Absolute top tier DAC for standard res Redbook CD

BTW, I have a Plasmatron as well and use it every day for my DAC and other digital front end gear.  Highly recommended.

Steve N.

Empirical Audio

I have been listening to speakers and amps for probably 10 years or more at shows where I exhibited.  The recent standouts for me are:

1) Wells Audio

2) Morrow Audio (digital amp)

3) Dan Dagastino

These all sound very natural and approach the best in tubes.  I have recently retired my modded Parasound JC-1's and I'm using SE 805A monoblocks now.  These tube amps are wonderful, but the Chinese output tubes only last one year.  Trying different tubes soon.

Steve N .

Empirical Audio

Matt - I have several orders for ODSX, so I'm building those.  I decided on the Duelund caps after testing the Jupiters and they melted...  Bass got crazy after they melted...  I am shipping it with my original USB module right now.

The new USB interface (in beta testing) started working on 176 and 192 after a change I made, so I thought I had fixed it.  Next day it didn't work.  Bummer.  Still experimenting to determine why it worked then not.

Second Ethernet module is being prototyped.  It is powered up, but not programming from the flash yet.  Very excited about this one.  It may end up being the default interface for the ODSX.

Steve N.

Matt - when are you going to break-down and get the best server on the planet, the Antipodes?

This is what I have been using at shows with the Overdrive SE and SX.  Nothing touches it IMO.  Reviewed recently in Stereophile.

Steve N.

Empirical Audio

From what I have heard some of the newer class-D amps are as good or better than the best SS amps, but tubes still win IME.

I have custom DC-coupled 805 SE tube monoblocks that even beat the AR stuff.  Nothing I have heard comes close to these.  The right tubes are essential though.

Steve N.

Empirical Audio

Matt - I can see how everyone can have different experiences with the Antipodes server. The earlier ones did not have the really good linear power supply. Also, if they are not driving USB out to a low-jitter USB interface, this can make a big difference. That is aside from the systems that they were used in, which probably had lots of ground-loops and substandard active preamps etc... Lots to go wrong there, and usually does at these shows. Many times the room cost is shared and it ends up a mish mosh of different hardware and cables that adds up to crap sound. Very few exhibitors are even capable of achieving a great sound at a show, even on the third day when everything should be broken in.

The point is that if it sounds crappy at a show, it may be due to other system things.  On the other hand if it sounds great at a show, it will probably perform well at home.  Then you have to root out the real weaknesses in your system.

Matt - have you thought about breaking all of the ground-loops in your system?  This is the next major step IMO.  I have zero ground-loops in my system and it is dead quiet and spooky clear.

Steve N.

JWM - did you get to hear the Meridian MQA demo at CES?

As for eliminating my ground-loops, here are the three tactics I use to break them:

From server to USB interface - I use a XMOS interface with galvanic isolation

From USB interface to Overdrive DAC - The transformer on my DAC S/PDIF input takes care of this one

From DAC to monoblock tube amps - I use my Final Drive line-level transformer and selector - also converts balanced to single-ended

This eliminates all ground-loops in the system. I can put my ear up against my tweeter and nothing, even at full volume. Every component has earth ground.  I never had any hum, but there was RF noise that is insidious.  Don't know you have it until it is gone.

BTW, tube gear that is not earth-grounded is dangerous IMO.

Steve N.

Empirical Audio

Richfield - the ground-loops are obvious in any system.  Testing to see if they are gone is simple, just unplug the AC cords and use an ohmmeter between the grounds.

Steve N.

Empirical Audio

Paul - Signal transformers should not be  a problem.  You can always increase drive by reducing voltage with 2:1 ratio.

ctsooner - MQA depends on the licensing.  IF its anything like MPEG, I cannot afford it.  Hopefully, the chips and licensing will be affordable for the little guys.

BTW, they say the big win is the impulse response, which has no pre and post ringing.  My Overdrive already does this ;>)

Steve N.

Empirical Audio

Just like the OR5.  I2S, AES and S/PDIF out.

Steve N.

