There is little doubt that DIY audio equipment has the potential to offer incredible value. If you go into a local Hi-Fi shop and pick a piece of gear at random, it’s likely that the cost of parts only accounts for maybe 20% or less of the total price. In some cases it is probably much lower. This applies to everything in the audiophile market, from CD players to speakers to amplifiers. Let’s use a $1000 pair of monitor speakers as an example. Dealer markup is easily 40-50% of the price already. So let’s be generous and say the manufacturer gets $600 when they sell these monitors to the Hi-Fi shop. Now we have to factor in overhead for the speaker company to keep running: basic things like paying their employees, a lease on their facility, utility costs such as electricity/water/gas, insurance, as well as more complex variables such as marketing. Let’s again be nice and pretend those things add up to about 25% for each pair of these monitors sold. That brings us down to $450 actual cost per pair. But of course they are not in business to break even; they want to make a profit, or else it is not worth doing. Let’s say they intend to make 30% on each set they sell, which is really not much. That brings us down to less than $300 cost for each pair of speakers. Even if we could allocate 100% of that to the materials (which we can’t because we need to think about packaging them, shipping them, etc) we would still need to split it between drivers, crossover components, and cabinet materials. So it’s easy to see how a $1000 pair of “high end” monitors could contain drivers that are worth less than $100, and that was us being very generous with our calculations.
That little exercise means nothing if we don’t actually have the ability to construct our own equipment. Speaker construction requires at least a small amount of ability and equipment for making the cabinets, and all DIY projects assume the builder can handle a soldering iron. The good news is that you can do a compromised “mostly DIY” type build using things like pre-built cabinets, or you can even purchase the services of a more experienced builder to do the whole project for you. In many cases this will still result in a better price to performance ratio than you would otherwise achieve by going in to the shop and buying a commercial product. So if we invested $500 into a pair of DIY monitors, it would likely feature higher quality drivers and parts than the $1000 commercial product, for half of the price.
I’m mentioning all of this because the item I’m reviewing here is essentially a pre-built DIY project. The company is called Gigawork (often mistakenly referred to as Gigaworks) and the product has been called various names. A common title is the Gigawork Big DAC, to distinguish it from a smaller and less complex DAC they also sold at one point. There is a massive 380+ page thread about it over at diyAudio.com. I believe the DAC started its life as a simpler and lower priced version consisting of an assembled board and separate power tansformer, where the end user needed to wire it up inside the case of their choosing. That version contains a lesser quality output stage and USB receiver, and is still sold in kit form. There are all kinds of tweaks and upgrades that people have come up with to turn that good sounding design into a very high performance unit.
At some point the manufacturer must have seen a demand in the market for an even simpler project, and also wanted to capitalize on the full potential of their design. So they started offering an upgraded version in completely finished form, branding the case with the name “Lars Audio DAC1 MKII”. It goes for $330 shipped on eBay and that’s the version I’ll be discussing. From here on out I’ll just call it the Lars to make it simple.
At the heart of the Lars is the Crystal CS4398 DAC chip. As far as I know this is still the top of the line offering from Crystal, and it is used in many different products. The CS4398 sometimes appears in fairly low priced gear, but it is also at the heart of some very high end players from the likes of Bryston, Esoteric, Ayon, McIntosh, Classe, and Electrocompaniet. So obviously when done right it has a lot of potential.
Supporting the CS4398 is the Crystal CS8421 asynchronous sample rate converter which converts all incoming data to 24-bit/192kHz. This chip is implemented in such a way that the user can disable it if desired, by removing the entire daughterboard. I strongly preferred the sound with upsampling engaged on every DAC I’ve tried, so I did not utilize that feature, but it is nice to have as an option. I do wish there was some type of switch on the front panel to make this adjustment on the fly though. As it stands the user will have to open the case remove the board. Although it probably isn’t too difficult, it obviously would not be good for on-the-fly, back and forth comparisons if you had to take that step each time.
On the input side of things, the CS8416 digital audio receiver (again from Crystal) handles incoming data in either coaxial SPDIF or optical toslink form. It claims to accept all data rates up to 24/192, although I was only able to verify up to 24/96 due to the limitations of my Squeezebox Touch transport. The third input is USB and it is handled by the increasingly popular Tenor TE7022L which means it handles everything up to 24/96 with the exception of 88.2kHz streams. This is a limitation of the Tenor chip, not the Lars DAC in general, and I didn’t find it to be a huge issue because I only have a handful of albums in that format.
