General thoughts on dCS, Aesthetix?

[Hopstretch]

I've made an audio friend down here who imports the dCS range and has a slightly-used Debussy he's looking to unload at an advantageous (though still eye watering) price. Does the HC brain trust have an opinion on the dCS house sound in general or their Ring DAC topology in particular? I will have a chance to live with it in my system for a while before making any decisions. Another option is to wait for the first shipment of the new DAC from Aesthetix, who he also carries. The Pandora uses (as I understand it) roughly the same digital board as my current Ayre QB9, but with a beefed-up power supply and analog section (and toobs, obviously). Charles Hansen has been repping it pretty hard online and it's in a slightly saner price bracket. Can be ordered with an analog volume control, too, which is nice for my desktop application.

Largely gratuitous pics follow.

(Can also be ordered with silver faceplate, of course!)

(Closest I could get to internals. This is the CD player version.)

[dsavitsk]

I can't answer most of this, but I can make some huge over generalizations about tubes in DACs. And, that is, while I like tubes, and consider them to generally be the best active way to produce voltage gain, there isn't much place for them in most DACs outside of marketing, at least those using current out chips. The reason is that a current out DAC chip works best into a low impedance. That means either an opamp or the source/emitter/cathode of an active device. The input impedance for one of these devices is ~1/Gm. With a BJT or a FET, this is easily under an ohm. However, even a super duper high Gm tube, like a 7788, is still orders of magnitude higher. And, those are rare and unlikely to be seen in a commercial device. The highest Gm tube you are likely to see in commercial gear is a 6922, and it has an input Z around 80 ohms which is too high. Moreover, most higher end DAC chips run a pretty high current out, so you'd need to really push a tube like the 6922 to carry that sort of current.

Secondly, there is also often a buffer stage, but again, sand devices just work better here. A cathode follower really offers no benefit over a source/emitter follower sound wise and has worse drive ability. And, it costs a lot more to make. And it is likely noisier.

There are other topologies, obviously, so maybe there is some really novel circuit going on here to make tubes make sense, but it would be news to me. The place where tubes can work well in a DAC is with a voltage out chip (Wolfson for instance) where you want some voltage gain, but I don't imagine that this is the case with either of these.

Edited September 5, 2011 by dsavitsk

[Dreadhead]

Stretch you may want to send a PM to Filburt. Think he had some words on the dCS ring DAC design.

[Pars]

I can't answer most of this, but I can make some huge over generalizations about tubes in DACs. And, that is, while I like tubes, and consider them to generally be the best active way to produce voltage gain, there isn't much place for them in most DACs outside of marketing, at least those using current out chips. The reason is that a current out DAC chip works best into a low impedance. That means either an opamp or the source/emitter/cathode of an active device. The input impedance for one of these devices is ~1/Gm. With a BJT or a FET, this is easily under an ohm. However, even a super duper high Gm tube, like a 7788, is still orders of magnitude higher. And, those are rare and unlikely to be seen in a commercial device. The highest Gm tube you are likely to see in commercial gear is a 6922, and it has an input Z around 80 ohms which is too high. Moreover, most higher end DAC chips run a pretty high current out, so you'd need to really push a tube like the 6922 to carry that sort of current.

Secondly, there is also often a buffer stage, but again, sand devices just work better here. A cathode follower really offers no benefit over a source/emitter follower sound wise and has worse drive ability. And, it costs a lot more to make. And it is likely noisier.

There are other topologies, obviously, so maybe there is some really novel circuit going on here to make tubes make sense, but it would be news to me. The place where tubes can work well in a DAC is with a voltage out chip (Wolfson for instance) where you want some voltage gain, but I don't imagine that this is the case with either of these.

Was just poking around TentLabs site (hadn't been there for awhile) and noticed that the tube output stage he uses for his DIY CD player claims an input impedance of sub 1 ohm.

http://www.tentlabs.com/Components/page4/page4.html

The IV converter uses an E88CC valve as transimpedance amplifier. It fully operates in the current domain, and is a new and novel design. The advantage over existing tube based output stage is the low input impedance (typically well below 1 ohm) which reduces the distortion generated in the DAC chips compared to the situation when these are loaded with say 100 ohm, which seems common practice these days. After the E88CC stage, a passive filter follows. The actual IV conversion takes place using an Audio Note tantalum resistor, signal coupling is achieved using Audio Note copper foil paper-in-oil capacitors. 2 pairs of outputs are available, a straight output with a moderate 2.5 k-ohm output impedance, and a buffered one with output impedance well below 200 ohms. The IV stage is fed by a custom made toroidal transformer, the well regarded 6X4 takes care of high voltage rectification.

[b_b]

So, which one did you get? How does it compare to your exsiting dac?

[Smeckles]

Very late reply but I am demo-ing the Aesthetix Romulus in my system now.

I am no techie and know nothing about the circuit topologies of either unit, but the Romulus is good enough that I would seriously consider replacing my current CD player (AMR CD-77) with it. Both players use tubes in some configuration. Both have a USB input, along with SE and balanced connections. That fact that the Romulus is less expensive is a bonus, I guess. The Pandora is about $1000 cheaper than the Romulus if you have no silver discs to spin.

What changed: mostly the width and depth of the soundstage well beyond the one presented by the AMR-77 and then the Romulus also solidified the placement of instruments within the soundstage on even so-so recordings. I hate to sound cliche and use the descriptor "holographic" but that's really the difference it brought. It wasn't subtle. My definitely non-audio girlfriend noticed (and approved) immediately.

It did need a good 40-50 hours to break in and stabilize, but that's hardly in Audio Research territory for break in. For reference: my 2 channel system is Maggie 3.7s with an ASR Emitter 2 Blue amp, wired with Nordost Frey cables throughout, and an iMac playing Pure Music software so it's a fast and fairly neutral system, but it's definitely not a "theater in the round" type . The Romulus added that extra dimension to this, without losing the other positives in the system. It just seems very synergistic with this set up and my room.