Empirical Audio

I use the OR6 with both Dynamo and Power Block to good effect. I wish I knew when it will be available. The OR6 has the galvanic isolation, so the Power Block and server is on one side and the Dynamo and OR6 is on the other side of the isolation. OR6 has an initialization bug that comes and goes that prevents 176 and 192 without errors, so it is not ready for prime-time. I did not design the custom XMOS firmware for this, it was contracted, so that may be the issue.

In the meantime, I am working on the Interchange, which is a Network Streamer with wired Ethernet connection. If the OR6 does not pan out, the Interchange will replace it.

Steve N.

Empirical Audio

This network streamer is only a renderer.  It uses another computer or ipad etc. to drive the music to it over the internet.  I was told to call it a streamer, but I think this might cause confusion.

The pricepoint for the Interchange is $2600-2800 with the Dynamo power supply.  The supply alone is almost $800.

Steve N.

Empirical Audio

Matt - I recently built an ODSX for a customer that did not care about 176 or 192 file playback, so I installed my new XMOS USB interface in it for him.  This is the first one like this.  This combo is simply incredible. Hands-down the best SQ I have ever heard, from analog or digital sources.  Even really old tracks, like Peter Frampton live sound amazing.  He also ordered my Power Block USB cable supply to go with it, which makes a big difference.

I am still working hard on my new Ethernet interface, which supports all formats, PCM and DSD.  I should have one running in the next week I hope.  This I think will be the default interface for the ODSX.

Steve N.

Empirical Audio

Matt - Focal speakers!! At the LA Audio show I heard at least three exhibits using Focals. All sounded really good. I have added Focal to my list of recommended speakers, along with Vapor and TAD.

BTW, did you ever compare the Aurender N10 to the top of line Antipodes server?

Steve N.

Empirical Audio

Here is a good test to see if your system is "truly resolving":

Take a FLAC, uncompressed FLAC, ALAC or AIFF track and convert it to .wav using DBpoweramp or XLD. Play both tracks alternately.

If you cannot hear any difference and the .wav does not sound better, you need to tune the position of your speakers or add acoustic treatments. If you still don’t hear any difference, you have an offending component or cable in your system, maybe more than one.

When I hear people say that all DACs sound very close, I have to wonder about their system.

BTW, Baja - that Antipodes is a very good server. IMO, to get the best out of it, you should be using a good USB converter or USB DAC, not S/PDIF from it.

Steve N.

Empirical Audio

geoffkait - you got it, however, I would start by selling the CD transport and look for a good computer audio source.  Too much mucking constantly with CD's, besides you end up listening to tracks that you don't want to hear, and only 10 or 12 of them.

Then, of course, you have servers or playback software and digital cabling and Ethernet cabling and isolators and LPS for this and that, and then there is format....

I don't consider these tweaks, but they are essential to getting the best sound from a computer source.

There is one tweak that I actually use:

Plasmatron from VHAudio.  Highly recommended AC regulator.

Empirical Audio

Willemj - I’m trying my Hifi-Man phones with my iPhone 6S today!!

The one thing that I’m leery of is using that much digital volume control.

Steve N.

Empirical Audio

Lately, I’ve been using music with non-music in it or lots of percussion in it, like:

Andreas Vollenwider - Trilogy - Behind the Gardens

Doors - Riders on the storm

Pat Metheny - On the Road to You live - Half Life of Absolution

These are great to determine liveness and imaging.

Steve N.

Empirical Audio

Matt - I had the Chord Qute here and it needed a better power supply and even then the bass was disappointing. Highs had amazing focus and detail however.

Is the Hugo better on bass?

BTW, ctsooner - I have designed a dual-mono, fully balanced-only class-A headamp based on the output stage of my Overdrive SX.  Really excited about this product.  Calling it "Eliminator". First protos hopefully early next year.

Steve N.

Empirical Audio

CTsooner - I plan to if you are who I think you are.  Be patient.  It's worth the wait.

Steve N.

Empirical Audio

Bill - these older D/A chips are very musical, but simply not as accurate as the modern designs, as well as not supporting higher sample-rates.  I offered a ladder-DAC tube output DAC myself for several years.  It was very sweet sounding, never harsh, but never really live either.

