Author Topic: My next experiment re digital audio.  (Read 799 times)

Offline Rotten Johnny

  • ROONed for life.
  • Administrator
  • Trade Count: (+31)
  • *****
  • Posts: 14,483
My next experiment re digital audio.
« on: January 22, 2017, 06:32:56 PM »
As a self-confessed sceptic where audiophoolery is concerned it's long irked me when people make claims of day/night differences, veils being lifted etc. achieved by switching out a wall wart for a linear PSU, changing USB transports or doing both.

Cases in point:

- I've seen many discussions where switching out the wall wart powering a USB transport purportedly markedly improves sound quality
- similarly, where a DAC's USB port is powered by the incoming USB cable it's been reported that a linear PSU independently powering the USB port again significantly improves proceedings.
- some guys slapped together a custom SOC board, called in the marketing guys claimed it's 'USB optimized for audio', borrowed open source software efforts without acknowledgement,  had it blessed by Goodness and sell it for $700. For perspective the ODROID sells for $39.

I should soon be armed with a purpose built linear PSU that measures extremely well (and better than any I've seen selling for the price of a decent DAC)  and I'll be running a number of tests with it, including:
- powering my ODROID-C2 which serves as a USB transport
- powering the DAC's USB port.

For first run I'll use the Joty and Sennheiser HD800 to see whether there are any obvious audible differences I can detect.  Thereafter I'll try the same on my main setup.

If I can get access to the right equipment I may even compare DAC output with and without the LPSU. Should be interesting...

Later on, if I can get my hands on a blessed audiophile USB I'd like to measure what comes out of the business end of a DAC when using the ODROID and compare that with what is produced when the USB has Goodness' stamp of approval.

Following that I'm hoping someone is brave enough to accompany their audiophile music sever to my place, again to measure what comes out of the business end of the DAC compared with what's produced using my stock mini-itx.





« Last Edit: January 22, 2017, 06:52:45 PM by Rotten Johnny »
Audiophile: There is almost no other group that prides themselves more on wasting good money on utterly worthless ****, and then trying to furiously blow smoke up their own ass to justify it.

Free your mind...and your ass will follow.

Offline scrarfussi

  • Trade Count: (+21)
  • AVForums Veteran
  • *****
  • Posts: 5,828
  • Quantum In Me Fuit
Re: My next experiment re digital audio.
« Reply #1 on: January 22, 2017, 06:42:36 PM »
this will be interesting  :coffee:
suum cuique pulchrum est


"I refuse to prove that I exist,'" says God, "for proof denies faith, and without faith I am nothing."

Online CAD

  • Trade Count: (+53)
  • AVForums Veteran
  • *****
  • Posts: 6,053
  • Keep Calm,Be Quiet and Listen to the Music
Re: My next experiment re digital audio.
« Reply #2 on: January 22, 2017, 06:47:14 PM »
Looking forward to your objective reviews as always RJ  ;)
The way of the warrior is NO match for the way of the wife.
Light travels faster than sound. This is why some people appear bright until you hear them speak.

Offline scrarfussi

  • Trade Count: (+21)
  • AVForums Veteran
  • *****
  • Posts: 5,828
  • Quantum In Me Fuit
Re: My next experiment re digital audio.
« Reply #3 on: January 22, 2017, 06:50:58 PM »
Looking forward to your objective reviews as always RJ  ;)

You mean Biased reviews :whistler:
suum cuique pulchrum est


"I refuse to prove that I exist,'" says God, "for proof denies faith, and without faith I am nothing."

Offline Rotten Johnny

  • ROONed for life.
  • Administrator
  • Trade Count: (+31)
  • *****
  • Posts: 14,483
Re: My next experiment re digital audio.
« Reply #4 on: January 22, 2017, 06:55:19 PM »
^^^ the measurements won't lie....you can come sit beside me to ensure I don't gippo readings, and I'd be delighted if there were in fact improvements...it could make the spend on the LPSU worthwhile.
Audiophile: There is almost no other group that prides themselves more on wasting good money on utterly worthless ****, and then trying to furiously blow smoke up their own ass to justify it.

Free your mind...and your ass will follow.