Lots of attention appears to have been placed on the power supply of the unit. Indeed roughly half of the PCB is populated by the power section, and that’s not including the space taken by the R-core transformer. The design boasts 5 individual stages of voltage regulation and a large row of smoothing capacitors.
The output stage uses a pair of Linear Technology LT1361 op-amps. These are high speed op-amps and fairly well respected by the op-amp rolling crowd. They are also socketed so the user could try their hand at various other op-amps if they so desired.
Capacitors seem to be from various sources. I see Nichicon, Elna, Sanyo, and Chemicon. All other parts appear to be of good quality too, including Dale resistors and small Wima capacitors.
Externally, the unit is very simple. On the front left side we find a single knob for power on/off switching, and a matching knob on the other side for selecting one of the three sources. Out back we see the three digital inputs, the single ended RCA outputs, and a standard AC plug.
Upsampling circuit installed on the daughterboard
The board can apparently accomodate various DAC chips
Extensive power supply section
CS4398 DAC chip
The Lars is surprisingly well built. The internal design seems fairly straight forward, but the more you look the more you notice that it is packed with components. The workmanship seems as good as many commercial DACs I’ve encountered and there is no evidence of corners cut or sloppy assembly.
Externally, the Lars manages a sort of simplistic classiness, but still seems to whisper its DIY roots. The case is really the giveaway. I don’t mean that it looks bad though; the sort of brushed aluminum texture on the top and sides, the thick front panel, the vented top area, the tight consistent panel gaps, all say quality. The knobs move with a satisfying click. I really just think the case is similar to DIY projects I’ve seen in the past, and that’s what keeps reminding me. It just has this universal look to it… as if it could be a DAC, an amp, a big outboard power supply, or anything else you needed it to be.
About the only part that I can think of to take issue with are the feet on the bottom. They are actually quite fine as far as functionality, and sit nice and level, in contrast to some other budget gear I’ve encountered. But the cheesy gold appearance is obviously just a thin layer of some sort of coating, and it feels like I could peel it right off. It’s not the gold itself that causes the problem; my Violectric V200 amp has similar gold feet but they look and feel very classy. It’s just the execution of these particular gold feet. I’d rather they just have left them black. The good news is that they sit underneath the DAC so you rarely have opportunity to inspect them closely, and from a distance they look acceptable. In a less conspicuous place is the “Lars Audio DAC 1 MKII” printing on front. It looks solid enough but I really wish they had made it smaller, or used a different font, or done something else to make it a little more exciting. I realize that they probably didn’t want a blank faceplate, but a small logo or something would have been a much better choice in my opinion. There is a small red LED that lights up to indicate power on status. I was hoping that it would turn green to indicate signal lock, but that is not the case. There is a row of similar LEDs inside the case in roughly the middle of the PCB. They light up too but I have no idea what their function is. If they are there for troubleshooting or to indicate system status, that message is lost on me. The last thing to mention is the slight “clicking” sound that sometimes comes from inside the unit. I assume it is from some sort of relay. You hear the noise when switching from one input to another, but also when playing a track recorded in a different resolution. So when I left my Squeezebox Touch playing music at random, the clicking sound appeared more often than I anticipated due to the mixture of CD standard and hi-res music in my library. It is a very minor distraction at worst but something worth mentioning.
I removed all the screws on the top plate for these photos - note that they are countersunk for a smooth look when the screws are in place
Pardon the fingerprints
The Lars DAC arrived in a generic shipping box with no Gigawork logos or markings. Inside I found the unit itself enclosed with a mix of packing materials and some wadded up newspaper. It was protected rather well but was not pretty. The experience was more like buying a used item off the classified section here rather than a new item from a dealer. Still, it arrived unharmed, so I can’t complain.
