Can't get my computer playing high res to sound better than my CD transport

[longinc]

I have been trying to get my laptop (macbook air with M1 chip) playing high res files (using Audirvana) to sound better than my CD transport, but I haven't been successful. I am hoping some members from this forum can shed some light.

 

I use the Merging NADAC (with the Merging external power supply). This DAC supports the audio over IP protocol Ravenna (ethernet), AES and SPDIF inputs. No USB input. I have an old Cary CD player (but it's built like a tank with a heavy power supply and has a well regarded Philips CD reader mechanism) that i use as CD transport piping into the NADAC via SPDIF RCA, and that sounds wonderful.

 

I connect my Macbook Air -> Belkin USB/C-to-ethernet converter->unshielded Cat 6 ethernet cable -> NADAC. In "theory", I was expecting that since it's using the IP protocol, you get the benefits of error correction (so I assume I am getting the correct and complete set of bits), AND the NADAC has the master clock (it also buffers and re-clocks the bits) meaning it should be jitter free/has low jitter. The NADAC has a good clock. I am playing high res files local to the Mac (so I am not dealing with networking issues associated with streaming real-time off an online service). But, bass sounds muddier, instruments/voices sound less right/fulfilling to my ears (almost like there are time alignment issues and you are not hearing the fullness of each tone), and the images are less defined. For comparison, I have a couple of albums where I have the CD's for and also the high res file equivalents on the Mac. I am puzzled that my laptop is sounding "worse" than my CD transport even though I am playing high res files from my laptop.

 

What other gremlins am i dealing with here? 

 

(btw, i use unshielded (U/UTP) ethernet cable because I was told shielded ethernet cables can cause more issues (than help) because it "connects" the ground of the 2 connected devices and you lose the isolation that one is supposed to get from ethernet connections). 

 

Anyone have any thoughts on my situation?  Thank you for reading!

[One and a half]

OK, Cary to DAC is good, Mac to DAC is crap.

Unshielded cables are ok, no issue with them.

 

The source is full of noise in the Mac, suggest:

1. Add in a DX filter, they are cheap and remove bad noise that affects digital signals on the network

2. Replace the source with a Melco NAS, Antipodes or similar

 

Just because files on the MAC are either hi-res or well recorded Redbook, the noise from the computer attaches itself to the Ethernet, or USB for that matter. This noise is difficult to remove and is responsible for what you hear, mud and loss of definition compared with the Cary CD. The choices are to improve the source, as to what and how much, well that's what this forum is all about.

 

[MarcelNL]

I beg to differ, and would not start adding stuff rightaway. Taming the abundance of activity in the Mac probably works too and should be for free, although even Red Book files will sound better than a CD player there is some work to do before you get there.

 

Shielding and cabling is only a minor part, settings and software is where you start until you get a taste of things and start building/buying a high end dedicated audio server. As self confessing Mac abstainer I have no links to audio settings and tweaks, I'm sure there are others well versed in that OS around here.

 

Good luck and have fun, save settings in a backup file if you can even change anything 'under the hood' in a Mac without jailbreaking. 

[Slainte]

Who says high res should sound better?

[MarcelNL]

nobody, I dare say that even a red book file can sound better from a computer than out of a CD player. It totally depends HOW the high res file was made and processed if it can sound better than a red book CD. 

[davide256]

12 hours ago, longinc said:

I have been trying to get my laptop (macbook air with M1 chip) playing high res files (using Audirvana) to sound better than my CD transport, but I haven't been successful. I am hoping some members from this forum can shed some light.

 

I use the Merging NADAC (with the Merging external power supply). This DAC supports the audio over IP protocol Ravenna (ethernet), AES and SPDIF inputs. No USB input. I have an old Cary CD player (but it's built like a tank with a heavy power supply and has a well regarded Philips CD reader mechanism) that i use as CD transport piping into the NADAC via SPDIF RCA, and that sounds wonderful.

 

I connect my Macbook Air -> Belkin USB/C-to-ethernet converter->unshielded Cat 6 ethernet cable -> NADAC. In "theory", I was expecting that since it's using the IP protocol, you get the benefits of error correction (so I assume I am getting the correct and complete set of bits), AND the NADAC has the master clock (it also buffers and re-clocks the bits) meaning it should be jitter free/has low jitter. The NADAC has a good clock. I am playing high res files local to the Mac (so I am not dealing with networking issues associated with streaming real-time off an online service). But, bass sounds muddier, instruments/voices sound less right/fulfilling to my ears (almost like there are time alignment issues and you are not hearing the fullness of each tone), and the images are less defined. For comparison, I have a couple of albums where I have the CD's for and also the high res file equivalents on the Mac. I am puzzled that my laptop is sounding "worse" than my CD transport even though I am playing high res files from my laptop.

