Clash of the titans: Benchmark DAC-1 vs. LeVry DA-10 vs. upgraded Zhalou DAC
Mar 5, 2006 at 5:30 PM Thread Starter Post #1 of 845

Ori

Member of the Trade: Oritek Audio
Joined
Oct 23, 2005
Posts
674
Likes
11
Those fotunate souls who have attended the recent San Francisco meet had a lot of toys to play with... One of those was a Zhalou DAC with its cover removed. People passed by it on their way from the impressive AMB display and the mighty Orpheus to the left, ignoring this unit. In the end, lamark took possesion of this unit for the rest of the day.
How sad!!!
confused.gif


This "review" summarizes two short sessions, conducted on two different speaker systems. The first session compared the Benchmark DAC-1 and the built-in DAC of the Grace m902 amplifier to the upgraded Zhaolu DAC. The second session just concentrated on upgraded Zhaolu vs. a Lavry DA-10 DAC.
The Zhaolu has hundreds of hours on it; the Benchmark is an "old" unit as well (mfg late 2004) and should be considered fully broken-in. The Lavry DA-10 is a brand new unit with only a few hours on it. To conduct a more precise test we should have given all the units ample time to break in and conduct a long-term listening and usage test. However, I feel that the differences we observed are quite large and a long-term test would not change fundamentally the overall conclusions.
Since this is a headphone forum, the question that one might want to ask is why these sessions were conducted on speaker systems?
The answer is very simple. We briefly listened through the headphone jacks of these DACs and through a headphone amp with a Senn HD-650. Headphones have a lot of sonic attributes, but they also lack many details that are extremely apparent on a speaker system. Soundstaging and spatial information do not come even close to the detail of a good speaker system. I've noticed the same phenomena with cables, so this is not unique to source equipment. The three DACs sounded very close through the headphones, almost to the point that it made no difference which one you pick. Perhaps our headphone amps are not good enough; perhaps there is another sane explanation. Bottom line is that both speaker systems were more revealing than the headphones available to us, so we proceeded with this approach. Additional comments about headphone sound added by zdogg, see the end of this article.
The assumption is that the better sounding DAC through a speaker system would be the better DAC for a headphone system. That is an assumption, of course. Listeners have different preferences and sonic tastes and perhaps different needs. If all you want is a good DAC just for your headphone setup then your pick may be much simpler.
With all these caveats, let's proceed to the actual test. We wanted to avoid noise problems, so the DACs were disconnected from the power line when not used. You might say that they need to warm up to sound their best, but then we have to contend with digital noise issues. We preferred this approach.
We picked very few tracks and repeated them with each setup. In many cases we would listen to one DAC for a few seconds, pause and immediately change to the other DAC. This takes a few seconds and then level matching is required as well, but it is the best method we could come up with. It should be also noted that we used 44.1 KHz material only.

Modded Zhaolu vs. Benchmark DAC-1

We had three people in attendance for this test: zdogg, lamark and ori. The general consensus was that the DAC-1 sounded somewhat harsh and lacked soundstage precision and tonal accuracy, with unrefined treble and “loose” bass, when compared to the Zhaolu and the Grace amp's internal DAC. In fact, we preferred the built-in DAC in the EastSound CDP to the DAC-1.
When comparing external DACs, we sometimes used one coaxial and one optical cable to facilitate quick switching. We swapped these cables between units, just to make sure the connections were not the main factor in what we were hearing. I can safely say that the coaxial connection was superior on all units, but that's old news. To further eliminate noise issues, the DACs were powered through a separate passive line filter (Monster hts3500, OneAC cp1105 or TrippLite LS-600b).
Again and again, the DAC-1 could not match the openness, spatial accuracy and finesse of the Zhaolu. It came across as a "fuzzy" presentation that would easily become boring and fatiguing if the audition had taken more than a few minutes at a time per DAC. The Zhaolu always presented a very coherent, musical and lively sound. It was hard to find any faults in it reproduction. Perhaps the very upper harmonics were just a tad smoothed out and the bass a tiny bit shy vs. our "other reference", the venerable Revox B-225. For those who don't know this unit, find one and have a listen... It represents world-class - regardless of budget!
Bottom line of the first session is that we were all scratching our heads... Why is the DAC-1 a "reference" unit for some noted audio writers?!

