Hi, so there are map 1, 2, 3 I’m my Rossini
Which one of the is the old standard that was before 2.0 update and which one is the one introduced at the 2.0?
Hi Givara,
Mapper 2 is carried over from 1.x and Mappers 1 and 3 are the new ones.
Thank you, but which one is the new “default” and which is the “tube like”? Sorry for noob questions )
IIRC all our current products ship with Map 1 enabled by default, but feel free to interpret that as a starting point, one has to start somewhere! Maps 1 and 3 both operate at double the frequency of Map 2, which provides more headroom for noise shaping. They use different rule sets to achieve the same ends by slightly different means, which represent slightly different trade offs between noise and distortion.
Neither is right or wrong per se (which is why we offer you the choice) and neither is intended to be “tube like”. Indeed the way the differences manifest may be system-dependent. When setting up systems for demos, which I do reasonably often, my own workflow is to start with Map 3. But there are certain amplifier/speaker combinations where I find I prefer Map 1. The only “right” answer is the one you prefer.
1 Like
Thank you, I’ll try both then
Dear Andrew, can you plz advise when it’s better to use DXD, DSD and DSDx2? I mainly listen to CDs on my Rossini (DAC+CD all in one) and is there any connection between DXD, DSD, DSDx2 and mappers? Like DXD usually better with map 1 and so? I understand it’s very subjective but would like to known what you prefer.
Hi Givara,
All the signal path settings (Filter, Mapper and Upsampling) result in small changes to the “flavor” of the resulting sound. You will have to believe me when I say that if one were right or “more right” than the others we really wouldn’t offer a choice. However as well as accommodating personal preference they also allow a small amount of tuning for different types of music and different downstream systems.
In particular the Filters all represent different compromises between frequency domain (image rejection) and time domain behavior which can mean that some work better with certain types of music, as described in the manuals:
That said IMO, even for an expert, trail and error is really the only way to determine what sounds ‘best’ in any given situation (although rest assured Filter 1 is a great all-rounder).
Fwiw my own listening (for work and pleasure) is now dominated by streaming, so most of the sources are PCM files (ca 90% std res, 10% hi res) and I tend to keep them that way so I default to DXD upsampling. My personal go-tos for these use cases are Filter 1 & Map 3, but as I said I can find that Map 1 works better for me in certain scenarios. But plenty of folks here default to Map 1 as well. On the rare occasions I am playing a DSD source I tend to keep it as DSD and default to Filter 5.
4 Likes
Again thanks, for me it’s valuable info.
uUst to add a little to @AndrewS excellent post, you also asked:
No, they are separate things. Mapping concerns the allocation of bits to the current sources in the Ring DAC. Upsampling multiplies the base sampling rate ( DXD) or recodes the data to a delta sigma format (DSD). This is a one bit format with very high sampling rate.
As Andrew says, what you prefer is up to you.
I live on a diet of mostly newly recorded classical music - virtually all of it is delivered commercially as “high res” so 24/48 up to 192.
I’ve found these settings work well for my listening: Map 1 and Filter 5 for 176,192 and DXD and f1 for 48, 88 and 96. I use separate f5 for 44. DSD is filter 5.
Andrew has me curious about Map 3 so i may give that a try.
@AndrewS thanks for sharing the revised chart above this is helpful.
2 Likes
AS you probably know I mainly try not to comment on subjective findings. However as I have similar tastes to you:
I would suggest trying F4 for frequencies 48,88 and 96, You may not like it of course as this is all a matter of personal taste. Nevertheless I will stay with my suggestion. Perhaps some people may be put off by dCS’ suggestion that is best with rock? It’s not limited to that in my book.
I would also recommend people not to settle on these choices immediately. Take time to revisit the choice over a period of time. In many cases the difference may seem very subtle and not obviously affecting those audiophile measures of bass , midrange or treble response, imaging etc. but can lead to an ill defined feeling of dissatisfaction after time if a better choice could have been made.
1 Like
Thanks @PAR for the suggestion I appreciate it! I will give it a try. Out of curiosity, which filter do you use for 176 and 192?
For these high PCM rates I use the asymmetric filter which is F6. One thing to watch out for; this is a similar filter type to F5 for 44.1, however for rates at and above 176.4 F5 is now a Gaussian type ( not to my taste).
That’s just me of course. Feel free to disagree 
.
I have landed on mapper 3 F1 and DXD upsampling since my recent acquisition of the Vivaldi upsampler. It used to be mapper 1 so very dependent on the rest of the setup. The upsampler I have on F5 setting, however that is very subtle. Might have to try some more but I have it pretty dialed in to my taste currently.
I am contemplating trying out different AES cables too since my setup changed. To get some bearing I started back with the standard cabling. Changing over to BJC Belden 1694A video cables for clock duty gave a nice improvement though.
3 Likes
1694A is a portal for mapper 3 in my experience. Try it with F4 for richer details within the notes. It works for me with a Lina&clock with a Naim NC NAP250. My XLRs are wide open Gothic Audio Nublu.
2 Likes
I believe that “Tube like” statement comes from John Quick’s quote captured by Stereophile’s coverage of the Rossini 2.0 Update;
Of the new mappers, MAP 1 has a slightly higher proportion of third-order to second, which is considered more ‘solid-state’ sounding; MAP 3 is the inverse, with more second-order to third, and is considered more ‘tube-like’ and benign.
Ah, thanks Anup, had completely forgotten that quote.
The first part of what John said is (objectively) correct, Map 3 does have a slightly different distribution of 2nd and 3rd harmonics. However the levels of both are extremely low and the difference between Maps 1 and 3 is only a couple of dB, so the view here is that the characterization as solid state vs tube-like is highly subjective. While decibels are fact these kind of similes are very much opinion and may lead readers to draw wildly different and possibly unintended conclusions.
I guess I personally share JVS’s own view on this:
2 Likes
Thanks Andrew you are absolutely correct to make these points.
Yes tubes may produce more second order distortion products compared to solid state designs where third order distortion may be prevalent. However the converse of a design with a higher amount of second order harmonic distortion than third may not necessarily sound “tube like” ( whatever that means) at all . That is to confuse correlation with causation.
This is slightly off topic, but I would like to ask @AndrewS (or other member of this forum ) if, in is experience, someone likes more Mapper x or filter y in a Lina DAC, the same person would have the same preference with a Bartók, Rossini & Vivaldi DAC’s.
In other words, does the ”flavor” translates to different topologies in the dCS family?
Does Apex changes things?
Great question Pedro! My first thoughts are:
- In the same system, between Bartók, Lina, Rossini and Vivaldi (non-APEX) I would expect preferences to be consistent. The DAC/output stages are identical across B/R/V and substantially similar in Lina.
- In the same system, between the three APEX products (four if you count Varèse) I would expect preferences to be consistent by the same rationale.
- I have heard it mooted that the consensus might have shifted from a slight net preference for Map 1 to a slight net preference for Map 3 in conjunction with the APEX launch. However per my earlier point system differences make this really hard to track accurately. It might be the case however I suspect this is a hunch based on very sparse datapoints.
I’m really interested to hear folks’ views on this, especially those who have upgraded pre-APEX units. Did your Map preference change?
2 Likes