That’s awesome that you’ve had some experience with HQPlayer; I’ve dabbled with it and had similar experiences. But lately, I’ve been really stoked about this little workaround using Roon for converting the PCM to DSD and then letting my dCS DAC take it from there. The difference in sound quality has been undeniably impressive, and I’ve been really happy with this discovery.
Roon also has some pretty advanced algorithms that are updated much more frequently, so I feel like it’s a solid option. If you’re interested, I highly recommend giving it a shot and seeing how it works for you.
And it’s fantastic that you enjoy dCS’s Vivaldi Upsampler! However, it’s good to continue to keep an open mind, even when we think we’re satisfied, and explore other options too. So I suggest giving Roon and dCS a go yourself to see if you experience the same impressive results as me. Why not give it a try? Another pleasant effect of this pipeline is that keeping everything in the same DSD domain with static parameters when entering the DAC spares me from the unpleasant and unenjoyable task of dealing with the filter-switching problem. Hey, who knows, right? Maybe once Lina gets updated to 2.0 with the latest mappers, I won’t even need to develop my own workarounds anymore. Fingers crossed!
I tend to test out everything I can get my hands on. I’ve got a Roon perpetual license from years ago and use it daily - especially when I explore new album releases. I’m not a fan of any DSP processing on Roon.
Well @Anupc, it’s great to hear that you’re an adventurous music lover, always testing out new things. But, as they say, you can lead a horse to water…and when it comes to DSP processing, I totally get your point - sometimes it’s best to let the music do the talking. But hear me out; every once in a while, a little processing can pleasantly surprise you. That’s what’s great about the technology and competition we have today; things keep getting better and better, and we end up with more options than ever. And let me tell you, that is definitely a good thing in my book.
With all that said, it’s fascinating how some of us have agreed that letting the PCM to DSD conversion happen outside the DAC can yield better results than doing it all on the DAC itself. Naturally, it’s always important to take a cautious approach when testing new settings, especially when it comes to something as subjective as audio quality. But, like @PaleRider, I think it’s a good idea to leave it as is for a while and compare it to my usual settings to make sure it’s really worth the switch.
Exactly. In my experience, the more processing you apply to sources outside of the dCS DAC, the more you’re going to lose subtle cues in the music, especially with live recordings; lost spatial cues, soundstage compression, delicate sounds captured near the noise floor gone, etc. Little nuisances that make live performance interesting will disappear.
There are benefits to off-board processing of course, like room-correction for the lower octaves for example, but other benefits are typical what I consider as low-fi ones like better slam, edgier, etc., which may excite some listeners, but delicate nuisances are totally lost.
That’s what happens with HQP, which is far more sophisticated than the DSP Engine within Roon [for upsampling]. So, no, IMHO, Roon upsampling is not better than what the dCS Vivaldi Upsampler does where these little nuances are well preserved.
In any case, probably a good idea to open a separate thread to discuss off-boarding processing (rather than driving this Lina thread off-track ) ), I’m sure there’s lots to chat about, including room correction processing etc.
@Anupc I totally understand where you’re coming from in terms of applying too much processing, potentially causing the loss of subtle cues in music, particularly in live recordings. However, I have quite a bit of experience using HQPlayer and Roon. As a result, their algorithms deserve more credit. They can sometimes preserve those delicate nuances while enhancing overall sound quality to varying degrees - although it often results in emphasizing certain aspects over others.
As for Lina and how it relates to my initial point and the topic at hand, it’s worth noting that, at present, there are no minimum-phase PCM filters available to choose from, which is where the real magic of this particular experiment lies. I’m really hoping that dCS will address the filter-switching issue promptly and introduce some new 2.0 features to Lina, such as high-quality minimum-phase PCM options. Not only would this significantly improve the user experience, it would also allow for even more opportunities for sound customization.
By the way, @PaleRider, you might want to try upsampling in Roon to the max PCM with a precise minimum-phase filter and then letting the dCS DAC handle the DSD upsampling. I’ve been attempting this workaround and found it quite compelling, further highlighting my excitement for the potential improvements a Lina 2.0 update could bring, including new filters and mappers. It would be fantastic to no longer have to resort to workarounds to get the most out of our Lina DAC. Needless to say, I can’t wait for 2.0 to arrive! AND SOON!