The only thing included was the DAC itself. I don’t even remember getting an AC cable, although I could be wrong about that. I have about a dozen AC cables laying around for some reason so it was not a problem for me. Indeed I’d rather them not tack on the extra $5-10 that it would cost to give me a generic power cable. Something I did somewhat miss was a user manual. Despite the unit being very straight forward and self explanatory, I’m just accustomed to flipping the pages of the manual when I get new gear. I was especially concerned with the information about disabling upsampling, and since there was no manual to explain it, I had to email the seller. He answered my question satisfactorily but it would be nice if I had that information from the start.
This is the associated equipment I used for evaluating the Lars DAC:
Source: Lexicon RT-20 universal player, custom music server with a Squeezebox Touch frontend
Amplification: Violectric V181 and V200, Yulong A100, DarkVoice 337SE, Matrix M-Stage
Headphones: Sennheiser HD800, Lawton LA7000 Lite, AKG K701, custom Yamaha YH-1 orthos in Sennheiser HD600 shell, Unique Melody Merlin, Westone AC2, JH Audio JH13
These are just the impressions of one guy. I do these reviews for fun, not profit, and I don't claim to be any special authority. Many people have agreed with my assessments of other gear but some have also disagreed, and I totally respect that. We all hear differently on a physical level and we all have different preferences as well, so I think it is almost impossible for one person’s impressions to apply to every other person. As with all my reviews, I hope you enjoy reading them and I hope they help our hobby to some extent, but I don't pretend that they are anything more than my opinion.
My initial impression of the Lars DAC was that it was very involving. It had a warm, analog flow to it that really seemed above par for the price class that it sits in. I was especially taken with the low frequency extension and control, which right away stuck out as the star of the show. I’ve put in many hours of listening by now and my opinion at this point essentially remains the same.
Starting with that low end; it is deep and very impactful, yet still very tight and controlled. Not boomy or muddy, just big sounding. It almost seems ever so slightly boosted, just because it draws your attention so much. But that could just be an artifact of the general sound signature. In any case, I thoroughly enjoyed the presentation on all types of music. From The Ohio Players to James Brown to Bill Evans to Metallica, the Lars DAC was consistent in giving everything plenty of drive and authority.
Mids are very flowing and organic. Smooth, but not overly so. When something is recorded poorly, the Lars will probably show it. But there is an inherent ease to the sound which is different from a lot of other DACs in this price range, and even some more expensive models. This isn’t one of those ultra detailed DACs where you get to hear a pin drop on the sidewalk outside the recording studio. Those sorts of micro details are simply not the point. Instead, the Lars concentrates on being very musical and rich. The end result is that an alto sax sounds distinctly different from a tenor, or a soprano, or a clarinet, and yet all of them have a clean organic flow that should be enjoyable on practically any headphones or speakers. It’s a great balance between “just enough” detail and great timbral accuracy, and it simply sounds great.
Given my description thus far, you might expect to find slightly rolled off highs. That is at least partially true. Extension does not go as high as some others I’ve heard, and once again there isn’t as much fine detail on display. But what is here is very smooth and very clean. There is a distinct lack of grain that makes it sound fairly lifelike despite being a little recessed. As I listen I don’t really find myself missing much until I compare it back to back with another DAC. But in some instances this presentation is actually beneficial. There is some music that, when played on some headphones, is just too bright for me to enjoy. Often these are older recordings that I otherwise love, and the Lars DAC can help take the edge off. I realize this is not the ideal presentation, but short of using a much more expensive setup it might be the best option. In contrast to this forgiving nature is a somewhat limited ability to portray excellent recordings to the height of their potential. That’s just the price that must be paid to achieve this level of smoothness at this price level, and for me it works much more often than not.
One of my favorite aspects of the Lars DAC is in terms of dynamics. With a thinner DAC you would have to crank the volume to extract more emotion out of the performance. The Lars is great for low or medium level listening because things still sound full and rich. But they stay clear and smooth at high volumes too, so the user has choices.
Soundstage is very nice, with great layering and realism. It is not the largest I’ve ever heard, nor even the most specific with regards to imaging. But it still somehow manages to convey a certain liveliness which I find difficult to describe. A comparison would be listening to a studio recording, and then switching to a quality recording of the same band playing the same song live in concert. The studio version might be more spacious, with more well defined placement of the instruments, but that doesn’t necessarily make it more “real” sounding. There is a certain charm offered by the live take that is hard to capture in words, but you just know it when you hear it.