I consider the digital filter in most DACs to be the #2 impediment for good digital playback (jitter is #1).  These older chips solve that problem.  They usually don't incorporate any digital filter or force the chip to automatically select a filter.  They make it possible to have only an analog filter and no digital filter.

The problem is that even if you only care about 44.1kHz playback, the HF transient response of these older chips is just not like the live performance.  The attack and decay never seem to be like the live event.

The is why in the Overdrive DAC (one of Matts picks), a newer technology chip was selected that supports up to 192kHz, but allows one to manually select the digital filter.  I recommend to select only the 192 digital filter (there are three) on the Overdrive, even when playing 44.1kHz tracks.  This is how I use it.  This approximates what these older chips do at the same time delivering accurate transient response.

A number of companies have gone to extremes over the last 10 years to develop their own digital filtering techniques, in order to eliminate or reduce pre and post ringing on impulse response of their DACs.  This was an effort to improve the deficiencies in digital filtering in off-the-shelf D/A chips.  So far, I have not been impressed by the sound of these filters.  I still believe minimal or no digital filtering is best.

There are a couple of modern D/A designs that use the old R2R ladder technology, but also support higher sample-rates.  The TotalDAC and the DaVinci II are such designs.  I have read great reviews on both, so they evidently have decent transient response.

Steve N.

Empirical Audio

Wow, I guess I'll have to get to work and build one then....

Steve N.

Empirical Audio

CT - I think crossfade/stereo depth for headphones is the realm of good DSP software.  Doing it in hardware is like upsampling in hardware.  Never as good as good DSP software.

For IEM's, should there be a low-gain switch maybe?

Steve N.

Empirical Audio

shadorne - agreed, however, the devil is always in the details. How to make it sound like analog and not electronic.

Upsampling has its own issues, particularly in hardware. There are always compromises made in these designs. My own reclocker is extremely good, rejecting almost all incoming jitter and delivering a low-jitter upsampled output, but when compared to my best USB or Ethernet playback which is not upsampled, the non-upsampled tracks win for best sound quality. Both scenerios use essentially the same clock and power supplies/regulators.

This is precisely why I don't put an upsampler in my DAC, as so many other DACs do.

Steve N.

Empirical Audio

So if this inter-sample overs effect happens only on peaks, why do I hear differences even in low-level signals?

I think the SQ issue with upsamplers, albeit small, has to do with hardware implementation of upsamplers.  Doing it is S/W seems to deliver better SQ.  I use Izotope, supposedly the best S/W resampler, and the SQ is usually great, better than the original.

Steve N.

Empirical Audio

So, if the track is recorded at low-level so that no clipping occurs after upsampling, there will be none of this effect?

Steve N.

Sounds like you need more "non-music" tracks to compare these two.  Falling rain, thunder, canoe paddling in water, wind are all helpful.  I use all of these to determine accuracy as well as well-recorded percussion tracks.  You need to hear the decay of gongs etc..

Start with a good recording of Doors "Riders on the Storm".

Steve N.

Empirical Audio

Merrill stuff is good.  One of the few SS amps I recommend, along with Wells Audio and D'angostino.

Steve N.

Empirical Audio

I'm am also getting superb results from my Interchange Ethernet interface with Jriver using DLNA now.  I am currently modeling a Roon interface, so we will see how it stacks up against the DLNA Jriver player.  I can do a fairly quick A/B because it supports both DLNA and Roon.

BTW, Matt, I finally got my XMOS USB interface working.  Sounds great.  If you wanted to upgrade the Overdrive SX and swap this module in for your older M2Tech firmware module, let me know.  Much lower jitter.  It only needs a driver for Windows. 

I also have new mods for the Overdrive SX that take it to a new level yet.

Steve N.

Empirical Audio

If you are planning to spin CD's, first reduce the jitter from the CD transport, then think about a new DAC.  Either CD player as a transport is fine, but you can minimize the jitter by adding a Synchro-Mesh reclocker ($599.00) and a reference S/PDIF coax cable ($499). LPS upgrade also available.

Here is a typical transport jitter:

http://www.audiocircle.com/index.php?topic=154408.0

Much higher than you would expect it to be.

