Empirical Audio Overdrive Diamond SX with Ethernet to I2S Converter

0 Members and 1 Guest are viewing this topic. Read 3649 times.

ketcham

  • Industry Participant
  • Posts: 285
Greetings Everyone. 

I wanted to update you regarding Steve Nugent's  Empirical Overdrive Diamond SX with the optional internal galvanic isolated Ethernet direct I2S converter cpu.  I have run the unit 24/7 for over 1000 hours and I must admit the caps needed this time to break in.  This will be the first dac that I felt needed over 6 weeks of almost continuous play to break in.  However, this is one instance where patience paid dividends. 

Since we all have our own perception and certainly favor different qualities, I am not going to delve into a lengthy review.  The qualities that stand out and I believe most would agree, the dac is well balanced on all fronts and certainly demonstrates design implementation is more important than the dac chip itself.  Historically, I have been approached by others who felt the ODSE was overly analytical.  I humbly disagreed as I found it to be detailed, accurate and perhaps unforgiving.  While I found the SE emotionally engaging, the SX is addictive.  The more dacs I audition, I find myself with a preference to non-oversampling design implementation.  My goal is to compare this unit to a R2R ladder dac, Totaldac, and will report my results.  At 4x the price, it would have to engage me on an entirely different level.  Steve's dac chip is proprietary and I feel discussing design theory should be left to the electrical engineers.

At first blush, I had found little difference between the ODSE and the ODSX.  Over time the Diamond SX bloomed, creating an even more precise presentation of the music.  Well-recorded intimate recordings immersed me into a room of musicians.  Big venue live recordings placed me back from the stage, closer to the soundboard, creating a deep and viable soundstage.  Both the ODSE and the ODSX have been the best interpolating PCM data and recreating an accurate presentation of the recording, especially piano, wind instruments, and particularly the mid and upper range of percussion decay.  I still hold to my initial assessment years ago.  I have yet to find a DSD dac that performs better than Steve’s implementation of PCM.  However, I have not heard a ground up designed DSD only (non PCM) dac.

The chassis is no different that the ODSE.  The faceplate now states SX in place of SE and the knob has a very subtle recessed diamond.  Nice touch but for those who upgrade their SE to SX, I would keep your old faceplate and save money.  The rear panel is the same.  Mine had a manual cutout where the Ethernet input replaced the USB.  This does bring me to another point:  For those of you who will upgrade, the ODSE/X run hot.  The morphology of the circuit boards change and this will affect the soldering.  I highly recommend you pay the extra nominal fee and use a virgin board.   Longevity is essential.  While I am sure Steve will service these units well into retirement, save yourself the headache.

Those of us who owned these units before know these units run hot.  While breaking in, I ran the unit in my office on my desk, which is quartz composite.  The unit ran hot, up to 208* over the Ethernet board.  Steve assured me this is not an issue.  When I placed in my system on Bubinga wood, the temperature dropped by 50* while being relocated in direct sunlight and reset to high-output line out.  Surfaces matter, ventilation matters.  I have run this dac hard without failures on any level. 

The Ethernet DNLA/UPnP is simpler, more robust, and strait forward.  Unlike USB, Ethernet packets of data are checked for errors and packets with errors are resent.  The more I read about USB implementation, more validating why I despised it and found it universally inferior to AES/EBU.  ROON uses its own RAAT protocol in place of DNLA/UPnP.  The third party company which designed the Ethernet to I2S converter CPU chip has no interest in RAAT.  ROON states they will implement DNLA but with no timeline.  Because of this current incompatibility, I have not pursued ROON at this juncture.  One manufacture who uses this CPU has figured a work around and was able to implement ROON RAAT.  However this was not to an I2S conversion.  Through Bubble UPnP and Linn Kazoo (both free) I am able to stream Tidal without any issue.  Linn Kazoo is not on the same league as ROON or even Tidal’s own controller.  To say it plainly, the user interface sucks.

There are many third party controllers.  Some are free, while some are cheap to purchase.  They all work well, and I give credit to this third party manufacturer.  Not all Ethernet dacs/Renderers have this level of performance.  Please note:  I find those controllers that send a file direct from NAS to the DAC without processing sound better.  Those who have interest in this topic please PM me for more detail and discussion.  I would love to learn where others have found success and discovered absolute ‘do not’s’.

I wish to commend Steve on a well designed dac.  Thank you for reigniting my passion for this hobby.  This last year has been somewhat frustrating for me.  High-end components do not necessarily play well together and this certainly has been a journey.  When my wife leaves a concert and says that sounded as good as listening at home, she is complementing the venue…and Steve the credit belongs to you.
« Last Edit: 3 Sep 2016, 01:42 am by ketcham »

ketcham

  • Industry Participant
  • Posts: 285
Although currently i am not using Steve's volume control, it is designed in such a way to present the information in bit perfect form nor loose dynamics in the output stage.  The old Meitner Museatex bidat had a similar design. 

