ESS Sabre 9018 vs Wolfson WM8741

[AT0MAC]

I currently have the Arcam rDAC but today stumbled upon the AudioLab 8200CDQ - witch could let me sell both my pre-amp, cd-player, dac and interconnects and replace it with this little all in one wonder...

 

Its a very intriguing thought, but how good is the ESS Sabre 9018 compared to the WM8741??

 

Anybody have the ESS Sabre chip in they DAC, and how does it perform?

 

[VandyMan]

It is a very good chip, but the quality of the DAC involves many other factors. There are both great and so-so implementations on the market.

 

[AT0MAC]

Any words on how the AudioLab 8200CDQ sounds as DAC compared to the Arcam rDAC?

 

[Tog]

Silly question I'm afraid - you might as well ask diesel or petrol engine - which is better?

 

The dac chip matters far less than the implementation and how will you be able to glean anything meaningful from people talking about a Sabre chip in one dac compared to a Wolfston chip in a completely different product?

 

It is hard enough comparing different approaches to dac inputs or reclocking or DSP!

 

Both are good chips - audition some dacs and buy the one you like the sound of.

 

Yours, Tog

 

[VandyMan]

>> Both are good chips - audition some dacs and buy the one you like the sound of.

 

That really is the best advice. I went with the Benchmark HDR in part because they offer a home audition (30-day return).

 

[Mercman]

My Wavelength Crimson is modular for the DAC board. I have the Numerator which is a Wolfson chip and the Denominator which is the ESS Sabre. The ESS Sabre, at least the way Gordon executed the design, is superior sounding.

 

[spdif-usb]

Just over a month ago I purchased the new Eastern Electric MiniMax DAC Plus, which uses the ESS Sabre³² ES9018 chip.

 

I had been looking at the Benchmark DAC1 HDR and saw it was a giant-killer but its price still scared me away in the end. Next, I discovered both the Wyred4Sound DAC-1 and the Audio-gd NFB-3. Finally, I noticed Johnny Darko's review on the Eastern Electric MiniMax DAC. As luck would have it, this unit had already been discontinued by the time. So I decided to bury the whole idea for yet another while.

 

Now that I've taken the plunge with the successor of the MiniMax DAC, called the MiniMax DAC Plus, all I can say is I'm sorry to sound daft but you should just see the grin on my face! Forget about dual Wolfsons, this new unit shows what the ESS Sabre chip is truly capable of. No, its implementation is not necessarily bad and no, USB input is not necessarily bad either. You'll be pleasantly surprised. You can either believe it or not. I'll be ordering a pair of cherry picked Psvane 12AU7-T vacuum tubes soonish.

 

But doesn't it sound too clinical, you may ask? Well, no, no and no it doesn't. Too synthetic? Nope. Just fantastically detailed all along the entire frequency range, wide open, huge soundstage and separation, no noticeable coloration, but slightly warm and ultimately spot on timbre. There's more than sufficient dynamics and air. A very worthy musical experience. The background is so dark you don't even miss the balanced XLR outputs. And it plugs into my small netbook using nothing but a dirt cheap USB cable because it has an asynchronous USB interface from M2Tech that's 24/192 capable, with hassle-free device driver software that's downloadable from Eastern Electric's website. What was that again about giant-killers?

 

[barrows]

have mentioned-it is virtually impossible to directly compare DAC chips, although Steve (mercman) probably comes the closest. Different chips really require different implementations to perform their best, so one cannot just plug a 8741 into the same design as a 9018 to compare them.

That said, my DAC with the ESS 9018 is really, really good-it does take very dialed implementation (especially power supplies) to get the best out of the ESS 9018-but when done right, it has a lot to offer.

 

[Sri]

I'm not familiar with what that means in terms of designing and engineering a DAC, can you please elaborate? An example of showing how a "dialed" design differs from and "undialed" design would be helpful.

 

[barrows]

This is not simple to explain, but I'll try and then give some examples:

 

The ESS 9018 chip is very sensitive to power supply implementation, it really needs very good power supplies to perform its best. Additionally, for it to operate in current output mode (where it offers best performance) it needs to be coupled to an output stage with which offers very low input impedance (< 1 ohm).

 

Every DAC chip has different requirements for the surrounding circuitry, so you cannot just plunk a DAC chip down in the same circuit to listen to the differences: the circuit has to be optimized to suit the requirements of each individual chip.

