Music Servers

My "transport" is an Eversolo T8, with  an SSD.

I stream from Presto and Qobuz but have about classical 1,000 works stored on the T8.  They are a mix of purchases from Presto and Rips of CDs.  

My internet is somewhat slow and can be glitchy, and disappears with strong winds, ice or snowstorms, hence an alternative to streaming.

Also, when in NYC I go to Academy Records and pore through their media.

The T8 could also access a NAS, but the stuff I had on a NAS is now copied onto the 4GB ssd.

If your DAC has a first class clock, and a similarly good implementation of asynchronous USB, all sources will be essentially of the same quality as the data is buffered in the DAC and timing is solely the responsibility of the DAC's clock.

Other source to DAC connections may not have this isolation, in particular IIS depends critically on the source clock, which is transmitted to the DAC, and the IIS cable.

Any thoughts on how the quality of the server impacts the streaming sound?

@fire_water

that's interesting because I was told by a dealer at a reputable high-end store that in most cases music files bought and stored sounds better than streaming it.

I was told the same thing by a high-end store's owner.

As to your original question, I have something somewhat similar, which is a laptop with locally stored music files, feeding a Berkeley Audio Design Alpha USB Series 2 re--clocker (transport) via an Audioquest Diamond USB cable, feeding my DAC via a Shunyata Research Sigma-X AES/EBU cable, and it sounds great (but there are additional steps I am considering, for further improvement (more on that, below).

I started off with a Windows 10 laptop feeding my DAC locally stored music files, via a mass produced USB cable.  That was terrible.  My local high-end dealer loaned me an Audioquest Diamond USB cable, and it was seemingly magical.  Apparently, that Audiquest cable significantly reduced the noise, that my laptop was generating, from reaching my DAC.

A few months ago, I added the Berkeley re-clocker / transport, and that significantly reduced the jitter that I never noticed, and it also filters out noise.

The owner of the store from where I purchased the Berkely box told me that for best results, do the following:

1) Play locally stored music files.

2) Use Audirvāna as your music app.

3) Use an Apple Mac for the hardware.

I am already doing #1.

For #2, I have been using JRiver's Media Center, and I love it.  But I never tried any other competitor's offerings.

For #3, I have to find an answer to using a short USB cable, from the Mac to my Berkeley box.  I do not have that problem with my Windows laptop, because I use that laptop as a Remote Desktop server.  So it sits right next to my Berkeley box.  I have a client laptop that sits with me, that controls the other laptop.

I have no idea if Macs have that option.  I need to look into that.

Anyway, I highly recommend using that Berkeley Audio Design USB re-clocker.  According to reviews, all streamers benefit from it, including price-no-object streamers.  It accepts only a USB input, and outputs a choice of AES/EBU or TOS.  Berkeley recommends using the AES/EBU choice.

I own the Series 2 model.  They also have a Reference model, which is twice the price.  But a quality AES/EBU cable might cost more than the Berkeley box.

https://www.theabsolutesound.com/articles/berkeley-audio-design-alpha-usb-series-2-noise-isolation-device/

-- Robert Harley:

I know a longtime industry veteran who travels the world setting up and fine-tuning state-of-the-art systems. He carries with him his laptop loaded with music—and an Alpha USB. Whatever DAC is on hand in the system is greatly improved with the addition of the Alpha USB. He wouldn’t think of doing a setup without it.

More regarding the recording/file source. Apologies for thread hogging, it may be a little off topic, more music sub-forum based, but I feel the need to do this followup to my post above regarding HDDT vs. Qobuz streaming. It is related to streaming vs. local files at least. 

As I mentioned, I was elated with the sound quality of the downloaded album suggested above compared to streaming, and not so much with another recent release. I decided to recently try another download what was a highly regarded release from HDDT: Sonny Rollins Saxophone Collosus. It is a direct "pure" DSD transfer from 15 ips 2 track tape, 8 reviews all positive. 

Downloaded the DSD256 file, and via Roon it sounds great. Now to compare to the 24/96 Qobuz album I have stored in my library. (not purchased, but stored in my library- meaning when I listen to it it is streamed and not direct from the internal hard drive of the streamer) I have previously gone through many of the 14 different versions (!) of this album that Qobuz currently has available and I must have preferred the 24/96 file "restored" in 2024 by Mouton records, over the other high res 24/44.1 release Van Gelder remaster; all the other versions are 16/44.1.

And the Qobuz streaming sounds more detailed; It is quite a bit more vivid, maybe too much? A little hot? I haven’t decided whether this is more natural or not- It still could be considered more natural yet fatiguing over the long term? Doesn’t sound digital just louder. And yes, I matched volumes. 

I listened to another version: Original Jazz Sound 16/44.1 version, and it sounds different again, even brighter, still natural but I just moved up 10 rows in intensity. 

DSD should sound smoother over PCM, sure, but not to the point of missing 10% of the details in the music. The shocking part is my DAC does prefer DSD over PCM.

Anyway, longwinded post to state I don’t know anymore! it depends! But at this point based on what I am hearing I wouldn’t suggest purchasing the download as I don’t see a huge advantage over streaming this album, (yet). 

Of course I could purchase the Mouton or Van Gelder from Qobuz and convert it to DSD512.... 

This is exhausting.