I found no major differences when switching between toslink, coaxial, or USB inputs. In fact if anything I sometimes thought USB actually sounded superior. That was a first for me. Obviously there are variables involved since one input is being fed by a CD transport and the other from a file on a hard drive. I tried my best to find a conclusive instance where I might be able to choose the USB input in a blind test, but I don’t think I ever got to that level. Still, the superiority of the USB input is something that I did continue to notice, albeit mildly and inconsistently. The key point here is that USB is not inferior, which is a common problem among lower priced DACs (or even higher priced models unfortunately). This must be a byproduct of the CS8421 and its reclocking function.
Ignoring matters of build quality, convenience options or features, etc, and focusing purely on sonics, I find the Lars to compare favorably with many of the DACs I have around the house. It is confidently superior to some of my cheaper DACs like the Hot Audio units or the Audinst HUD-mx1, which are all nice in their own right. Yet it is clearly inferior to some of the higher end models like the Audio GD Reference 7 and Wavelength Cosine, which of course cost a lot more. In the end I found that it compared best with two somewhat similarly priced competitors: The Matrix Cube and the Yulong D100.
The Matrix Cube costs about $60 less than the Lars. It also has a nice built in amp function, making it possibly a better value. But since we are strictly dealing with sound quality, the Lars ultimately comes out ahead. The two models are actually more alike than different, with the Lars sounding something like an enhanced version of the Cube. Specifically the Lars has a more defined, extended low end, as well as the ability to maintain more composure during extremely complex passages. Both strike a good compromise between detail and musicality, but the Lars just seems to have a slight edge when it comes to richness of tone. The Cube maintains the lead when it comes to soundstage spaciousness, which I think is its biggest strongpoint. But as I mentioned earlier, bigger is not always better, and some might prefer the more direct approach as presented by the Lars. So despite the Cube remaining a phenomenal value for an all in one device, the Lars does have a more pleasing performance overall.
The Yulong D100 is roughly $120 more than the Lars, once shipping is considered. It is my benchmark for reasonably priced DACs and holds its own even when compared to more expensive units. The Lars is an interesting comparison. On one hand the Yulong is more resolving, more linear, and more accurate. It produces a more technically correct rendering of your music without being too analytical or sterile. The Lars sort of comes at things from a different angle, not necessarily aspiring to technical perfection but rather appealing to the emotions. It’s like listening to live music in a club, where maybe the acoustics aren’t perfect and the background noise is not ideal, yet it has soul and character and you don’t want it to stop. Both DACs are neutral enough in the grand scheme of things, with the D100 maybe falling slightly on the brighter and cooler side compared to the warm, organic sounding Lars. If I didn’t know better I might almost think there were some tubes hiding in the case on the Lars, while the D100 definitely has the stereotypical solid state sound. Ultimately I think the D100 remains my pick for a reasonably priced DAC with high end performance, but the Lars in an interesting alternative. I can easily see some people choosing the Lars over the Yulong. For example, people who still prefer vinyl to even very good digital playback will find a willing companion in the Lars DAC. It would probably be a perfect match for someone like my dad, who is all about vinyl, tubes, and vintage horn speakers.
The Lars Audio DAC 1 MKII, AKA the Gigawork big DAC, is a very high value piece of equipment. Despite its DIY roots, it appears about as well made as any commercial product in its class. And perhaps because of the DIY design, it seems able to offer an intriguing sound that transcends its price point by a significant margin.
To sum up the sound: it is organic, rich, and dynamic, with exceptional low frequency performance, beautiful flowing mids, and clear inoffensive highs. It is not the last word in detail but is very musical and very accurate when it comes to timbre. For those who are more concerned with the overall presentation on a macro scale, and are willing to sacrifice some inner detail to achieve that goal, the Lars DAC is the best I’ve heard at what could be considered a reasonable price. It actually sounds like a junior version of my Wavelength Cosine which was quite a bit more expensive.