 

What other gremlins am i dealing with here? 

 

(btw, i use unshielded (U/UTP) ethernet cable because I was told shielded ethernet cables can cause more issues (than help) because it "connects" the ground of the 2 connected devices and you lose the isolation that one is supposed to get from ethernet connections). 

 

Anyone have any thoughts on my situation?  Thank you for reading!

Figuring out the audio interference bugaboo's for a 1 box USB out media server is a PITA, I've given up on it.

Good results are much more predictable with streaming to a low power USB out endpoint such as a microRendu or Atom CPU endpoint with a good LPS

[MarcelNL]

8 minutes ago, davide256 said:

Figuring out the audio interference bugaboo's for a 1 box USB out media server is a PITA, I've given up on it.

Good results are much more predictable with streaming to a low power USB out endpoint such as a microRendu or Atom CPU endpoint with a good LPS

I beg to differ, the results from using  NUC on an LPS were decent and surely better than using a laptop or similar...the results with my server project are at a completely different level.

The beauty is, there is plenty of info in this very subforum to get you on the right track real fast.

 

[davide256]

15 minutes ago, MarcelNL said:

I beg to differ, the results from using  NUC on an LPS were decent and surely better than using a laptop or similar...the results with my server project are at a completely different level.

The beauty is, there is plenty of info in this very subforum to get you on the right track real fast.

 

Sorry, BTDT.  When good results require gambling more than $2k on computer components for uncertain results, I start thinking off the shelf solution

that can be auditioned.

[MarcelNL]

26 minutes ago, davide256 said:

Sorry, BTDT.  When good results require gambling more than $2k on computer components for uncertain results, I start thinking off the shelf solution

that can be auditioned.

Totally understandable! And the total investment exceeds 2K considerably.....think 8-10K for a high end audio server, but that gets you a solution that can compete with 20K off the shelf solutions.

Audio has interesting markups, manufacturer, importer and/or distributor they all need to heat their place and eat....just figure what value the kit in such a 2K solution is worth, likely around where I live off that 2K just the VAT is 400, dealer margin used to be around 40% leaving 1100 or so to buy whatever is in that product, software and putting it together and warranty. 

[longinc]

Thanks Everyone for your perspective. Very helpful.

 

@One and a half thanks for sharing the DX filter (looks like it's similar to the Network Acoustics passive filters). Trying to better understand this "noise" that attaches to the ethernet signal. Since I am only using an ethernet connection (and no USB or any other connection downstream), and I also get the error correction of the IP protocol, and the NADAC has the master clock (and does buffering and reclocking), what you are saying is that even though I am getting the "correct bits" in the "correct timing", this other "noise" (i assumes its high frequency RFI/EMI coming through the ethernet connection) is messing with the digital to analog conversion process despite it is getting correct and accurately-timed bits - did I interpret that correctly?

[AudioDoctor]

I would use a headless server running Ubuntu Linux server, and send the music to your DAC via an ethernet connection to an endpoint that connects the network to your DAC.

 

A multipurpose computer, with a display, and a full desktop OS, will never sound as good as a headless and dedicated server running only the necessary services to play music.

[longinc]

Thanks @AudioDoctor. forgot to mention (a constraint I have) that because my DAC uses Ravennea (AOIP), my player end point or music server/player combo machine needs to be based on MacOS or Windows because of the Ravenna drivers. They have drivers for linux too but those (I have heard through the grapevine) are not implemented as well. 

[AudioDoctor]

2 hours ago, longinc said:

Thanks @AudioDoctor. forgot to mention (a constraint I have) that because my DAC uses Ravennea (AOIP), my player end point or music server/player combo machine needs to be based on MacOS or Windows because of the Ravenna drivers. They have drivers for linux too but those (I have heard through the grapevine) are not implemented as well. 

 

Ok, good to know. After a look at the website I think the easiest route for you would be to get a small NAS and put your music on that, use the computer as a control point, or an iPad/iPhone. 

 

I think there are 2 things working against you. 1 being the computer itself and 2 is the USB to Ethernet adapter.

[longinc]

Thanks @AudioDoctor. Sorry still getting used to how all of this works. My DAC does not have player on it. It is purely a DAC with an ethernet input, so i assume i will still need the player software installed on a machine "playing into" the DAC. Or are you suggesting there is a way to install the player software on the NAS? 