 

The Wolfson 8741 is a voltage out chip, with onboard current to voltage (I/V) conversion, the ESS 9018, when paired with a low impedance output stage, is a current output design, and how the following I/V stage is done will have big influences on the sound. Since the 8741 does I/V on the chip itself, the designer cannot implement their own I/V stage for best optimization.

 

OK, examples with my DAC: The ESS 9018 DAC chip uses a local masterclock, and 4 local, separate power supplies: two supplies which power the analog outputs of the chip, and two supplies which power the digital sections of the chip, additionally, a fifth local supply is needed to power the local masterclock. Most designs will implement these supplies with local, IC chip based voltage regulators. These regulators are good, but not great. The Twisted Pear Audio Buffalo DAC uses five individual, discrete, shunt regulators for these supplies. Shunt regulators are lower in noise, and much lower in output impedance (across a broader bandwidth) than ordinary IC regulators. Using the shunt regulators results in much improved sonic performance, this is probably mostly due to their low output impedance, which stops the individual supplies from being modulated by each other (better isolation from one supply to another).

In my DAC I then go one step further-the DAC board (with its 5 onboard shunt regulators) are fed with another shunt regulator, acting as a pre-reg. This pre-reg is fed from its own transformer winding to isolate it entirely from the other power supplies (which are three other shunts, one for the USB receiver, and two for the +/- 15 VDC analog output stgae supplies). TPA also provides a very, very good masterclock on the Buffalo DAC (Crystek CCHD Series), this clock is much better than the clocks used on most commercial DACs.

Additionally, The Twisted Pear Audio offers two output (I/V) stages designed specifically for the Buffalo DAC. I use the Legato II (now there is a III version) which is a discrete bipolar input MOSFET output design with no global feedback. This output stage presents a less than 1 ohm load to the ESS 9018, and this low loading allows the ESS output to operate in current mode, resulting in the best possible performance.

 

The above illustrates that there is a lot more to making a DAC perform really well than chip selection. I am not an engineer, and I am sure if any of them are reading this they probably find my examples overly simplistic, and that they could probably point out many other areas of implementation which matter to the performance.

 

[4est]

Not to put words into another's mouth, but "dialed" is short for "dialed in" which means "adjusted optimally" or "tuned" for a situation. Many products are not because it can be expensive and or labor intensive to do so.

 

[bleedink]

FWIW it does seem to have much to do with implementation. My receiver has DACs for each channel that are identical to the new flagship Integra AVR. Obviously (or perhaps not so) even though I am very pleased with the sound of my receiver I'd be foolish not to think that the implementation of the DACs in the Integra are probably much better suited to the audiophile. I have not compared the two but many of these chips are just reference for building systems. It is the way they are implemented that will make more of a difference. Also the dac in my Oppo is supposedly better than the DACs in my receiver. However when processing in the Oppo it sounds terrible, or not as spacious as my receiver. Better chip, worse results. Worse chip, better results. I would certainly try to find as much as I could about the DACs properties but I would not say that is going to tell you a thing about how it sounds. One manufacturer may choose to use one DAC to drive 2-8 channels. The next uses one DAC for each channel. So identical DACs in that case would likely have very different signatures. After all, the work gets spread through multiple DACs and presumably they would have less computations to do. It's nice to know the little details like does the DAC decode DSD or what have you but specs are not the definitive way one determines what sounds good.

 

[MykhailoM]

As it has been said - the Chip implementation is the KEY! and speaking from experience, the same Wolfson Chip in one design either from Arcam or Naim or Sugden etc. sounds a bit or a lot more different. Now in professional DAC Tube design it is a whole lot different story and therefore a different level if and ONLY if a premium EXPENSIVE parts used in very good design. I personally prefer Tube design such as MB™ DAC -WM1 Tube DAC it takes a real High End product to wrestle with it. Small unknown companies can do a wonderful product.

[Emcee]

FWIW...

Of course implementation is a major factor but I compared my original EE Minimax (ESS) with a similar priced Rega DAC (Wolfson) and the Rega sounded sluggish and boring. It was however more tonally rich than the EE but that was in a rather brightish system. In a more balanced system the EE gains in tone and keeps its inherent PRaT. The EE does have a few modifications which might have contributed to its superior characteristics (Belleson regulators for one) but nonetheless...

 

I also tried the supposedly giant killer and considerably less expensive Beresford Bushmaster with a Wolfson chip and it sounded flat and 2D. Couldn't understand what the fuss was about.

 

On the other hand a Yulong DA8 (ESS), while very dynamic, was just too aggressive for my tastes.

