Has anyone had a chance to compare A/B between Bartok 2.0 and Mola? I know in Youtube there were some comparisons but it was 1.0 Bartok which was not as good as 2.0. Jay Audio Lab considered Mola as the best under 20k dac…Hope to have your opinion.
Bartok and Mola Mola Tambaqui are not really comparable. Bartok is a DAC streamer, Tambaqui is solely a DAC.
Bartok v. 1 and v.2 are not different units. This denotes a software upgrade which is free and is carried out using dCS Mosaic. All new Bartoks are v.2 of course and I expect that almost all used Bartoks have been upgraded to it. If you come across a v.1 (n which I doubt) then you just press the appropriate button on Mosaic and it will upgrade.
All new Bartoks are Bartok Apex. This increased the price but I do not know if Tambaqui remains in a comparable price bracket in your market ( wherever that may be).
There is no best! There is only the best for you , your needs and your tastes. That goes for everything.
1 Like
Hi Pete
Informative reply. For the original poster it might pay to note the Mola Mola is ROON ready. In other words it provides access to both Qobuz and Tidal streaming services without need for any additional external boxes. The same applies to any local ripped files as well.
At the end the best advice would be for the original poster to demo both units.
Gregg , many thanks but unless I have been misunderstanding things for Roon Ready you need Roon. That requires a subscription to the service and another box in order to run Roon Core.
Yes, auditioning both is the only reliable way to make the choice.
Oh - am I bad🤣
Actually Pete, as Gregg indicates, like the Bartók, the Tambaqui has a built-in Streamer with an Ethernet port for direct connectivity to the Home network. The Tambaqui also comes with a built-in Headphone Amp, and Pre-Amp functionality, and just like dCS, Mola Mola has its own (Android & iOS) App to control the Tambaqui.
So, functionality wise, the Tambaqui really is the right comparison against the Bartók.
Dat Vu,
I’ve not done a direct comparison of the Tambaqui against dCS, but I did spend a full day with the Tambaqui and Aries Cerat Kassandra DAC, flipping between them (when I was auditioning some speakers).
I found the Tambaqui to be very bland, flat and uninspiring, contrary to almost every review I’ve read. My suggestion to you, same as Pete and Gregg, is to go listen for yourself, don’t rely on someone else’s opinion.
2 Likes
Thanks Anup. That means that the review that I checked before writing was crap. Who could believe it ? 
1 Like
Interesting comment. I passed this information onto a friend (current Vivaldi stack owner) who has always been enthralled with the Mola Mola.
I did find one piece of interesting information (if, like Pete, I am to believe an online review). It was stated the Mola has separate dual mono DACS. I found this interesting in light of the recent Varèse
development.
Thx all of you for sharing. I am very surprised hearing that because I also read many good review about Mola. Mola is very smooth, natural, no harsh at all even with high volume. But the sound stage is a little bit small. That’s what Jay Audio said. Do you feel the same.
(It’s quite hard to find a Mola in my area to listen but I will try)
Technically, any DAC that’s built on discrete components and not with merchant silicon chips (ESS, AKM, TI etc), like many of the R2R DACs, are architecturally “dual mono” internally within a single box, including the Mola Mola Tambaqui.
That “dual” nature is limited to the actual D-to-A conversion stage and does not extend to the Power Supply, or the input and output stages; which means both channels share a single power railing and the incoming data channels are separated on-board, and the output amplifier circuits share a single power supply etc.
In other words, they’re really not mono at all, at least in the full traditional sense of the meaning.
The width of the Tambaqui’s sound stage was fine, and clarity/transparency was very good. but it was flat in a sense that it lack 3-dimensionality/depth, and it made every track sound tonally bland in my opinion - at least when directly compared to the Aries Cerat Kasandra DAC (an R2R DAC built on AD1865 chips).
4 Likes
Excellent points above, here are a couple of other things to bear in mind:
- You probably read the Herb Reichert review of the Tambaqui and comparison with his long term reference Bartók in Stereophile. Long story short he thought there were things to like with both of them:
Bear in mind that was the original Bartók with the 1.0 firmware.
- Since then the Bartók has been updated with 2.0 firmware (software upgrade, free of charge) and Bartók Apex (hardware upgrade, charged), both retrofittable to older units. Herb kept the Bartók and still uses it as his long term reference (his review of the Apex upgrade here).