Modded Zhaolu vs. Lavry DA-10

Next comes the Lavry DA-10. The unit was not available for the first session but zdogg has done a preliminary comparison with the DAC-1 and it too smoked the DAC-1. zdogg reports hearing more refined treble, deeper soundstage, tighter more defined bass and improved musicality overall; also, the built-in headphone amp was vastly superior to DAC1’s. The battlefield was now set for the clash of the titans - Lavry DA-10 vs. the mighty Zhaolu.
On first look, the Lavry DA-10 looks very professional. Nice front panel and a sturdy case. A peek inside revealed a very basic commercial power supply. The modified Zhaolu has a massive toroid transformer and a massive power supply PCB with many power devices regulating the supplies of different parts of the circuit. Nothing fancy, but well thought out. The Zhaolu also included a clock upgrade, which added its own supply and regulation, using a dedicated tap from the transformer. As it turned out, this has no function for the audio signal reconstruction. What appeared like a nice touch ended up being useless, though not harmful...
Next I noticed two basic crystal cans on the Lavry main PCB. The cans were not even grounded (LAvry claims that would have a negative effect, which I disagree) and their marked frequencies were 22 and 24 MHz.
So now we established that the Zhaolu has a robust - and hopefully "clean" - supply, but not necessarily a better on-board clock reference. What about the guts, the DAC and the other electronics?
The Zhaolu utilizes the AD1852 converter chip. The Lavry has an AD1955. Both are capable of 24/192 and both are mass production devices. Which is better? Your guess is as good as mine.
The analog section is very different. The Lavry has a volume control circuit in the signal path. According to their explanation, it consists of analog muxes (perhaps the ADG508AKR chips) and a resistor ladder. The Zhaolu has none. That is a fundamental difference between the two units. The Lavry DA-10 is designed as a DAC/preamp. The Zhaolu is a DAC only.
The analog circuitry of the Lavry consists of discrete transistors with a few OP275 opamps, presumably outside the signal path. The upgraded Zhaolu has two OP2604 opamps and a single DY2000 (replaces an OPA2604 in the basic configuration). Later we found experimentaly that the third opamp implements a DC servo, albeit the DC offset still measures 30-100 mV in a few samples.
Both provide XLR output that we haven't used in the second session. The Lavry was connected through a modified XLR/RCA adapter, which floats the unused pin. This Tara Labs adapter was found sonically better than the Hosa adapter previously used by zdogg. It yields decent sonic results in a speaker system.
To sum up, the analog section of the Lavry is more complex, but the implementation of the amplification stages is closer to a purist design. On the other hand, the Zhaolu resorted to a basic opamp circuit with upgraded opamps. It actually has sockets for the DIP8 devices, so rolling opamps would be a breeze. It's also important to mention that the Zhaolu has a relatively high (30-100mV) DC offset on its outputs, which may present a problem for some gear (check your manual or with the manufacturer). To my taste, opamps always have a sonic signature that usually results in masking of high treble information, so we are left with a question: which implementation degrades the sound more?
My experience with digital volume controls is not vast, but implementations such as a Levinson 380S preamp left me not too impressed. Given that volume control does not change at a high rate, the relay-based implementations are much superior to the chip-based ones. However, cost and packaging issues are a factor for the Lavry designers and I can understand their choice. I would have liked to see a true bypass for that part of the signal chain, perhaps relay-based... Speaking of relays, the Lavry DA-10 does include a few of these near the analog outputs. It is by no means a cheap implementation!
The headphone amp section of the Lavry is very impressive, especially compared to the two-opamp implementation of the Zhaolu. The latter is a total waste in its original form and simply does not match with the quality of the DAC itself. You may outfit the Zhaolu with a discrete headphone amp at additional cost, but we didn’t have that option on our unit.
The Lavry implementation has the opposite problem. It seems that the headphone amp is a higher quality design than the DAC that drives it... Bad source and a good amp or good source with a bad amp yield the same mediocre sound quality. The upgraded Zhaolu can at least be used as a credible DAC, so I'd favor its "compromise"...
So, after a long dissection of what's inside, the question that remains is how do they compare as pure DACs?
The Lavry DA-10 has many switches on its front panel. One of them allows selecting a clock source between the internal oscillator and the transport clock (with two options, wide and narrow PLL filters). Trying all combinations, I was saddened to see that the cheap transport had the less jittery clock than the internal oscillator of the Lavry, yielding a better spatial precision when on the “narrow PLL” setting vs. the “crystal” setting. Good clocks are the real foundations of digital sound reproduction and the Lavry misses that essential piece of the puzzle. The Zhaolu offered an unbelievable soundstage precision, left to right and front to back as well. Perhaps a better transport would produce better sound with the Lavry, but the Zhaolu seemed immune to transport jitter altogether. No options to confuse you - what a great concept!
A drum piece brought to the spotlight the slight boominess of the Lavry vs. a much more neutral bass definition of the Zhaolu. The tone was just "right" on the Zhaolu and a bit "electronic" on the Lavry. Voices tend to compress more on the Lavry. To top all that, the Zhaolu also had a better treble extension. In some recordings, it produced more sibilance, but overall I believe that it's a matter of treble rolloff on the Lavry vs. the "right" treble balance on the Zhaolu. This is a very interesting fact. The Lavry DA-10 has the more sophisticated analog circuitry with its discrete signal path, yet the overall sonics do not reflect this fact at all.
Another factor to note is that the volume setting on the front of the Lavry affects the sonics; we found 42 to be best. Maxxed out at 56 (and adjusting level on the external preamp), the Lavry sounded noticeably worse, despite hearing no evidence of clipping. The Lavry responded to different power cords (a basic Quail power cord was not “sufficient”), but nothing tightened the sound up enough to rival the Zhaolu.