ODSE > Final Drive > Allnic A6000 mono blocks was far superior to ODSE > Allnic L3000 > Allnic A6000, as if wool blankets three layers thick were removed from the speakers providing more depth, detail, improved clarity and imaging.  The Allnic retails for $13,000.

I since replaced the Allnic L3000 with the $23,000 L5000 DHT, which I purchased used for $10,000.  The sound was different but of a comparable quality.  With a DHT/Transformer based system, there is limited benefit adding the final drives.  Inline, it does provide significant gain on the 0 gain setting. 

Per Steve's recommendations I am trying a loom of Antipodes Reference cables.  My system likes balanced cables better and I will sell off my High Fidelity Ultimate Reference Loom.  One of my amps gets a 60/120 Hz low noise with balanced cable while using the VH Audio Plasmatron.  I believe the Plasmatron adds significant benefit to Steve's DAC.  Matt Clott tried his on my recommendation and didn't like it.  I reviewed this further, and although I like the Plasmatron, the balanced cable adds more benefit. 

This is my long winded answer.  The Final Drives are passive using quality transformers.  For the price, I have yet to see anything better.  His DAC plus Final drive to amplifier is an excellent combination and definitely a reasonable price.  For me buying and selling gear often is a money losing proposition and getting the sound I desire was met with this consequence, more so than I care to admit.  Through this process, I learned to trust Steve's ear. In hindsight, I wish I invested in the ODSX with Ethernet input/Findal Drive combination and focused my resources elsewhere.  This would negate the need for a server, preamp and allow yourself to focus on an amplifier or speakers, if needed.

Hope this information helps.

audioengr

Thanks so much for putting so much thought into this lengthy post.  Lots of great info there.  It was my goal with the SE and SX to approach the SQ of NOS ladder DACs while providing hi-res capability and more detail.  I do some unusual outside the box things with filtering in there which may not look great to the measurement gurus, but these really deliver the result I'm after.  I am so pleased with the performance of my output stage (because it is as good as a tube), that I have plans now to do a headphone amp based on the same technology, but with further improvements that I cannot do inside the DAC.

I was never personally blown away by DSD and in fact found it to be "over the top" in most cases.  With the advent of many new technologies, like DSD and recently MQA, there seems to be a short-lived splash and then they slowly die out, reverting back to PCM, which has the largest set of content.

Looking forward to more feedbacks like this.  I will also be building a demo DAC I can send out for professional review as soon as I can make time between orders.

Steve N.

ketcham

  • Industry Participant
  • Posts: 285
After some more experimentation, I am changing my thoughts regarding the Final Drives.  Over the years I have learned that great equipment may not play well with each other.  There is a certain synergy between components and accessories.  The last 2 years there have been many cabling systems that either were trialed or purchased.  Steve understood my frustration and recommended Antipodes Reference cables.  The synergy is amazing.  Anyone who is interested in building a new system, consider Ethernet-ODS e/x>final drive with Antipodes loom.  The package will save you a source component and a preamp.  The value to cost ratio can not be under emphasized. 

audioengr

After some more experimentation, I am changing my thoughts regarding the Final Drives.  Over the years I have learned that great equipment may not play well with each other.  There is a certain synergy between components and accessories.  The last 2 years there have been many cabling systems that either were trialed or purchased.  Steve understood my frustration and recommended Antipodes Reference cables.  The synergy is amazing.  Anyone who is interested in building a new system, consider Ethernet-ODS e/x>final drive with Antipodes loom.  The package will save you a source component and a preamp.  The value to cost ratio can not be under emphasized.

I'm glad the Antipodes cables are working out for you.  One thing that is important when driving cables from DAC to amp is that these cables have low capacitance per meter.  The difference between a typical 1m cable and a low capacitance cable is a 10% loss in amplitude at 20kHz, where the loss at 100 Hz is zero.  Longer and higher capacitance cables will be even worse.  Then there is the dielectric absorption due to conformally coated conductors in the cable.  This effect was nonexistent in my original cables and in these Antipodes cables due to their construction.

paul79

  • Industry Participant
  • Posts: 901
I agree completely. Using the Antipodes Cables is kind of like giving your system a shot in the arm if you have been using plastic cables. There is freedom and effortless to the sound that is difficult to describe. You just gotta hear it.

Their beauty I believe, is their simplicity. No plastics, no shielding, good metal. Simplicity always wins with audio.