Best DACs Series: Mola Mola Tambaqui

• The problem of spatial presentation
• The problem of bass in real rooms
• The problem of dynamics
• The problem of digital distortions
• The problem of recording standards
• The problem of visual images

While much progress has been made on signal purity from input to output, we assert that further gains there, given how much has happened in the last 75 years, are not likely to be the source of believability breakthroughs. And, we assert, most of the work on the 6 big problems requires a different and more systemic framework than is normal for component engineering.

So, we approach the Mola Mola Tambaqui with that in mind. I’m trying to cover it from the perspective of more than ‘is this a good DAC and is it competitive for the money?’ I think it is, but I think a bigger reason to consider spending $13,500 (the US price of the Tambaqui) on a DAC is if it solves or addresses a difficult and fundamental problem of audio. If that kind of money makes your head explode, you may want to watch my Bluesound Node Icon review for a $999 streamer/DAC with good performance and a great feature set.

I thought the Tambaqui might fit well in this series of reviews because its lead designer was Bruno Putzeys. I have found that many of the candidate DACs for this series have ‘genius-level’ designers behind them, maybe because these are the kinds of engineers who go ‘outside the box’. In any event, Putzeys has worked extensively on the advancement of Class D amplifiers, power supplies and DACs with Grimm, Hypex, Kii, and Purifi.

For the Tambaqui, Putzeys designed a unique DAC circuit or set of circuits and software really. He points out that a key goal of the design was a signal-independent noise floor, because noise that varies with the signal is a giveaway that something is wrong – the result sounds digital. But he makes the significance of a custom design even easier to understand when he points out that most DACs (whether delta-sigma or ladder type) have at least one failure mode where certain kinds of signals produce odd, clearly not music-related distortions like whistles. There exist a bevy of such distortions in digital, including aliasing and pre-ringing as well as modulated noise. Some of these distortions occur in the decimation process used to take the very high-res (e.g. 384k) master recording and reduce it to standard digital file sampling rates and word lengths. These may not be addressable in the DAC. But some errors can still occur in DACs.

To address these problems, Putzeys designed a special PWM (pulse width modulation) set of switches, stable clocking and custom digital filtering determined by listening tests, all of these optimized by sampling rate. Plus, a “secret” analog filter design.

Sound Quality

As we point out, I can’t and you probably can’t, reason from design details to sound quality. So, we listen.

I need to articulate the key audible failure modes of DACs that are part of this adventure. If you are accustomed to thinking in frequency balance terms, these may seem odd, although once you notice them, you can’t unsee them. Or unhear them, perhaps more to the point.

I believe many audiophiles and music lovers notice them, in fact. There is, for example, a passionate interest in analog playback via vinyl (and a tiny group interested in analog tape). You can write this off as a misguided passion for a bygone world. But I’ve spent a lot of time listening to high-end vinyl, and some to tape, and there is a there there. It isn’t hard to hear. Unpacking what happens is harder.

Now before you all get your knickers twisted, I will also say that analog, especially vinyl, has distortions many, many times higher than digital by most measures. So, I’m not arguing that analog is better. I’m saying analog shows that digital has some flaws relative to the absolute sound that ideally would be corrected if we could correct them. If you get the chance, spending some time listening to very good vinyl, or in a concert hall, will make the thoughtful among you wonder if you can have your digital cake and eat it too.  That’s what this series of DAC reviews investigates.

A core revelation of listening to analog (or, even better, listening to real instruments in real space) is that certain distortions become more obvious. Cymbals, bells and the trailing edge of female voices often show a hashy, edgy sound that isn’t right when referenced to the absolute sound. These distortions can make digital sound “bright” to some listeners, although frequency response may be flat as a pancake (aka, the ear/brain is not a microphone and D/A converters are not amplifiers). For some listeners, this kind of distortion is a distraction from believability that, while it is occasional rather than constant, tells the ear/brain something is wrong. It is also distraction of musical attention, somewhat like an audience member coughing, or tics and pops on vinyl.

The Tambaqui does a very good job with these signals. Cymbals, to take my favorite example, are rendered with greater naturalness and detail than is common with chip-based DACs. To give you a visual analogy of sorts, with normal DACs it is as if the signal from some instruments, which of course unfolds over time, is scrambled. If we imagine the 3-second-long signal consisting of 5 parts: 1,2,3,4,5, it is, with some DACs, as if the playback is 1,3,5,2,4. The sounds are there, but the order is only somewhat recognizable as the instrument you know. The signal is seriously messed up. With the Tambaqui, high frequency instrument signal coherence is better preserved, independent of the mastering choices about how bright recordings should be. I think the Tambaqui sets a benchmark here, in my testing so far.