BTW, I also have a demo Overdrive SX DAC that might be available soon.  It could be demoed right away with a refundable deposit.

At least look at the top DACs reviewed by audiostream.  "best bits"

Steve N.

Empirical Audio

Would my Modwright Marantz SA 8005 have to be sent in or is this a plug-play mod? Curious if you have done any jitter samples on this unit but imagine it’s similar to the Oppo?

No, this is an add-on device that inserts in the S/PDIF cable. a reasonable cable going in and a reference cable going out is what I recommend. I offer an excellent reference cable.

I don’t have any jitter measurements on the Marantz. Almost no Transports have the 20psec of jitter that you get with the Synchro-Mesh.

I may not even want a Dac if this will make it even better... Any installed pics of this unit? Curious on how it’s connected.

Folks with NOS DACs say that the Synchro-Mesh (SM) really wakes them up and improves the dynamics, imaging and detail.

I don’t have installed pics, but it’s really simple:

1) DC power inlet that can be from the included wall-wart or an upgrade LPS.

2) S/PDIF coax on BNC or RCA or Toslink inputs

3) S/DIF coax on BNC or RCA or Toslink outputs

All you need is a cheaper coax cable, like Blue-jeans connected from transport to SM and a reference coax cable from the SM to your DAC.

Steve N.

Empirical Audio

any of your customers use this between the PS Audio Directstream DAC and the PS Memory Player transport? Good results?

Most that have PSAudio DS are using HDMI I2S from my Off-Ramp 5.  They can also use my Interchange Ethernet renderer.  Same outputs.

If you wanted to try the Synchro-Mesh using my coax cable, it would likely outperform even the I2S, but this depends on the S/PDIF path inside the DAC.  The signal will have lower jitter, but the path may make it a lot worse.   30-day money-back if you want to try it.

I may do a Synchro-Mesh in the future with I2S on HDMI, but other things are more important now.

I can tell you without hesitation that the Interchange with HDMI I2S to your DAC will be hands-down the best solution, best SQ.  It blows-away any USB, even my own.   It's identical to the Off-Ramp on the webpage, but has Ethernet input instead of USB.

Steve N.

Empirical Audio

I'm assuming that I'll be using both in/outputs on the SM going into my Marantz Digital in/outputs?

Transport output connects to the Synchro-Mesh input

Synchro-Mesh output connects to DAC input

Steve N.

Empirical Audio

Some sobering thoughts and measurements.....

Not so sobering IME.  I have an RPi and an even better hat board, the Allo Digione.  I also have a Digi Pro board.   I have made jitter measurements on all of them.  The Digione is quite good for $100, but I still cannot get 176.4kHz or 192kHz files to play through my DAC on either board.  I have been messing with it for a week, including modding both the RPi and the HAT boards.

Steve N.

Empirical Audio

In the case of DACs, it’s my understanding that delta-sigma is "lossy" (i.e. all of the source PCM music just isn’t converted and played back) whereas an RxR DAC is "lossless", so all of the music is played. And if that is true, it becomes hard to imagine how delta-sigma could possibly ever sound better, regardless of how it is used or any other factor for that matter.

Not true. I think you are maybe confusing this with lossy formats. You can think of the Delta-Sigma as an engine that creates the R2R function repetitively. Instead of having passive banks of resistors to create different voltage levels, it uses a single engine that iterates to create these dynamically.

Delta-Sigma can definitely sound better than R2R, particularly older R2R chips. Newer Delta-Sigma chips not only support 24/192, they have excellent measurements which include very low distortion and noise as well as higher S/N ratio. They tend to reveal more detail at high frequencies too. The older R2R chips are often implemented without any digital filter, so this solves this issue, but it does not solve the high-frequency detail problem.

The thing that sets the Delta-Sigma apart from the R2R is usually the digital filter implementation. If this is executed well, the Delta-Sigma can beat the R2R, depending on implementation and particularly power delivery to the chip.

This is one down-side to R2R discrete implementations. They tend to require much more board space than the Delta-Sigma, so this makes it more difficult to get di/dt power to the devices involved.