[mitch751]

I am using three Mac computers for playing my music library and streaming connected to Chord Hugo TT2 via different endpoints :

 

1. Mac mini 2012 (Uptone JS2 LPS)

2. Mac mini 2018 (Paul Hynes SR5T LPS)

3. Mac Studio Ultra

 

I am very happy with all these Mac, and all hardly heard any difference as compared to my Goldmund Eidos (2004 model) as a transport to Chord Hugo TT2, however, dCS Network Bridge with Mosaic is slightly better (10%?), 

[mitch751]

8 minutes ago, longinc said:

Thanks @AudioDoctor. Sorry still getting used to how all of this works. My DAC does not have player on it. It is purely a DAC with an ethernet input, so i assume i will still need the player software installed on a machine "playing into" the DAC. Or are you suggesting there is a way to install the player software on the NAS? 

Adding an endpoint e.g. UltraRendu between computer and a DAC via lan cable

[longinc]

Thanks @mitch751 for sharing the diagram and your current setup. The UltraRendu won't work unless I am able to install the Ravenna drivers on the UltraRendu for it to "see" the merging NADAC. Drivers are only available for MacOs and Windows. The linux driver (i hear) is marginal. And my DAC does not have USB. It does have AES and SPDIF RCA, but I was hoping to see if I can leverage the ethernet input so i can take advantage of higher res files...

[AudioDoctor]

1 hour ago, longinc said:

Thanks @AudioDoctor. Sorry still getting used to how all of this works. My DAC does not have player on it. It is purely a DAC with an ethernet input, so i assume i will still need the player software installed on a machine "playing into" the DAC. Or are you suggesting there is a way to install the player software on the NAS? 

 

Ok, hear me out here. HQPlayer has a Ravenna version of the Embedded player that runs on Linux. Maybe @Miska can explain if this would all work flawlessly, if it will, that would be a great way to get the best sound quality. I use and love HQPlayer and I think it is the best sound quality I have had in my system.

 

 

[longinc]

11 hours ago, AudioDoctor said:

 

Ok, hear me out here. HQPlayer has a Ravenna version of the Embedded player that runs on Linux. Maybe @Miska can explain if this would all work flawlessly, if it will, that would be a great way to get the best sound quality. I use and love HQPlayer and I think it is the best sound quality I have had in my system.

 

 

 

Thanks @AudioDoctorthat would be great (more options). Let's see if @Miska has any further thoughts on this possibility....

[One and a half]

On 5/26/2022 at 6:47 AM, longinc said:

Thanks Everyone for your perspective. Very helpful.

 

@One and a half thanks for sharing the DX filter (looks like it's similar to the Network Acoustics passive filters). Trying to better understand this "noise" that attaches to the ethernet signal. Since I am only using an ethernet connection (and no USB or any other connection downstream), and I also get the error correction of the IP protocol, and the NADAC has the master clock (and does buffering and reclocking), what you are saying is that even though I am getting the "correct bits" in the "correct timing", this other "noise" (i assumes its high frequency RFI/EMI coming through the ethernet connection) is messing with the digital to analog conversion process despite it is getting correct and accurately-timed bits - did I interpret that correctly?

Yes, that's about right. All the error correction and 'bit perfectness' is no match for parasitic gremlins that take a ride on the Ethernet signal. Still, there are products that create a barrier across Ethernet, the most popular and effective is the EtherRegen (if you can find one). I think the speed where Ravenna operates is OK for the 100Mb/s EtherRegen, however routable Audio IP used to be restricted to certain switches, but things may have change in that respect. 

Elimination or an attempt to reduce parasitic noise is one major goal that's worth pursuing, since it can clean up the SQ quite markedly.

[longinc]

@One and a halfI finally got the DX filters. Wow! night and day change for the better. Players and instruments were better defined, darker background, and things just sounded fuller/more natural. This is the cheapest tweak that made such an improvement. Still shy of the sound quality I am getting from my CD transport (but close), and more importantly, now I know how much "noise" my source is giving off and clearly tells me where I need to improve (spend $'s on) to make further progress to sound quality. Thank you! 

[One and a half]

1 hour ago, longinc said:

@One and a halfI finally got the DX filters. Wow! night and day change for the better. Players and instruments were better defined, darker background, and things just sounded fuller/more natural. This is the cheapest tweak that made such an improvement. Still shy of the sound quality I am getting from my CD transport (but close), and more importantly, now I know how much "noise" my source is giving off and clearly tells me where I need to improve (spend $'s on) to make further progress to sound quality. Thank you! 

Great stuff! When you find a reduced noise source, thay you find equal to or better than the Cary, please let us know. Media servers can be really expensive diy or ready made, it's not an easy path, even small details like wiring inside on ground paths makes a difference. 

 

At a high level, what does work is having music files on one box with server software,  another box, even the laptop or ipad should be ok in a function of control, where you pick the music, and the end point as you have now connected.

 

DX filter from the box with files to the end point cleans up the rest. The laptop is then out of the picture as far as noise is concerned since it asks the server to send to the end point.