I find it interesting that one design consisting of a sigma-delta DAC, dedicated upconversion module, and opamp based solid state output, can sound rather similar to a non oversampling, non upsampling, passive I/V design with a tube output stage. The Wavelength has more detail and a bigger soundstage but the general feel remains the same. This just goes to show that we should try to keep an open mind and not always assume things like “tubes are always warmer than solid state” or “that specific DAC chip/opamp has a certain sound”. The end result in a piece of audio equipment is always dependant on the implementation and not just the individual parts used. In any case, the Lars DAC seems to use reasonable quality parts in a well thought out implementation, and the result is great sound. I can very easily recommend this DAC for folks looking for this type of sound signature.
There is another interesting option though: I see that the seller still offers the DAC in semi-assembled kit form. It sells for just $120 shipped. It appears to be mostly the same, but uses an inferior USB implementation and cheaper opamps. Since the opamps are socketed, that would not be an issue. You get an assembled board and what appears to be the same R-core transformer. It would be up to you to find your own enclosure and wire it up, but that could potentially be an even cheaper option for someone who is semi-capable of some mild DIY action. But for the rest of us, who either don’t have the time or the skill, this pre-built Lars DAC option is probably the way to go.
In my review of the Matrix M-Stage, I gave it sort of a twist ending when I revealed that it was basically a clone of the highly respected Lehmann Black Cube Linear headphone amp. The Lars has a similar twist although the story is more complicated.
Some readers may already know this, but there has been much speculation that the Gigawork DAC has a strong connection to the more expensive Decware ZDAC-1. That product sells for $875 although it shows “sold out” at this moment so it may be discontinued. A link to that product: http://www.decware.com/newsite/ZDAC1.html
There might be some variation in things like capacitor brand, but the circuit is clearly the same and the resulting sound should be very close. Of course, people have differences in opinion about that sort of thing. It’s up to the reader to look at the Decware website and the pictures shown there, and decide what they think. Something of value to read is this article by Decware which sort of addresses the issue: http://www.decware.com/newsite/paper151.html
In it they claim a few things unique to their version: a special vibration damping process when they install the board into the chassis, and a special coating that makes the board nearly waterproof. They also say they completely re-solder everything once the assembled board arrives from China. Again I’ll leave it to you to decide what you think of their statements. Pay extra attention to the section where they use the words “Chinese parts are complete junk” when describing “RCA jacks, IEC connectors, chassis, and just about everything else”.
I inquired with Gigawork about the relationship with Decware. Their version of the story is that they base their design on a reference Crystal board, with some of their own modifications. They have the product manufactured at a factory in China. They used to supply the boards to Decware who would simply swap in better opamps, add the power and input selection switches, and throw it in their own case. When Gigawork read some statements by Decware that they felt were disparaging (possibly something like the comment about China from above), they stopped supplying the boards. Decware then contacted the factory directly and bought the remaining supply. Gigawork has since switched factories. The Lars DAC uses the same opamps as the ZDAC, and since it does have power and source selection switches prewired it remains nearly identical to the ZDAC.
I don’t have any experience with the Decware ZDAC so I can’t definitively make the claim that the Lars DAC sounds the same. But after writing this review, I browsed online to find impressions of the ZDAC. Descriptions of the sound it makes sound eerily similar to what I’ve written about the Lars DAC. Reviews have been universally positive and most called it a very good value for the $875 price.
I don’t know which company is telling the truth, or if it sits somewhere in the middle. Decware implies that they put some deliberate errors in the schematic to avoid being copied. Gigawork claims they are coming out with an all new DAC in the future to distance themselves from the whole fiasco. One thing I will say is that if Gigawork is telling the real story, it would explain why Decware is out of stock: They don’t have any more boards available to them. But that’s just speculation.
Whatever the truth may be, the fact is that the Lars Audio DAC1 MKII gives you the opportunity to own a very high performance DAC for a relatively low price. Imagine this: one person buys the Lars DAC and the Matrix M-Stage, an excellent combination, for less than $600 shipped. Another person buys the Decware ZDAC-1 and the Lehmann Black Cube Linear for around $2000. They basically have the same equipment but one paid nearly three times as much for it. He gets more pride of ownership because he has respected name brand gear, but he doesn’t get any more actual performance. It’s up to you to decide who you think made the better buy, but I know which one I’d choose.
The Lars is cheaper than almost everything else on this table (except the headphone stands) but it is a very solid performer and does not feel out of place when used with more expensive gear
Edited by project86 - 5/27/11 at 12:04pm