A second down-side to R2R discrete implementations is the tolerance of the resistors and their accuracy. The Delta-Sigma only needs one or two circuits to be very accurate. The discrete R2R needs hundreds of individual resistors to be extremely accurate.

A third down side to discrete R2R implementations is the speed at which they can process new words.  Because of the distributed nature of the resistor networks and the many drivers needed, it is more difficult to get a really fast response from these circuits, like you would with an single integrated circuit.

Steve N.

Empirical Audio

Okay, time for me to weigh in again....

I also manufactured an R2R DAC in the past, using the laser-trimmed PCM1704 binned for performance. It was a very sweet sounding DAC, and not because of the tubes (I used Siemens CCA NOS tubes, very fast). The detail rendering however was never as good as modern Delta-Sigma chips. Here is what the PCM1702 datasheet says about R2R and Delta-Sigma:

"However, even the best of these (R2R’s) suffer from potential low level
nonlinearity due to errors in the major carry bipolar
zero transition. Current systems have turned to oversampling
data converters, such as the popular delta-sigma architectures,
to correct the linearity problems. This is done, however,
at the expense of signal-to-noise performance, and the
noise shaping techniques utilized by these converters creates
a considerable amount of out-of-band noise. If the outputs
are not properly filtered, dynamic performance of the overall
system will be adversely effected."

So, both techniques can have good S/N (Delta-Sigmas have improved a LOT since the 90’s), but the issue with Delta-Sigma is the filtering. The issue with R2R is this non-linearity. This is what I had said all along. Eliminate the bad digital filtering and you have something really fine with Delta-Sigma. This is what my ears tell me.

As for comparing the Bricasti to the TotalDAC: I think the devil is in the details. Can you even make it apples-to-apples? Even if both were driven S/PDIF, does Bricasti resample and the TotalDAC not? If one is using Ethernet and the other USB, it’s apples to oranges. If even one USB, S/PDIF or Ethernet cable is different in the two systems, all bets are off.

The only valid comparisons would be maybe Ethernet to Ethernet and USB to USB using the same playback software from the same computer or server and the same cables. And we have not even talked about the system of preamp, amps and speakers yet.

The point is that no one can make any credible claims about one DAC over another if the device is not tested in the same system, preferably at the same time with the same music. Particularly if both DACs are very good quality. This is exactly why Matt’s experiments are so important, even more important than professional reviews. Even reviewers systems change over time. If reviewers did more shootouts like Matt has done, we would learn a lot more.

Steve N.

Empirical Audio

jwm and charles1dad - I think the problem might be the system.  Each of you heard the same two DAC's, but in different systems.  It is very possible that one preamp or set of cables made one DAC sound better than the other and the opposite in the other system.

There is certainly expectations and listener preference, but I would suspect that it might be more the system effect.

Steve N.

Empirical Audio

My Oppo UDP-205 (delta-sigma) has a user preference setting whereby the filter can be changed. Way back, I did some research as to what this entailed, and found it had all to do with "pre-ringing", "post-ringing" and basically everything at the 20 KHz point. I tried all (7) available filters and none made the sound "better" than the default (mini-phase fast). It only made it as good or worse, to my ears.

So, what "filter" is the issue with delta-sigma?

Filter response, such as "apodizing filters" are not what I'm talking about. If you want to really improve Delta-Sigma filtering, you have to either eliminate the digital filter completely, replacing it with analog filters, or set the filter frequency much higher than most chips do automatically.  This effectively removes the effects from the audio range.  This is what I do in my DAC.  The user can select the 192 digital filter when playing back 44.1 files, and this is how most customers use it.

Steve N.

Empirical Audio

gdhal - what source is driving your Yggy?

What are your Delta-Sigma DAC's?

Steve N.

