Improving SQ of Streaming

I‘m afraid I have same experience :rofl:
(And I pimped up my roon like hell…better Ethernet cable, dedicated PSU, roon Nucleus, isolated stands, super powercable, roon on MC vs Win vs dedicated hw etc etc). Imho nothing beats Mosaic/upnp

2 Likes

Hello Lee,

Of course it is difficult to compare set up’s and systems. To many varieties for sure.
One point never mentioned, what is your Internet speed? Could this be the bottleneck?
My Internet speed is not that high for The Netherlands but at least 350 mpbs down- and 35 mbps upload.
I also set my Rossini Apex in Buffer On.

1 Like

Hi,

Just an observation - not a specific comment on Jeffs reply per-se - but so frequently people get hung up on their internet SPEED.

Anything above 10meg is capable of streaming even uncompressed two channel 24/192 audio (theoretically) …

What is the real killer with internet connections is routing and latency issues and there is very little (well, nothing really) that you or your ISP can do about those as “the internet” is inherently designed to be dynamic and self-routing / self-healing on the whole which is great but can mean that sometimes it does get its knickers in a twist when you get things like DNS servers not updating correctly and when that does happen basically all you can do is wait for “the internet” to sort itself out.

7 Likes

Maybe it was already suggested, but AC isolation like a PS Audio regenerator, quality power cords, and quality power supplies (if they are used in your playback chain) are essential and can produce a big improvement in SQ. it should enable your ability to hear more deeply into the recording because of blacker, quieter backgrounds.

Another thing that really improved SQ for me was switching to Audirvana Studio using DLNA with an ethernet bridge. I also use Qobuz with Audirvana Studio.

1 Like

Got to say streaming from qobuz is very good into my vivaldi.
I have it running through a phoenixnet first, and its so good i use it over my ripped stuff stored in the melco.
Would love to see what a £30 K taiko extreme or the pink faun could do to inpove on it, if they are worth it.

2 Likes

Do you use the Vivaldi upsampler or Melco USB?

I’m struggling to imagine via what mechanisms a Taiko or similar might justify its price tag in sound quality terms, especially into a PhoenixNET which will do a damned good job of mitigating any noise on the line. Ethernet being ethernet, it can’t be a clocking thing.

Where can I find a simple and short exposition of the case for a Taiko or similar, without disappearing into the rabbithole?

Thanks

Vivaldi upsampler

1 Like

So am i, but reading about how great it is on the “what’s best forum” certainly is interesting, but it does make you think how genuine it all is

2 Likes

Me three
I know someone using a $25k Taiko with a $10k Chord Dave. On the Naim forum we refer to something like that as a Mullet system.

2 Likes

Personally, I take it with a very big pinch of salt. Nothing I’ve read on WBF w.r.t Taiko has any tangible technical value for Ethernet connected DACs. If anything, the Taiko maybe cleans up USB; in fact, nearly 70% of Taiko users are USB connected; IMHO, paying $20K to clean-up USB is the most absurd thing to do :rofl:

3 Likes

Cleaning up USB does strike me as a bit non-sensical. Surely it’s better not to use USB in the first place.

I do wonder if there’s an improvement to be had for some dCS DACs by having some functions done outside the box, though — as dCS themselves have done with the Vivaldi stack for example: the DAC having no Ethernet input.

Continuation of that line of thought: how might a Vivaldi or Rossini DAC be via certain transports (Oladra, Taiko etc.) instead of the Upsampler? What would be lost/gained in not having dCS’s own solution? Especially given that the original Upsampler is a decade old now. I imagine that there might be downsides in going AES/EBU into a Rossini instead of the onboard Ethernet input, but maybe not. Pondering continues while I eat lunch…

2 Likes

Agreed, before my previous dCS Bartok, I had a Chord Hugo 2 (along with their M-Scaler) and, once I added the Phoenix USB, it was clear that USB had been holding the show back (however the Phoenix USB did a stunning job).

On having some functions done outside of the dCS Rossini, if it were possible to power off the streamer inside the Rossini, it would be interesting to see how close this (with their upsampler) could get to their Vivaldi DAC.

My guess is they’ve invested more heavily in the clocking, along with the output stage and, possibly, the PSU’s of the Vivaldi, to separate the two.

As for one of their own DAC’s fed by a K50 via S/PDIF, my own impression (owning a K50v1) is that the Rossini with their own Rossini master clock beats the K50v1.

I have to say, the K50v1 surprised me. Without the master clock, the two would likely have been closer together. The K50v2 is, no doubt, even better and the OLADRA likely better still (I have no way of knowing which way the latter would go).

1 Like

Interesting stuff, Jeremy :+1:

Where does the master clock come into things, though? Perhaps I’m missing something — I’d still use my Vivaldi Clock in both cases (A: AES/EBU from Oladra, B: Ethernet from switch).

My understanding is, for AES:

The clock signal recovered from the AES (or S/PDIF) signal typically has priority (whether the dCS is controlled by an external clock or not).

I seem to recall reading you could override this, but if the DAC is driven by its own clock, there is a danger of buffer overruns/underruns due to the clocks not being precisely the same (one could be running faster than the other, meaning the DAC can run out of bits to process or runs out of buffer space to store them).

In a Vivaldi DAC + Upsampler + Clock combination, they are all locked to the same clock signals (44kHz and 48KHz), so the problem doesn’t arise.

The OLADRA has a word clock output, but not one that corresponds with the that used by the dCS boxes.

However, when using Ethernet:

The dCS controls the flow of data and can therefore rely on its own clock (influenced by any external master clock, if present)

2 Likes

Very source first :joy:

I don’t have a separate clock, i run the upsampler from the dac, so they both are tied to the vivaldi dac

My friend has an Oladra. Personally, on first listening at least, I preferred his system with the Innuos Statement; the Oladra had something in the mid-high frequencies but, to these ears, lost too much impact in the bass to be my personal choice. I am fortunate to be able to hear the system-with-Oladra reasonably frequently and, without that direct comparison, I have to say it sounds sweet indeed.

1 Like

I would love an OLADRA to replace my K50 :slight_smile:

I found it rather strange that it supposedly supplies a 500 x base PCM Word Clock rate. What was it designed to connect to?? :thinking:

1 Like