- Should you wish to, the performance of the Bartók can be further elevated with the addition of an external Master Clock (Lina, Rossini or Vivaldi at three different price levels). Not an option with the Tambaqui.
- The Bartók has a choice of 7 reconstruction filters allowing further fine tuning of the sound to your taste, the Tambaqui only offers one.
- The Tambaqui lacks single-ended outputs (XLR-only) which may or may not be important to you.
Per the comments above, only you will really be able to say which one you prefer the sound of. However objectively, one would have to say that while the Tambaqui is by all accounts an excellent DAC the Bartók is more flexible, more tunable and more upgradable.
4 Likes
I have owned both Tambaqui and Bartok2.0. Did extensive comparison of their dac function with external Eversolo streamer. Bartok is more versatile DAC with mappings and filter choices. Bartok mapping 3 renders sound that is quite similar to Tambaqui except the Tambaqui has slightly more bass slam. I felt that bass was more natural on Bartok.
Kept the Bartok and sold Tambaqui in the end.
7 Likes
Hi struts001. Your sharing ís excellent source, comparing very detail between Mola and Bartok 1.0. After reading it, I definitely continue to use my Bartok 2.0 
Hi Supa, great to have you here. Maybe you are the only who have both Bartok 2.0 and Mola at the same time.
I am happy to know that Mapping no. 3 brings the same sound as Mola. What about filters??
Besides could u share more detail about the sound differences between them?
I’m a little bit puzzled by this comment, wondering whether @supa meant Filter 3?
In his measurements for Herb Reichert’s review John Atkinson found the filter in the Tambaqui to be a linear phase type with a roll-off that reaches full stop-band suppression at 26kHz (where the red line flattens out):
The Bartók (with 2.0 firmware) offers six different filters (+1 for MQA) with Filter 1 having the steepest rolloff (full attenuation at 22.05kHz) and Filter 4 the gentlest (full attenuation at 30kHz):
Unfortunately John doesn’t show the graphs for Filters 2 and 3 which lie somewhere in between but if they are evenly spaced that would mean Filter 2 bottoms out at around 24.7kHz and Filter 3 at 27.35kHz so the Tambaqui filter lies pretty much mid-way between the two (hence my comment above).
The Mappers randomize the selection of the resistors in the I/V conversion stage according to different algorithms. This subtly affects the overall noise floor, not just the out-of-band noise. Maybe one of the Mappers in combination with Filter 2 or 3 sounds more like the Tambaqui than the other?
Mapper 3, not filter 3 although it is shown in my previous photo.
Filter changes are subtle in comparison with mapper change. My personal filter preferences are 3 and 5. I listen to Jazz mainly.
When you say favourite filters, remember that filter numbers are only that. The characteristics of a filter depends on sample rate as well as digital format.
So filter 5, for example, is not the same for a 44.1 kS/s file compared to a 96 KS/s file. Hence current dCS DACs actually offer 42 filters not 6 (not sure that total includes MQA). Much on this relates to the GoldenSound contretemps earlier this year.
Hi Pete, Accordibg to John’s measurements Filters 1-4 all behave similarly accounting for the different sample rates. At 44.1kS/s F1 starts rolling off at just below half the sampling frequency and then rolls off sharply. F2-4 roll off successively earlier relative to FS but thereafter have progressively gentler slopes. So they are ”families” that you could describe as ”anatomically similar”. However he found F5 to behave quite differently at different sample rates. Ho hum.
Yes. For “group” 1-4, not only does frequency of the sampling rate change the filter regarding its “turn-over” point for Nyquist image rejection but also its slope and transient response, Filter 5 at 44.1 has an asymmetric response. However filter 5 at 176.4 and above has a Gaussian response with the asymmetric response now at filter 6.
Of course the group of filters for DSD are also numbered 1-5.
Skim reading of the details published so far suggests that they may have adopted a clearer scheme for Varese.
The scheme to date seems to cause all kinds of confusion judging from posts here over time and from external reviews.
Nice HPA. I have a Luxman integrated and always wondered what a Luxman HPA would be like. Today I only use the Bartok for headphones.
1 Like