 

Years ago I do remember enjoying Rega CD players (Apollo, Saturn) which, if I'm not mistaken, use(d) Wolfson chips. Good implementation or is it just my memory?

[NeroMetalliko]

Hello,

I'm a newbie here, but I found this forum very valuable and interesting, so I want just share my limited and humble experience on this topic.

 

Few weeks ago I was looking for a replacement of my Marantz DV9500, died due to a fail in the spinner mechanic for the second time (already replaced one time in the past by an official Marantz support center).

After a lot of struggling thoughts I realized that I still needed a CD/SACD spinner, because I have some hundreds of not ripped CD and some dozens of SACD which I don't want to leave gathering the dust.

 

I think that the future for my habits will be a balanced mix between buying new hi-quality hi-resolution files (Sound Liaison, 2L and the like), streaming services (CD and possibly hi-res quality which Tidal is supposed to start to offer next year) and still owned media (old CD collection of titles which are not available on streaming services and the few real SACD discs already owned).

I plan to keep to a minimum the future buy of physical discs opting for more liquid alternatives, if possible.

 

Yes, I seriously think that I have to start to put all my CD on lossless ripped files: it will be a long and boring work, but once done it will be forever. If I will be able to rip even the SACD this will be ideal, because I will finally free from the need of a physical disc player which is my long term goal. I'm currently looking for a right PS3 to perform the SACD rip and I have even subscribed to the list of the PS3 player USB PC connected project already present here in the forum.

 

But all of this will require a lot of time and in the meantime I realized I still needed a decent disc player to fill the gap.

That said, unluckily my available budget for the new player was not too high, I needed a universal player with a good audio quality in the given price range and I wanted it to be possibly used even as DAC for the growing collection of files I plan to slowly build, mainly connected via a standalone USB disk.

 

Looking around I ruled out the current Marantz SACD players because they don't offer a decent capability to play hi-res and DSD files from the front USB port, which is mandatory for me at this point.

After a lot of additional thoughts I was convinced that a possible viable solution could be the Oppo 105D, which should have delivered a good price/quality ratio, SACD and file playback with decent audio quality and a lot of flexibility (like Tidal player and HDMI DSD output for a future external DAC upgrade, if the case). I was even curious regarding the new Cambridge CXU, which, even if more limited in some aspects (no XLR analog output and no back USB PC DAC port, just to say) was a little bit less expensive.

 

So I decided to do a listening test of both before pulling the trigger, and after a lot of research I was finally able to find a dealer here around (north Italy) that offered to me the possibility to listen to both for real. It is absolutely not easy to find this kind of possibilities here, so I got my car and in roughly 2 hours I was there with some USB key files and some SACD/CD of my taste.

 

Just to note: I usually listen to all genres of music, from chamber music to rock, lately more classical/chamber and jazz, but I have a lot of good old pop/rock album sometimes I still like to hear (first Genesis albums SACD remastered series, for example).

I own/play a (little) coda piano, a keyboard, 4 guitars (acoustic and electric with valve amps) and I regularly go to chamber music concerts in the winter season from, say, 30 years...

Music is a long time passion for me and I like to listen to the "real" live things very often, so I have good references when I'm listening my music at home. I'm interested only in stereo reproduction, not multichannel.

 

I entered in the shop heavily thinking to buy the Oppo, because of all the good opinions/reviews as giant-killer I had read, because the Sabre chip and the more complete package of features. But I wanted to give the CXU (which is a Wolfson based player) a try.

Frankly said, I had not too much time to spend there, so I rapidly checked some SACD, 2L/Sound Liaison files and even old (not perfect) CD recordings covering female Jazz voices, trios/quartets , trumpet, sax, guitars, piano, cello, orchestra and, finally some rock too.

 

For reference, the players were connected directly to a McIntosh integrated amplifier (I don't looked at the model, sorry) and using a pair of Sonus Faber Stradivari loudspeakers. Same signal cables for both players and no signal switchers or whatever.

I think this combo is not bad at all (and way more expensive than the players), probably the loudspeakers are on the bright/analytical side, and maybe this played a role during the evaluation.

 

However, the difference was evident from the first second with the Diana Krall "Besame mucho" voice until the last one of Peter Gabriel in "Selling England by the pound" (both SACD). The results were very consistent even with all the acoustic instruments solo/trio or jazz/classic (like Charlie Byrd or Jacques Loussier Trio) or big orchestra (2L "Magnificat" file) or even a humble low res mp3 of the Who cover "Saturday night is al-right for fighting", which was very challenging.