Detailed listening Impressions
(copied from a follow-up in this thread)

I've used mainly the following tracks for close scrutiny:
  1. Male vocal: Sting/Angel eyes, Taj Mahal/Honkey tonk women
  2. Female vocal: Vivaldi/Tu M'Offendi (a great operatic piece...), Cowboy Junkies/Helpless
  3. Instrumental: Bach/Partitas "presto"/Rachel Podger (violin), Chopin/Barcarolle in F sharp Major, op 60/Vladimir Shakin (piano), Take 5/Manhattan Jazz Quintet (mainly percussion solo)
As you can see, I go mainly for the "least congested" pieces when comparing equipment. The vocals are very distinct and cover a huge range. I really like Taj Mahal with Jim Cotton on the harmonica. The harmonica sound exhibits, well, a lot of harmonics... Sort of a controlled "broken-up" vibrant sound. this particular recording is extremely clean. You'll find it on HOB "Songs of the Rolling Stones" CD, a must-have for the R&B fan.
This clearly improvised piece has a lot of ambience and tonal qualities that are very distinct. In addition to the harmonica and the throaty vocalist, the metalic dobro guitar adds even more fun to the "party", with that metalic sheen that edges on harsh brightness, but not exactly...
The vocals through the Zaholu are extremely clean. The Lavry exhibited a slight compression. In a minimalist piece like this one, it is really easy to tell soundstage and depth. The Zhaolu lets you hear where the vocalist mouth is and where his guitar rests. The Lavry was much more "fuzzy".
At the end of this piece you'll hear Jim cotton expressing his sheer joy in a 4-letter word that the mix engineer and the producer thankfully elected to leave on the track...
eggosmile.gif
You can tell where he's located on the far left, but the Zhaolu also tells you that it's far back in a 3D room; the Lavry is just "vague".
I'm deliberately dissecting this piece, given that it demonstrates a lot of the qualities that I've mentioned before, but one might suspect that the Zhaolu is perhaps too analytical?! Not at all. In fact, the extremely coherent presentation makes you want to forget that this is a review session. You want to just lay back, relax and simply enjoy the terrific music!
The Chopin piece is one of my favorites for equipment reviews, and in particular the first ten seconds of it... There is a lot to be learned from the tonal qualities and harmonic reproduction of a piano stacato. A one note "hammer" gives you a complex sonic "space", filled with mechanical sounds, fundamental tone, string after-vibrations (as in after-shock...) and echoed upper harmonics. One big mash that is easy to mess up - as many systems often do...
confused.gif
Again, the Zhaolu was just more believable. It reproduced the piano note very cleanly and put it out in front of you, perfectly aligned with all the harmonics, without exxagerating anything, without ommiting a squeak. This was just as good as I've ever heard this piece, to put it in perspective.
The unusual rendition of Take Five as a drum solo is in this distinct mix for obvious reasons; no review is complete without a drum piece... The Zhaolu reproduces size and locality extremely well. The Lavry was again more vague but also sounded more "boomy" in the midbass and overall gave you a feel of being more rolled-off and less coherent than the Zhaolu.
Percussive tones are much more distinct and believable through the Zhaolu. There is one transition in that piece, where the kick drum is struck repetitively for 3-4 seconds. It was distictly a kick drum through the Zhaolu, and something between a drum and a wood block through the Lavry. This particular drum piece gave a $15K Meitner a run for the money - and here you have a $300 el-cheapo play it in a way that leaves you smiling from ear to ear...
k1000smile.gif

And that just about says it all.