Empirical Audio

Musical Fidelity M6si > Belden 5T00UP > Golden Ear Triton Reference (pair)

Schiit Yggdrasil > Canare L-4E6S XLR > M6si

Oppo UDP-205 > Belden 1694A (Canare RCAPs) > Yggdrasil

Emotiva ERC-3 > Belden 1800F AES/EBU > Yggdrasil

Samsung SMT-C5320 > Mitsubishi Eska POF > Yggdrasil

The Yggdrasil has a very good S/PDIF receiver, the AK4113. Same one I use. The results should be stellar with any good low-jitter source. Unfortunately, you are using very poor S/PDIF cables. See my cable jitter plots here:

http://www.audiocircle.com/index.php?topic=154425.0

The DAC in the Oppo is not great, but should sound decent with a low-jitter source using the balanced analog outputs. The Emotiva is inexpensive, using many mediocre op-amps, if its the one I am thinking of. I think the analog section of the Yggy is probably much better, as is the I/V converter and power supply. I don’t think it’s really apples-to-apples to compare Delta-Sigma to R2R with these particular DACs.

If you are using Oppo as a transport with 1694A, you are getting a LOT of jitter see these plots:

http://www.audiocircle.com/index.php?topic=154408.0

Steve N.

Empirical Audio

gdhal - This is an awesome gift!  My wife says the Hall and Oats concert.

Thanks,

Steve N.

Empirical Audio

Anyway, what is your opinion of using USB in the following configuration:
Gateway NV79 > Belkin e129760-c usb > Yggdrasil

A better USB cable would improve things.  The problem is that most USB interfaces built into DACs are not that great. There are a couple of decent ones.

Should USB produce better results than the Belden 1694A coax?

Depends on the USB interface and the S/PDIF interface and associated clocks.  There are a lot of variables.  I would think that a low jitter source using a good S/PDIF cable to the Yggy would outperform the USB input.  My BNC cable has RCA adapters with it, so you can do any combo.

Should there be improvement if I were to switch the optical from my cable box to the Oppo? (note that according to Schiit, AES/EBU is best to the Yggy, followed by coax)

If you use this for TV, movies on Netflix and 2-channel audio, then I would recommend using the IFI S/PDIF iPurifier for $149.  You can go optical in and coax out using my Standard BNC cable.  This is what I do with my OPPO and my Smart TV to reduce jitter.  It supports all formats, including Dolby Digital and DTS.  Big difference with TV and movies.

I only have AES/EBU out from my Emotiva. I also recall reading from Georgehifi that some type of glass optical is best, which I’m also not using.

I thought the Emotiva was a DAC.  If so, then the AES/EBU is an input, not an output.

I'm getting confused as to what is driving and what is receiving here....

Glass cables can be very good compared to cheap plastic, but the plastic I use from my Smart TV is even better than glass.  From btpa.com:

https://btpa.com/?target=search&mode=search&substring=toslink&including=all&brandId=0

Steve N.

Empirical Audio

I studied atmospheric physics in depth too but I better stop there... as nobody likes to hear rational discussion on that topic which has become a new cult religion.

I do.  People need to be educated on this. Most don't realize that GW is an existential threat to mankind.  Weather catastrophes cost the US over $300B in 2017.  Worst year ever on record and its only going to get worse.  And here we are quibbling about $2B or $18B for a wall. We cannot even fund FEMA anymore.  People don't even realize that power plants are only half the problem.  Cars and other vehicles are the other half of the problem.  Time to by a Tesla.  They are fantastic cars!!

Steve N.

Empirical Audio

CTsooner wrote:

I'm getting The Memory Player with built in DAC and was wondering if I should use the SM to connect the TV to the internal DAC via BNC?

If you are only doing 2-channel, the Synchro-Mesh is a great choice.  If you want it for HT and have a SS proc., then the IFI S/PDIF iPurifier.

Steve N.

Empirical Audio

Shadorne can you explain then dark energy.

This is the energy needed to place all of the masses in space relative to each other that is not explained by gravity and space-time.  It's an unknown energy.

Steve N.

Empirical Audio

Matt - You are off to a good start.  One thing though.  The audio level on the YouTube video is a bit low.  I have to put my volume slider to the max to hear it well.

You definitely have the right kind of diffusors behind the speakers.  I would be interested in how these speakers work and what all those drivers are.  Also, I would expect more rugs on a hard floor like that.  It must be very live.

Steve N.

Empirical Audio