cockpit.  If you want to verify the sample rate of the input signal to the DAC3 you’re going to need to do a little basic math on the fly.  Sample rate is indicated by multiplying the value of the DSD indicator and word-length indicator values.  Huh?

1. 192kHz playback is indicated when the DSD 4x lamp and the 48 lamp are both illuminated because 48 x 4 = 192.
2. 176.4kHz playback is indicated when the DSD 4x lamp and the 44 lamp are both illuminated because 44 x 4 = 176
3. 96kHz is 2 x 48 and 88.2kHz playback is 2 x 44 – you get the idea

 
Can someone explain to me why I am I looking at the lamp labeled DSD when my source material is PCM?  That’s like clicking the old “start” button in Windows to shut down your PC.  It’s counterintuitive.  One would think the DSD lamps would be extinguished during PCM playback, but instead they serve a dual purpose.
 
Used in a studio setting, these switches might be set once and forgotten about.  Used as an all-in-one device in a home stereo setting, these inconveniences quickly grow tiresome.
 
The DAC3 could really use a digital display like the Oppo HA-1 which retails for nearly half the DAC3’s price.  Heck – even the $900 Gustard X20 has a good digital display.  One would hope that a “3^rd generation” product like the DAC3 would have evolved in usability with each generation.
 
Another usability quirk with the Benchmark is the 2x buttons that scroll through the inputs.  It seems like every other button has multiple jobs, yet we have one task, input selection, that requires two buttons.  One to go “up” the list, the other to go “down” the list.  Many devices either have a rotary switch, or uses a single button that cycles through all of the available inputs.   When the “up” or “down” button hits the end of its range, it doesn’t cycle back around to the first or last input.  …not a true gripe, just an observation of a unique quirk. 
 
Give me an DAC/AMP combo with the Benchmark’s sonic quality and the Oppo HA-1’s features and flexibility and it's “shut up and take my money” time. 
 
Summary
After living with the DAC3 HGC for a few weeks I can say that I prefer the DAC3 serving in the role of “stand alone” DAC.  I think it’s a fantastic sounding D/A convertor.  On well recorded material it does an outstanding job with detail and resolution.  It brought new life to CD quality 16/44.1 material - even on tracks that weren’t particularly well mastered. 
 
However, as an all-in-one AMP/DAC combo I think the DAC3 HGC falls short.   I’m not interested in grabbing a screwdriver and the owners manual to change something as common as gain. 
 
I found the integrated HPA2 amplifier in the HGC model to be very good, but not stellar.  it’s not that the HPA2 amp section does anything bad, it’s just that my external amplifier did everything better.  (For ~ $3k one would expect nothing less!)   Put another way – if you have $2,200 to spend on a DAC you probably already own a “reference class” amplifier capable of outperforming the $300 HPA2 option found in the DAC3 HGC. 
 
I think the DAC3 is a worthy and capable standalone D/A convertor.  I would pass on the HGC model with its integrated amp.
 
Footnote
I want to take a moment to personally thank Benchmark’s Sales Manager Rory for all of his help.  He quickly answered all of my questions before and AFTER the sale providing detailed answers to all of my questions whether by phone or by email.  He also made sure that the unit I received had the latest & greatest firmware.  Thanks again for the great customer service!

As part of the OPPO PM3 loaner program I had a chance to evaluate the PM3's for a few days.
 
Overall, I was very impressed with the packaging, build quality and finish.  The brushed aluminum is exquisite and the earpads & headband are soft & supple.  They fit my basketball-sized skull quite well.   
 
The cable is sturdy and terminated with nice accents - great attention to detail.  The phono plug adapter fits well and provides a solid connection when using a desktop amplifier.
 
This headphone is geared towards high-end mobile play and I think it does well in that regard.  However, that's not quite the setup and listening environment I use.  I have an Oppo HA-1 DAC/Amp combo to which I stream high resolution audio from my PC over a USB connection.  I don't expect anyone to buy these headphones to pair with a $1,200 amplifier.  However it does provide an interesting way to drive them harder and with better resolution than you can with your phone.    
 
One of my favorite test tracks is "Tied Up" by Yello.  It's a brutal track, with deep lows, shimmering highs and lots of sharp detail.   I was immediately impressed with the PM3's bass response.  Next up was the track "Lies" on the same disc.  I was missing that wide soundstage - it felt closed in - the air was gone.
 
I switched gears and queued up The Black Keys "Stackshot Billy".  I felt like the PM3 completely fell apart on this track.  It was as if the guitar and vocals got tossed into a blender and frappéd into an audio slush - zero separation.  The cymbals were practically non-existent and the guitar work butchered.  For me - this was the worst track I played on the PM3. 
 
I switched to something I thought would be better suited to the PM3 - The Talking Heads live performance of "Psycho Killer" from their Stop Making Sense album (one of my favorite Talking Heads tracks).  The PM3's kept pace with the bottom end, but the guitar lacked detail and brilliance.  "Heaven" is the next track on that album and the PM3's just weren't nailing that live guitar sound.  
 
I dug up two tracks to test vocal playback.  Norah Jones performing "Don't Know Why" and Queen's "Somebody to Love".  Norah Jones' vocals sounded good, but that open airy ambiance I've come to expect was gone.  Freddy Mercury's vocals were strong, but somewhat muzzled.  The bells on "Bicycle Race" were just flat wrong.   The PM3's did a better job with the darker and more aggressive "Another One Bites the Dust".
 
So overall, I can say that the PM3's are extremely well built and comfortable.  Sonically, they're just not my cup of tea.  I found them lacking in soundstage, presence and detail.