Streaming through upconverting DAC - can I even improve audio quality?


New to streaming but have been at this hobby for over 50 years.  Started by getting a Marantz SA-10 SACD player that enables streaming through its DAC only via SPIDF.  The SA-10 upconverts everything to DSD 11.2.  So I started with ifi Zen Stream via ethernet, good Belden 75 ohm cable, then upgraded streamer to ifi's linear power supply, which made an improvement I could hear through balance of system, Luxman 590axii and Yamaha NS-5000 loudspeakers.  Sounds good so I'v e started to look at other streamers thinking it could get better, including the new Eversolo DMP-A8 (where I could use its DAC or the Marantz) and even the $5200 HIfi Rose RS-130 (streamer only).  But here's the thing:  so far, just using free Spotify 44.1 redbook quality, 1) since the Marantz upconverts everything to DSD 11.2 would I gain anything from a premium streaming subscription? and 2) the Ifi's femto clock and it's purifier noise removal works on the signals into the unit (clock) and then out (purifier on coax) and then the Marantz' dual clocks take over, so is there anything to be gained with a different streamer's clock? Or have I gotten the quality to about as good as its going to get without investing significantly more in another higher end DAC and streamer?  Should I be able to get SACD quality via streaming or maybe I'm getting that now and it's just that streaming sounds different than a SACD.

Ag insider logo xs@2xeisen0169

The 44.1 redbook quality I am referring to is the sampling frequency of the linear PCM signal input to the Marantz DAC.  That's what the Marantz manual says.  What I'm hearing sounds as good as CD but of course the Marantz is upsampling.  Maybe the 44.1 is a default number since that's the lowest frequency supported by the Marantz and what's happening is the Marantz is upsampling and converting Spotify's 160kb signal.  I've never been able to find any really good review on the Marantz DAC when streaming through its coax input.

@eisen0169 

I’m not really big on the upsampling thought process as I believe that if the gear can’t reproduce 16/24/32 bit audio in the first place there’s no point in worrying higher sampling rates.

So much of this depends upon the design and build quality of each component as well. From what I’ve read in your first post I would start with comparisons between your SACD players output and how that compares to streaming content. If you’re playing CD’s this will be more of an apple to apples comparison. Note the differences of what you like or not and start your focus there.

I use a Naim Uniti Core to store/playback my CD’s. I have a lower model Lumin U2 Mini transport for streaming. Although I listen to the music through the Lumin almost everyday I still prefer the sound of the ripped CD from the Naim- regardless of what format or sample rate the Lumin is streaming the music in. I run the Lumin native- even though it will upsample to 512 or something crazy it’s always playing the file in its native format.

My point with this to start simple and work with what you have from a basic level rather than being lured into what seems to be greater features. You may find in the end that they really don’t perform with that great of a difference or that the gear you’re using isn’t really that good at performing them anyway. If you start with a basic comparison- CD player 16/44 Redbook through your chain vs streaming 16/44 through your chain you’re able to analyze the difference. Use that a while and get comfortable before moving up to comparing CD against upsampled streaming content. You may find that those features aren’t really that important to you and if they do turn out to be at least you’ll have invested the time into discovering the nuances between both. Have fun and enjoy!

Spotify uses a lossy compression algorithm (Advanced Audio Coding or AAC - link) to deliver audio while saving on bandwidth and storage space. This means that some of the audio information is discarded during the compression process. These are parts of the audible spectrum that a psychoacoustic model has deemed inaudible. The compressed audio is then decompressed into a format and sampling rate suitable for your DAC, in this case 44.1 PCM. No matter how much upsampling your equipment performs on the resultant audio data, the discarded information is lost forever.

 

Also, don’t be surprised if it’s difficult to distinguish between audio compressed with AAC and compact disc audio in a blind listening test. After all, these types of tests are used in the development of compression algorithms.

 

Finally re: SACD, you can think of the resolution as similar to that of 20 - 24 bit PCM audio depending on the demodulation approach so any lossless hi-res music service out there will get you good performance as long as the recording itself was mixed and mastered well.

Upsampliing is like putting Bud Light into Waterford crystal.

If you don't start with something better, that's all that you're gonna get. Target practice.

Upsampling does not add more data to a source. Nada! Therefore useless! It's not like analog where a more advanced stylus profile will follow those tiny groove undulations to extract more musical detail.