Offline kamikazi

  • Trade Count: (+17)
  • AVForums Veteran
  • *****
  • Posts: 5,922
Re: My next experiment re digital audio.
« Reply #5 on: January 22, 2017, 07:40:21 PM »
The NuPrime uDSD might be nice to try with or the Audioquest Dragonfly, but you'd have to put in some specific use cases that you would like to test. USB decrapifiers and linear PSUs have different effects depending on what you are dealing with. Often I've found the most benefit with things that are really crappy out of the box. While the RL Concero sounds a bit smoother with the external PSU that cuts the 5V and GND lines on the USB interface, the most benefit I've found actually was with my B&W MM-1 multimedia speakers (uses a USB transceiver and DAC integrated into the same chip and is partially powered off the USB supply). Noticeably less hiss with my in ears monitors using the headphone amp which should be measurable and would be a nice test case. When using the external PSU with the Concero or AP2 as digital USB transports I doubt I'd be able to identify the differences through blind testing and when not it would be impossible to rule out expectation bias so I don't really consider it that much of a benefit in that use case.

Again, many DACs don't use the 5V line from a computer at all. Many have complex clocking schemes that negate the benefit of external clocking/reclocking, so you've got to be sure that you aren't drawing conclusions from tests where the improvement might never be evident at all.
« Last Edit: January 22, 2017, 07:50:35 PM by kamikazi »

Offline gavinbirss

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 1,104
Re: My next experiment re digital audio.
« Reply #6 on: January 22, 2017, 09:16:05 PM »
There is a number of "folks" that reckon a USB 3.0 port is better for USB 2.0 audio. Better still the port must be powered by low noise regulators.

A good quality short USB cable is also a must.

My isolated XMOS chipset to Soekris are plugged to my PC Engines APU2C4 which has USB 3.0 ports. The Soekris inputs are also isolated nothing to worry about there.

PC Engines are running a new version of the APU3C board that has a 1xPCI express slot on the side.

My first thoughts was "Pink Faun i2s Bridge".
This eliminates USB completely...

http://pinkfaun.com/index.php/producten/streaming-audio/pink-faun-i2s-bridge-st-stereo-uitvoering-d-a-converter.html

Knowning how USB is implemented on the particular DAC helps, is it isolated, what chipset and so on.
« Last Edit: January 22, 2017, 09:25:27 PM by gavinbirss »

Offline gavinbirss

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 1,104
Re: My next experiment re digital audio.
« Reply #7 on: January 22, 2017, 09:40:28 PM »
A realtime Linux kernel is the other critical part of the chain when using USB.

Offline pwatts

  • Trade Count: (+3)
  • AVForums Veteran
  • *****
  • Posts: 1,095
Re: My next experiment re digital audio.
« Reply #8 on: January 22, 2017, 11:41:44 PM »
There is a number of "folks" that reckon a USB 3.0 port is better for USB 2.0 audio. Better still the port must be powered by low noise regulators.

This has to be clarified. USB3 has zero in common with USB2, and if one looks at the connector pinouts it's clear that it's all duplicated. USB3 has its own data pins and actually a USB3 connector houses a completely independent set of USB2 pins. All that is shared is ground, power and shell (earth).

USB3 needs to be able to deliver more current, but it all comes from the same PC 5V bus, which then passes through a current-limited load switch to turn off the current when exceeding the spec. As long as the current limit is not exceeded there's no difference and the "purity" of the current is the same regardless since it's just fed from the PC PSU.

There is a far more plausible reason for this though: because of this higher current ability, a USB3 host port typically has higher capacitance on the power rail of better quality than USB2. The spec says 120uF minimum, and for most USB2 applications a cheap and nasty electrolytic is used. For USB3, designers tend to opt for something a little larger and with better ESR specs to pass qualification.


Offline gavinbirss

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 1,104
Re: My next experiment re digital audio.
« Reply #9 on: January 23, 2017, 07:52:17 AM »
This has to be clarified. USB3 has zero in common with USB2, and if one looks at the connector pinouts it's clear that it's all duplicated. USB3 has its own data pins and actually a USB3 connector houses a completely independent set of USB2 pins. All that is shared is ground, power and shell (earth).