This is the first time I have heard of this issue, and I am sorry to hear you are having problems. It is a confirmed problem with Lina DAC that is currently being worked on by engineering with a fix coming by way of a software update soon, but I have not seen any other reports of Rossini units displaying the same behaviour - please feel free to drop me a private message on the forum or an email to [email protected] and we can go through this issue.
I know I am late to the party on this but it did peak my interest. Provided we can rule out interface jitter from the discussion by assuming we are only talking about asynchronous formats like USB or network, here isn’t any reason why DSD Upsampling would cause more jitter than DXD Upsampling. The DSD Upsampling is an extra stage added on at the end of the standard DXD Upsampling process, but the Ring DAC will oversample any incoming data (PCM or DSD) to DSD rates at 5-bits in the Lina DAC anyway (synchronously of course, so 2.822MHz for 44.1k based content and 3.072MHz for 48k based content) even if the unit is set to DXD Upsampling.
As such, if the higher rate itself was enough to induce jitter in the DAC, it would be present on any and all formats / settings. I can’t speak to the HiFi News review measurements, but there is no technical reason why DSD Upsampling specifically would induce jitter not otherwise present.
All theoretical of course, and the ear/brain is a funny thing - none of this is intended to counter what you are hearing.
Thank you, @James, for providing valuable insights into the potential lack of difference in jitter between DSD and DXD upsampling. Although it doesn’t entirely explain my experience with Lina’s dull and unengaging sound when using the Network setting with DSD upsampling, I find it compelling that using max-rate PCM or pre-converted PCM to DSD produces significantly better results. Utilizing this (hopefully temporary) workaround suggests there is still plenty of room for dCS to improve its algorithms and firmware. As you allude to mentioning the HiFi News study, it’s not always about measurements (accurate or not) but the audible experience and pleasure thereof. Your technical explanation is, of course, greatly appreciated.
In light of this, I am keen to learn more about potential improvements, such as a Lina 2.0, that could address this issue and other related problems, including the lack of minimum-phase choices for upsampling and the filter-switching problem. With continued development and improvement, dCS can further enhance the performance of its products and offer even better audio experiences for its customers.
Agreed. More filters are definitely a good thing. Note though, among the rest of the dCS family, there’s only one minimum-phase filter, Filter 5 (just to set the right expectation )
There’s no objective best filter as each makes compromises in either the temporal or frequency domains which different people will have different personal preferences. On my Vivaldi stack for example, for PCM I alternate between Filter 5 - for some rebook (Rock mostly) - and Filter 6 for a wider range which I find especially good for Jazz and Live recordings. While for DSD I always stick to Filter 1 for the widest bandwidth.
The HiFiNews review seems to suggest Lina’s F2 Filter is F4 on the Bartok, but they didn’t show any of the measurements
Interesting. Stereophile did something similar a couple of years ago when they drastically reduced the space for JA’s measurement section, and graphs become tiny.
At 2:54 why dCS decided to use the top of the range Xilinx FPGA in the Lina. I know that it was mentioned before here but it’s good to hear it directly from dCS employees.
I wonder how many outside this community (who may knock the investment cost of a dCS DACs), know how much these FGPAs cost. The street price of the 3 FGPAs in the Bartok, Rossini and Vivaldi is more than the retail cost of many standalone DAC products.
I just found that the price here has dropped to € 411,71 a piece, with a lead-time of 17 weeks. I am sure dCS has taken care of ordering large batches, so they will be well stocked.
chuckle … you should have been looking at this during the “Human Malware” outbreak … the cost of high end FPGAs went absolutely insane, like double or triple that (and don’t even start on availability but I’m sure that you can work out that if they tripled in cost then they weren’t exactly available to excess)!
It does amuse when we get asked at events “How many taps does it have”…
I tend to test out everything I can get my hands on. I’ve got a Roon perpetual license from years ago and use it daily - especially when I explore new album releases. I’m not a fan of any DSP processing on Roon.
) ), I’m sure there’s lots to chat about, including room correction processing etc.