Both players played all the media inserted without a glitch in a fast way and without issues: CD/SACD and even the Cannonball Adderley "Somethin' Else" HDAD 24/192. From the USB key both players played flawlessly the wav/flac files up to 24/192 (no DxD) and DSD64 (no DSD128), tried DSF and DFF.

 

I found that the sound of both the players was very good but not without issues: the CXU was clearly more round and warm, with more evidence of the mid bass and clearly less sparkling effects in the mid highs zone, so no fireworks here, but the details were there with a solid scene and really a musical and enjoyable experience.

The Oppo performed very well too, providing a more open and spacious scene with tons of details, but in some cases I found the mid high side was a little bit too much in evidence. This was noticeable in the female voices starting sibling, or in the poor Who mp3 becoming irritating after just few seconds. So, if the recording is perfectly balanced and of highest quality, the Oppo will give its best, but paying attention that the rest of the chain should be not too bright too, otherwise the experience could be a little bit less pleasant.

 

I kept in consideration that once back home the player should have been connected to my Bow Technologies ZZ-one amplifier and to a pair of B&W 805s loudspeakers. I know my components and I know that the speakers are already on the bright/analytical side so the best choice for me was the CXU, and in any case this was even the conclusion I got there, because the more important thing for me was that in some cases some strain and listening fatigue with the Oppo appeared, and I didn't liked the slightly sibling voices.

 

This was a little bit disconcerting: I really wanted the Oppo to be the one to get, because the Sabre chip and because it offer more features, including Tidal support, XLR out (which currently I don't use btw) and USB DAC PC capabilities.

Regarding the CXU, I can live with the lack of XLR and PC USB DAC capability but I don't like the fact that its implementation with the Wolfson chips convert/upsamples all the input material to 24/192, even the SACD discs, so the DSD stream is never natively played. Luckily there is an HDMI output capable of pure DSD stream for a maybe future external DAC upgrade (finger crossed). In addition there is no Tidal player and nobody know if/when it will come.

Once back home I have discovered that the CXU offers a gap-less playback option for USB (but not for network) which it seems to work and is nice to have (I don't know if the Oppo has it).

 

However, currently I have to admit that in my living room, the CXU plays my CD and SACD very nicely (even if converted to 24/192) and there is little to no listening fatigue (depending on the recording quality, obviously). I strongly hope for a Tidal support to come soon, in the meantime I want to highlight that the USB playback of some Sound Liaison recordings like “Bach reflections” or “After silence” is one of the most enjoying/emotional listening experiences I have memory of (in relation to not “live” events).

 

Finally, my test was short, obviously partial and not exhaustive, so I perfectly know that my findings may vary depending on the components used in the chain and cannot be considered something to generalize, but I knew exactly what to listen for, so, in few time the decision was done and, to my big surprise, even with all the caveats showed, it was in favour of the CXU, which was the player I finally got.

 

I think that, as already highlighted, this could be considered another example in which the importance of the DAC chip implementation (and the audio chain relationship) are to be judged only from the final user listening experience.

So, probably this will not add more arguments to a general Sabre vs Wolfson discussion, but nonetheless I thought it could have been useful to share my little piece of experience here.

Sorry for the long boring post, at least I hope someone could find it useful.

 

Have a nice listening time,

ciao.

 

Andrea

[MykhailoM]

Thanks Andrea for shearing your experiences. If you have a chance, try to audition Wolfson DACs or ES Sabre32 or any other leaders for that matter in Tube output configuration, that could be a real treat

[NeroMetalliko]

Hello,

thanks for the suggestion, I will definitely pay more attention in the future to all the various DAC implementation, including tube output stages.

 

I'm really interested in the pure DSD conversion using a simple low pass filter for example, just to say.

 

Regarding my previous post above:

I have noticed that I have wrongly used the word "sibling"

 

This was noticeable in the female voices starting sibling

 

and I didn't liked the slightly sibling voices.

 

but the right one I should have used was something like "hissing",

sorry for that (English is not my first language, obviously).

 

In addition I realized that maybe my post above is too long and even OT here, because it's more related to a listening comparison between two players (Oppo 105D and Cambridge CXU) than to the strict topic of this thread (Wolfson/Sabre).

 

If the forum administrator think that it could make more sense to move the post in another section, please, let me know if I can do something for this purpose.

 

Have a nice listening time,

thanks to all,

ciao.

 

Andrea