The verdict

Although I ain't no judge, I'll volunteer
rolleyes.gif
my own subjective opinion.
We played several instrumental pieces and switched back and forth between the two DACs with a consistent overall impression that the upgraded Zhaolu is just the more coherent of the two. The Lavry DA-10 is indeed better than the Benchmark DAC-1, based on my recollection, but it falls short of the refined presentation of the Zhaolu. This leaves me with a simple conclusion: if you are in the market for a new DAC, and I'd take the risk and say "at any price point",
very_evil_smiley.gif
you will do yourself a big disservice if you do not check out the mighty Zhaolu.
biggrin.gif
“Weighing in” at half the cost of the Lavry DA-10, the modified Zhalou is easy on your pocket and represents a true bargain in today's market!
k1000smile.gif
eggosmile.gif
biggrin.gif

My friend Bill sent this example to me and asked for my opinion. Given the price and upon a basic inspection of what's inside this unit, I wanted to make a few modifications and take it to the next level. After listening to this unit against the credible competition, I'm not sure I want to touch it at all. Doing so might be nothing more than a first step down the road of miniscule and very diminishing returns...
Before I conclude this write-up, I’d like to clarify that I have no connection with Zhaolu or the Hong-Kong dealer (listed below). I don't even own a Zhaolu myself and perhaps never will, for "other" reasons. However, I know a bargain when I see one and I just can't keep such a "secret" to myself. Enthusiastically recommended!
eggosmile.gif


More Listening Impressions

Being the audio geeks that we are, more listening is a natural thing to do on Saturday nights… zdogg added the following comments:

I've done some more listening at home now. With speakers in my system, the differences [Zhaolu vs. Lavry DA-10] sound more subtle (my ears may be toast now as well) but I think most of the findings still stand, and at worst there is less difference between the DACs (favoring the cheaper one...). Overall, there is a sense of slight defocusing with the Lavry compared with the Zhaolu; the highs sound more rolled off and slightly “syrupy”, and soundstage precision is lacking, although depth is good. Dynamics likewise seem muted a bit with the Lavry, and everything sounds “softer” all round. Bass is less controlled and less precise. With headphones, the Zhaolu is clearly superior in terms of treble extension and bass definition, as well as overall clarity of soundstage/instrument positioning. Interestingly, on careful listening I clearly preferred the Zhaolu -> Ray Samuel SR-71 -> HD-650 to the HD-650 straight out of the headphone amp on the Lavry! Just as noted in the previous comparison on the speaker systems, the Zhaolu headphone setup bettered the Lavry with a more precise spatial soundstaging, better and more controlled bass and snappier dynamics…


What’s in a Zhaolu…

The basic unit was already described in another active thread: http://www6.head-fi.org/forums/showthread.php?t=139545
A few Hong-Kong dealers seem to offer this unit and the upgrades. Search this thread for contact information and pricing. In general, I recommend the following configuration:

Basic Zhaolu DAC: $80-180
Opamp upgrade: SKIP!
Clock upgrade, 1-2PPM SKIP!
Power supply upgrade $48-80 (value: undetermined)
IEC AC adapter SKIP!
assembly fee $0-25
Speedpost (express mail, 2 days) $45-58




"Closing Arguments"

As usual, YMMV... Just in case, the flame suit is already on...
Many thanks to my friend Bill for letting me borrow his treasured unit, to lamark for the first session and to zdogg for the many comments and mainly for his great spirit of adventure...
 
Mar 5, 2006 at 5:35 PM Post #2 of 845
When my modded DAC-ah comes back, let's pit it against a modded Zao at the Nat. Meet. If the DAC-ah wins, by inference it beats the DAC-1 and DA-10, Score!
icon10.gif
Thanks for the impressions and comparo, we don't get many comparos here using speakers so that's refreshing.
 
Mar 5, 2006 at 6:45 PM Post #3 of 845
Very interesting comparison, makes me wish I had heard the Zhalou in my own rig at the meet, as the barebones tube amp it was hooked up to was, well less than impressive. What power cords did you find had an impact on the Lavry, and what difference did they make? I guess the next adversary for the Zhalou should be the G08.

zdogg, you going to chime in, or did Ori summarize your thoughts in totality?
 