I have not heard the Debussy, other than in a local showroom, and while it is undoubtedly a high end piece, I was very unfamiliar with the surrounding equipment (mostly Goldmund electronics and Kharma speakers) and could not determine how it would fit in my system. To be honest, I was not wowed by the Romulus in the showroom either, so there is obviously great value in auditioning in your system if possible.

Best of luck.

[mikeymad]

Why can't more first posts be like that.. thanks Smeck..

[charelanen]

I'm VERY interested in Romulus. So it would be great help for me to know what you think now about it,4 months later.

[Smeckles]

Hi,

Late again, as usual. Sorry about that.

I've returned the Romulus to the local dealer a few months ago and I still miss the huge soundstage. I'm trying to figure how I can get either it or the Pandora back from the dealer to re-audtion it to verify.

In the mean time, I've looks at a few other DACs with USB compatibility and none of them captures this ability so I am thinking it's either unique to Aesthetix or I was really on a bad bender during the review period. I hope it was the former, otherwise, my binges are bad enough that they affecting my girlfriend's hearing as well as mine.

Other units I've had since then (all were either floor demos or used, so were already well broken in by the time I plugged them in):

Antelope Gold w/Voltikus: most of my listening was with digital tracks and where I could, I used higher resolution files (24/96k). The Voltikus had a speaker width stage presentation, not the huge blooming one that the Romulus threw. The Voltikus sounded like it was pro-audio gear. It's fast and quick (noticable in my fast-n-quick system). Bass was controlled well. There was almost a "density" to the music (I really don't know how else to describe it), that made instruments fill up what soundstage was there. Instruments sounded bigger, like they occupied more space, but it really wasn't a lifelike recreation. Just big, "dense" images to make the space between the speakers sound crowded. With the Romulus, each instrument had it's own space but no more. Instrument sizes were roughly the size I would expect from a real instrument, and if they were miked or mixed with space between them, the Romulus would project that too. Volitkus made such spaces disappear. If Phil Spector designed a DAC for his Wall of Sound mixes, he'd like this one. The Voltikus' sound didn't make instruments "bleed" into each other, they just sounded ... dense. Some folks may like this type of presentation, and I do think highly of the unit, but for me it had the really hard job of replacing the Aesthetix unit.

Also, the Voltikus USB is flaky. Several times the IMac refused to see the Voltikus (failed handshake?), and I had to fully unplug the Voltikus and reboot the IMac to make it play nice. It worked fine once that ritual was finished. Turning off the Voltikus had no effect. Unplug was the only fix here. I've seen some other people online had the same issue, and they too came to the "unplug or die" answer, so I'm not alone with this.

Berkeley Audio Designs Alpha DAC (series 2) and USB (leashed together with Crystal Cable AES/EBU): Again, pro-level audio gear. Soundstage was speaker width. Bass got handled well on good recordings but lower resolution files sounded like low resolution. Meh. No sweeteners here.

My notes on this one features the words "percussion" or "percussive" again and again. This unit got into the details of "musical impact" more than any other unit I've heard. The impact of a stick hitting a snare head was so clearly presented, it was like a different recording than what I was used to. Likewise piano hammers hitting the strings in quiet space (I played the opening Aria from Andras Schiff's version of the Goldberg Variations again and again to verify this). I like the effect even more so on acoustic guitar with the pick audibly scraping across the strings and then the note or chord sounding. With the impact, there must also be decay. The Alpha DAC didn't highlight or emphasize decay the way it did with impact, thankfully. So much analytic detail would wear thin pretty quickly and distract from actual music. In otherwords, the DAC was very competent in the overall presentation with its real strength being this presentation of impact speed(?).

I don't know how Berkeley Design does this whether it's processor speed, software implementation or what. A really nice unit, but I can't say I would pay the ransom my local dealer wants for it, especially since the Pandora would be cheaper than the Alpha + USB box combo I would need. Add the Crystal Cable and I'm getting into some bucks.

Finally, I'm demoing a used Weiss DAC 2: This is connected via Firewire rather than the USB, so right away it's different than the other 3 I've talked about. I'm only a few days into the review now, so I'm still deciding what I like about this. This is an older unit (at least 3 years old, I think) and has been replaced in Weiss' product line by the DAC 202. In fact, my IMac refused to connect with the DAC 2 when I first plugged in because the supplied CD with Firewire drivers were actually obsolete. Fortunately, downloading the most recent drivers and installing them was a piece of cake. In short, with good quality (mp3 >256k or AIFF/WAV files) , it sounds great. Musical and really pleasant. Soundstage is the familiar speaker width, but everything flows well. On some of my older low-rez files or poorer recordings, the bass gets very "woofy" (electric bass, bass drum etc.) and I find myself steering away from those tracks to higher-rez waters. That is really the worst thing I can say about it so far, which is pretty good indeed.

The Weiss is very very interesting, and the price of the used DAC 2 is less than half of any of the other units here. The DAC 202 that replaces it on the other hand, is more expensive than any of the other units listed here. I do need to listen to it more, but if I can live with the saggy bass on some of the older tracks I have, I might get this for a bedroom or office unit.

But based on everything I've heard, I still long for the Aesthetix. Damn if that thing wasn't like audio crack. I really do need to get that one back to verify. If you can audition in your system, give it a try. It sounded so much different in my system than what I head in the dealer's showroom (in an all Aesthetix system no less). It was night and day with my own system on the plus side.

One last note: I am trying to get a Bricasti DAC unit to listen to but it's proving difficult since there isn't a local dealer I can beg for it. I will try though.