USB3 needs to be able to deliver more current, but it all comes from the same PC 5V bus, which then passes through a current-limited load switch to turn off the current when exceeding the spec. As long as the current limit is not exceeded there's no difference and the "purity" of the current is the same regardless since it's just fed from the PC PSU.

There is a far more plausible reason for this though: because of this higher current ability, a USB3 host port typically has higher capacitance on the power rail of better quality than USB2. The spec says 120uF minimum, and for most USB2 applications a cheap and nasty electrolytic is used. For USB3, designers tend to opt for something a little larger and with better ESR specs to pass qualification.

Perhaps there is also a little marketing hype?

Does USB 3 also need 90ohms? of series resistance or is it less?

Offline pwatts

  • Trade Count: (+3)
  • AVForums Veteran
  • *****
  • Posts: 1,095
Re: My next experiment re digital audio.
« Reply #10 on: January 23, 2017, 08:06:41 AM »
Can't comment on hype, but again, USB2 does not share the same data lines as USB3. A USB3 connector is a 2-in1, and if you use a USB2 link partner the USB3 lines are unconnected. The USB2 lines are implemented the same way as the other regular USB2 ports on the host. USB2 is not USB3 being down-throttled, it is a totally different bus that's crammed in with USB2 in a connector sharing only mechanics, power and ground in a clever way to ensure adoption.
An actual USB3 end node would be interesting though but that's a whole different kettle of fish.

Offline Rotten Johnny

  • ROONed for life.
  • Administrator
  • Trade Count: (+31)
  • *****
  • Posts: 14,483
Re: My next experiment re digital audio.
« Reply #11 on: January 23, 2017, 08:10:41 AM »
The NuPrime uDSD might be nice to try with or the Audioquest Dragonfly, but you'd have to put in some specific use cases that you would like to test. USB decrapifiers and linear PSUs have different effects depending on what you are dealing with. Often I've found the most benefit with things that are really crappy out of the box. While the RL Concero sounds a bit smoother with the external PSU that cuts the 5V and GND lines on the USB interface, the most benefit I've found actually was with my B&W MM-1 multimedia speakers (uses a USB transceiver and DAC integrated into the same chip and is partially powered off the USB supply). Noticeably less hiss with my in ears monitors using the headphone amp which should be measurable and would be a nice test case. When using the external PSU with the Concero or AP2 as digital USB transports I doubt I'd be able to identify the differences through blind testing and when not it would be impossible to rule out expectation bias so I don't really consider it that much of a benefit in that use case.

Again, many DACs don't use the 5V line from a computer at all. Many have complex clocking schemes that negate the benefit of external clocking/reclocking, so you've got to be sure that you aren't drawing conclusions from tests where the improvement might never be evident at all.
Agree, however, I'm most interested in so called high-end DACs where users with $$$$$ hi-fi claim to hear day/night differences, so the uDSD and Dragonfly are low on the list of priorities.  With Roon it's pretty easy to tell which DAC's take power from the incoming USB ... if they do the USB chip reports their capabilities whilst the DAC is switched off/unplugged.  I'll PM you ahead of any testing.

Audiophile: There is almost no other group that prides themselves more on wasting good money on utterly worthless ****, and then trying to furiously blow smoke up their own ass to justify it.

Free your mind...and your ass will follow.

Offline gavinbirss

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 1,104
Re: My next experiment re digital audio.
« Reply #12 on: January 24, 2017, 02:21:33 PM »
... With Roon it's pretty easy to tell which DAC's take power from the incoming USB ... if they do the USB chip reports their capabilities whilst the DAC is switched off/unplugged...

Am leaning to not agree with above?
If the USB chipset is isolated with digital isolators, the USB chipset will need its own power to be on/powered. It this case if the DAC is off (isolated side of USB chipset also) applying your logic you assume it is using incoming USB power because it will only report capabilities if DAC is powered on.

Think you are rather determing here if the USB chipset is isolated. If not it will be powered by incoming USB power and be on, even when DAC is switched off.

The max current draw/request from the USB endpoint may also be extra confirmation as enumerated by the USB chipset.

Quote
lsusb -v|egrep "^Bus|MaxPower"

The USB chipset I2S connection will be unconfirmed, until the DAC chip/board is powered.
If using a specific module (linux) it will fail until the DAC chip/board is powered.