Differences in SQ: Some people hear them, I do not... Why?

[steklo]

I recently did a comparison of several MacOSX audioplayer-softwares: Amarra 3.0.2, Audirvana 1.5.10, Fidelia 1.5.3 PureMusic 2.0.2 and JRiver MediaCenter 190154

 

I was astonished to find that SQ of all of them was so good and so close, I could live with any of them easily. That was quite a difference to my first testing about 2 years ago, when I found Amarra to be clearly the best.

More refined listening finally displayed, that I liked Amarra and Audirvana the best for sonics. For me Amarra seems to be a bit richer sounding while Audirvana seems to be a little more neutral and sometimes I felt Audirvana to have a bit better defined soundstage. But these differences are just so little, I could hardly tell them from each other.

I could not reproduce the large difference in SQ that others reported between stand-alone playback and iTunes mode, neither for Amarra nor with Audirvana...

 

 

In contrast to that upgrading my MacMini (2009 version with external PSU and white cover) from SnowLeopard to Mavericks made a really remarkable difference. I guess this is mainly due to the fact that Mavericks connects to my AR-T Legato USB/SPDIF-converter in integer mode whereas SnowLeopard (for what reason ever) did not.

 

 

Trying a laboratory linear PSU (vintage Systron Donner up to 20V/10A, plain electrics, heavy built, very low ripple, very good voltage stability over a large range of current loads) was not a remarkable success vs. the standard Apple SMPS. Not enough to make me switch for it.

Also I could not find a remarkable difference with the „Computer Audio Design OSX Audio Optimization Script“ applied (I cloned my system to a second partition so I could do an A/B comparison). I should mention I did some manual optimizations to my standard systems in advance such as switching off Spotlight, TimeMachine etc.

 

 

I’m a bit confused about my findings because I think that both my hearing as well as my hifi-system are not so bad to not reveal existing differences.

Is it possible that by installing the AR-T Legato, which has quite a good PSU inside and does not draw any power from USB, which is galvanically isolated from the DAC by a transformer in its output stage and which does have a very precise clock inside, a lot of issues on the computer-side are getting „healed“ so that in the end there are not so many things that really matter for SQ?

 

My system:

MacMini 2009 with Amarra 3.0.2 > AR-T Legato USB/SPDIF converter > Audio Note DAC 1.1x signature > Audio Note M1 Phono > Audio Note P1SE > DIY speakers based upon vintage Saba "Greencones" plus BMS4524 Supertweeter

[tranz]

Hi steklo,

 

It is not strange. There are so many variables that can be the reason.

 

In your setup it would not surprise me if the way the DAC handles USB 5v and ground is a major positive factor. You could try one additional thing: tape the 5v and ground pins on the USB plug you put into the computer.

 

On my old speakers I did not notice tweaks as much, and now I am used to listening for certain subtle changes. But my SPDIF converter does use the 5v and it is a pain to clean-up.

[coot]

Why not A+ 2.02?

[ericuco]

In your setup it would not surprise me if the way the DAC handles USB 5v and ground is a major positive factor. You could try one additional thing: tape the 5v and ground pins on the USB plug you put into the computer.

 

Agree with removing the 5v from your USB cable as being worthwhile. I taped the 5v pin but not the ground on the advice from a guy in the know. The USB XMOS board in my DAC is powered separately on both the non-galvanic & galvanic sides.

 

On the player side, you should give HQ Player a try, there is a 30-day demo license available (only plays for 30 miutes at a time then shuts down which is a PITA but gives you a chance to see what it can do). I had been a long time A+ user and very happy with it on a number of fronts (SQ, development, support, etc.) but IMHO HQ Player is in another league. Just made the transition yesterday and still feeling my way through all the options (lots of options) but like what I am hearing so far using PCM conversion to DSD. There is a recently released version for OSX. The other thing that intrigues me about HQ Player is the NAA set up using the SOtM sMS-100 sometime down the road.

[mrvco]

There are so many factors that can affect "computer audio" that I find it helpful to have a baseline to refer back to such as a turntable or CD player that isn't (potentially) subject to so many variables.

[mayhem13]

More likely?.........you're just not as receptive to suggestive bias as you've had more experience with different gear over a comparatively small period of time. Your findings are parallel with 95% of the music listening world with only the very small and insignificant sect of audiophilia claiming stark and wide ranging improvements.

[yanrijiah]

well,The USB XMOS board in my DAC is powered separately on both the non-galvanic & galvanic sides.thanks

[tranz]

Ok, how about a few pictures to illustrate.

 

A few amateur scope pics of the USB 5v:

 

Using IFI iUSB with probe on 5v USB wire and cheap LPSU. What I feed my Berkeley SPDIF converter.

 

 

Direct 5v USB connection from battery operated laptop (power lead unplugged) without iUSB. Using switching PSU likely increases the noise even more and this is what would be fed into the sensitive DAC.

 

 

All using AC coupling, 2ms/div, 5mV/div, 1x probe. LPSU was plugged into a power filter.

[steklo]

Well, as for USB-power: AR-T has addressed this in advance by putting a dedicated USB-cable without power wire to the Legato, thus I'm already benefitting from this suggestion, nevertheless thanks.

 

HQ Player definitely is on my shortlist. As far as I understand its advantages over other players is mostly about special filtering methods and upsampling / conversion to DSD. To me it's questionable if I can take advantage from HQ Player with my particular system, because it is some kind of specialized for Redbook: AR-T Legato is 44.1/16 only and Audio Note DAC 1.1x does in principle accept 96/24 but actually truncates data beyond 18bit. Still both units to me sounded much better than other highres-capable gear that I compared it to so that until now I didn't care much about highres-stuff.

But still it will be nice to hear what HQ player can do by passing Redbook files straight and unprocessed into my Legato. Furthermore of course it would be quite interesting to hear how a modern DSD-DAC plus HQ Player would perform vs. my old-school unfiltered-nonos-Redbook-stuff.