If I seem to be holding back on full gushing praise, it is because I think some of these distortions may be there on recordings, and I don’t want to mislead you into thinking DACs, and the Tambaqui specifically, can address these, although hope springs eternal. An example is the aforementioned trailing edge of female vocals. On many recordings there is a tearing sound on these tracks that just doesn’t occur in real life. But I note that this seems more a phenomenon in recordings from 20-40 years ago than newer recordings. I don’t know when studios switched to 32 bit/384k sampling, or if that is even what has generated the improvement, but this time-dependence makes me think this is a recording issue. In any event, the Tambaqui doesn’t solve these issues.

Another digital issue that good live music and analog reveal is a tonal thinness to many instruments. This isn’t a typical tonal balance thing – something that occurs due to the general tilt (or lack thereof) of the frequency response of most DACs. Rather it is a shift in the timbre of many instruments, with lower harmonics, whatever the frequency, seeming slightly diminished. This may be the midrange frequency version of the issue I discussed with cymbals. But its sonic features are quite different in that this distortion seems mostly subtractive, whereas the difficulty of DACs with high-frequency transients seems additive.

In any event, the Tambaqui again improves the sense of instrumental body and richness compared to many of the chip-based DACs I have heard. I also compared the Tambaqui to the even more expensive Berkeley Alpha Reference DAC Series 3, and I think this issue of instrumental body is even better addressed by the Berkeley. I will address this in a bit, but I think most of you would find these differences very subtle. Of course, add up ten subtleties and you’re in a different sonic world.

Another quality I must mention is the sense of “organic” sound wherein the instruments and voices sound naturally coherent. If we studied the engineering of digital, we would observe that a lot of effort is spent on the timing of signals. That’s because, unlike spreadsheets, digital data for music consists of bits in time, not just bits. With music, the timing is part of the data, just like in writing, the punctuation is part of the writing – it isn’t just words. It may be that small timing errors (the clock for hi-res audio is at 5 mHz or higher, conversions involve timing ‘guesses’ etc.) in A/D conversion are perceived as inorganic because the human brain has great familiarity with actual, natural sounds. Which don’t have timing errors.

This naturalness is an area where I think the Tambaqui shines. I will caution that, at first, I thought the Tambaqui might sound a little dead. But I have had the experience before, and did again with the Tambaqui, that very high-resolution devices encourage lower listening levels and a small change in volume to the reference level often gets you back to the liveliness you want in music. But…minus some of the distortions that you realize are part of many components and which generate liveliness at the price of naturalness. The Tambaqui doesn’t buy into this tradeoff, it just presents the music in a way that you focus there.

I do observe that DACs differ in their spatial presentation as well. While we’re on the subject of timing, perhaps different DACs preserve or damage phase information that is essential to spatial presentation? I don’t know, and there are many filter tradeoffs and artifacts to consider.

I came to think that the Tambaqui was doing a very good job spatially. Stage depth in particular was superbly conveyed and is often a weakness you don’t notice until you hear it. I wasn’t initially as comfortable with the left-right staging of the Tambaqui, but after another 30 or 40 test tracks, I came to think that it delivered a more precise placement in that dimension, and sometimes that meant the recording engineers had certain artists hard left or hard right. This, if I’m correct, is a classic audio tradeoff: do you want what is on the recording or do you want your equipment to distort it euphonically? With the Tambaqui, good and great recordings were more convincing in both the X and Y planes, so it really seems to be a tradeoff. Sorry.

As a final point, I sense that a lot of people are trying to tune and tweak and select their audio gear to address issues in the frequency domain. The want more bass or less treble or less bass or less upper midrange or… You get the idea. If the issue really is the frequency balance of your system, I don’t think DACs are a particularly effective way to address this. It is your call, but at least for the Tambaqui, and most other DACS we test, any frequency balance deviations from flat are minor in comparison with what some people seem to want (there is a reason for this which is the subject of another investigation). Now, if you want to avoid some of the distortions that call attention to treble, that is different and the Tambaqui can help there on some recordings.

Summary

I want to put all of these attributes of the Tambaqui in the context of “how big a deal is it”? I can’t tell you what you value or what your budget is. But it is possible to say things about the magnitude of differences. Here’s a way to think about it. In our model of the problems of audio, we have input to output signal purity at the center. This is mostly the domain of incremental changes that with any given component switch will generate relatively little difference, but where progress across an entire system (cables, streamer, preamp, amp etc) often add up to something quite noticeable and valuable. Then we come to the Big 6 issues of audio mentioned above. Addressing most of these produces obvious and dramatic differences (again, whether you value this is another matter). But the problems of digital live in an in-between state: I think the gains are obvious but they don’t happen all the time because they are triggered mainly by certain signals. So, you could say that what a DAC like the Tambaqui does is subtle. You could equally say that what it does is dramatic and valuable. I’m in the latter camp, but that should be irrelevant to you. You should hear it and decide for yourself.