Mar 5, 2006 at 7:08 PM Post #4 of 845
We only had the Quail and a home-made cord, based on commercial 14AWG twisted/shielded design with hardened vinyl cover (for vibration resistance) and commercial-grade plugs on both ends. The latter improved the bass focus and resulted in less boominess. It had also a very small effect on overall soundstage precision, as you would expect from the reduction of line noise and RF pickup it offers.
I suggest that you should try different shielded cords. The line filter is mandatory as well with the LaVry. If you are more adventurous, try soldering a piece of stiff wire to the oscillator cans and solder the other side to the PCB ground. This ought to remove some noise and would also add mechanical stability to the crystals.
BTW, that "barebones less than impressive" tube amp extracted the best sound out of the R-10's...
eggosmile.gif
 
Mar 5, 2006 at 7:27 PM Post #6 of 845
Hmm interesting stuff to consider, still not sure that I want to mod it though.

Quote:

Originally Posted by Ori
BTW, that "barebones less than impressive" tube amp extracted the best sound out of the R-10's...
eggosmile.gif



Hmm, well I guess that is where personal preference comes in. I though it made the R10's lose quite a bit of their focus and tightness, in exchange for a more smoothed-over and bassy presentation. I can see how that might make the R10 sound better for some sorts of music to some people, but I feel that by doing so you're compromising away what the R10 does best, and making it much less special than it could be. If I wanted that kind of sound, I'd just listen to the HD650 with a copper cable.
 
Mar 5, 2006 at 7:30 PM Post #7 of 845
Quote:

Originally Posted by Iron_Dreamer
Hmm interesting stuff to consider, still not sure that I want to mod it though.



Hmm, well I guess that is where personal preference comes in. I though it made the R10's lose quite a bit of their focus and tightness, in exchange for a more smoothed-over and bassy presentation. I can see how that might make the R10 sound better for some sorts of music to some people, but I feel that by doing so you're compromising away what the R10 does best, and making it much less special than it could be. If I wanted that kind of sound, I'd just listen to the HD650 with a copper cable.



Or use Foobar Convolver and give up even less.
wink.gif


-Ed
 
Mar 5, 2006 at 7:42 PM Post #8 of 845
Zhalou !!!! =)

man this makes me want to buy another Zhalou fully modded for backup.! in case mine ever kicks the bucket. hah.. and before prices increase..

it is a sweet piece, and even though I never did comparisons to the above mentioned and eastsound,... I knew in my mind it had to be damn close.

sweet review!
 
Mar 5, 2006 at 8:13 PM Post #9 of 845
Nice review Ori .I knew when i first heard the Zhalou i had to get the word out.I could not beleive how good the sound was for such a small amount of money.The fully modded Zhalou just improves an already nice sounding piece.

And i agree about lawrence,he is a music lover himself and has been right on the money with his advice.Highly recommended


best regards
 
Mar 5, 2006 at 8:53 PM Post #10 of 845
Very nice comparison!

Is this Zhaolu only available by contacting Mr. Chan by e-mail or is there some website or something? I was going for DAC-AH but this Zhaoly is getting more and more interesting.
 
Mar 5, 2006 at 9:15 PM Post #12 of 845
Quote:

Originally Posted by Patu
Very nice comparison!

Is this Zhaolu only available by contacting Mr. Chan by e-mail or is there some website or something? I was going for DAC-AH but this Zhaoly is getting more and more interesting.



Heh i ordered modded diyeden great-march dac from lawrence to Finland. When he shipped it after final mods and testing saturday it was at customs monday (2 days like he said).

Hm btw i think clock upgrade to zhalou is nowadays 1ppm instead of 2ppm in review
smily_headphones1.gif
 
Mar 5, 2006 at 10:35 PM Post #13 of 845
Quote:

Originally Posted by Edwood
Or use Foobar Convolver and give up even less.
wink.gif


-Ed



Foobar Convolver? There is no plugin for foobar2000 that could possibly reproduce the warm, natural feel of a tube DAC with a bunch of complicated programs! DSPs work entirely with ones and zeroes and are thus incapable of understanding the human condition! This is completely different from a tube DAC, where the tubes are the most important part and add warmth directly into the line-level signal!
 
Mar 5, 2006 at 11:21 PM Post #15 of 845
nice review, now i really cant wait to get mine.
icon10.gif
 

Users who are viewing this thread

Back
Top