Were the levels matched in that blind A/B test? If not, all bets are off. Why? That's because you can hear a difference from a same DAC if played at different levels. You don't have to change DAC to hear a difference, think about that.
Absolutely! If you listen to the louder source you think that it is clearer and better.
There is no need to change/match the levels since I have two exactly the same USB cables, one of these with separated leads for power and data. The levels are always kept the same.
My next guest for the A/B test was daughter who was told not only to listen but to tell the difference. She said that the words in the songs are more distinct - and every time it was with battery (powerbank).
I have received the response from one of the respectful Dutch forthcoming companies producing high level LPSU.
The question of course, was if the USB separate PSU could improve SQ, before entering the galvanic isolation in USB/SPDIF interface.
His answer:
Quote
Yes, it definitely will improve the sound quality, since the Signxer XMos USB receiver works on the voltage provided by the NUC board through the USB cable. The digital outputs of this receiver go through the galvanic isolation, but if is already noisy it will never get better. Unqoute
Regrettably, Allo Support, did not care to respond on any of my questions and remain quite unresponsive to this topic on their own thread except when Catana is involved.
To say bluntly, I would think twice if I am going to buy...
There is no need to change/match the levels since I have two exactly the same USB cables, one of these with separated leads for power and data. The levels are always kept the same.
My next guest for the A/B test was daughter who was told not only to listen but to tell the difference. She said that the words in the songs are more distinct - and every time it was with battery (powerbank).
I have received the response from one of the respectful Dutch forthcoming companies producing high level LPSU.
The question of course, was if the USB separate PSU could improve SQ, before entering the galvanic isolation in USB/SPDIF interface.
His answer:
Quote
Yes, it definitely will improve the sound quality, since the Signxer XMos USB receiver works on the voltage provided by the NUC board through the USB cable. The digital outputs of this receiver go through the galvanic isolation, but if is already noisy it will never get better. Unqoute
Regrettably, Allo Support, did not care to respond on any of my questions and remain quite unresponsive to this topic on their own thread except when Catana is involved.
To say bluntly, I would think twice if I am going to buy...
Galvanic isolation does not always result in complete noise decoupling. There is always some stray capacitance between the dirty side and the clean side. Noise at HF/RF may be significantly attenuated but not completely without remaining audible effects.
That problem was reported by one user with a particularly noisy computer. None of his other computers produced the same symptom.
Galvanic isolation chips have a hold off voltage rating and some physical spacing between the input and output pins. In some cases isolators chips are used along galvanic barriers designed such that clean and dirty ground planes come very close to each other over an extended circuit board area.
In other words, HF/RF isolation performance may depend on the particular overall galvanic isolation design. To be completely sure if there would be any audible advantage using a low noise clean power supply on the dirty side of a USB galvanic isolation barrier, one would need to try in under the exact circumstances where one intends to use it.
All the foregoing having been said, the very highly rated Benchmark DAC-3 does not appear to use USB galvanic isolation at all despite many other noise incursion reduction steps taken for other inputs such as SPDIF and analog audio. In other words, there is no guarantee that of absence of USB galvanic isolation will always cause any audible reduction of audio quality.
Either way it is something that may or may not cause problems. If a problem is found then the remedy used should be appropriate and suited to the magnitude and nature of the problem.
That problem was reported by one user with a particularly noisy computer. None of his other computers produced the same symptom.
Galvanic isolation chips have a hold off voltage rating and some physical spacing between the input and output pins. In some cases isolators chips are used along galvanic barriers designed such that clean and dirty ground planes come very close to each other over an extended circuit board area.
In other words, HF/RF isolation performance may depend on the particular overall galvanic isolation design. To be completely sure if there would be any audible advantage using a low noise clean power supply on the dirty side of a USB galvanic isolation barrier, one would need to try in under the exact circumstances where one intends to use it.
All the foregoing having been said, the very highly rated Benchmark DAC-3 does not appear to use USB galvanic isolation at all despite many other noise incursion reduction steps taken for other inputs such as SPDIF and analog audio. In other words, there is no guarantee that of absence of USB galvanic isolation will always cause any audible reduction of audio quality.
Either way it is something that may or may not cause problems. If a problem is found then the remedy used should be appropriate and suited to the magnitude and nature of the problem.
Last edited:
Without verification process, whether the levels were matched within 0.1 db or not is unknown. Often the audible differences in A/B comparison reported are due to level mismatch.
Well of course he's going to say that, it is what they are selling after all.
Honest question though, if you'd go so far to improve SU-1. Why don't you just buy Singxer SU-6 instead? It is powered by supercapacitors across the line...
So, you do not believe your countryman, hmm... Joking. Of course they want to sell, that is the point, right. And we are the suckers who want to buy.
Singxer SU-1 is fantastic piece, but this is beyond this thread. My music is strictly PCM, no DSD. My opinion is that the noise must be firstly take care of at the source and then along the chain.
Something like the glass of water with hole at the bottom. It is easier to plug the hole, then to wipe spilled water all over the table.
Ha- ha
Hmm yes, I also like my SU-1 (don't see need to upgrade). But I was just saying, if you buy Shanti or Sbooster or the like; might as well upgrade SU-1. There exists now SU-2(better clocks) and SU-6(temperature controlled oven + charged supercapacitors for power) from Singxer which makes more sense IMO than an LPS.
Ha- ha
Hmm yes, I also like my SU-1 (don't see need to upgrade). But I was just saying, if you buy Shanti or Sbooster or the like; might as well upgrade SU-1. There exists now SU-2(better clocks) and SU-6(temperature controlled oven + charged supercapacitors for power) from Singxer which makes more sense IMO than an LPS.
Well, I got your point. But I still think that better USB interface does not rule out the benefit of better LPSU, that is my point I tried so hard to prove.
I think I will wait with the SU6 upgrade. Speaking of the upgrades, your honest opinion about Intona in your rig?
But a good USB interface is self-powered, and does not even use the USB power, only the data. Imagine adding an LPS on an interface that pulls zero power
the SU-1 does need it though, but that was the first generation of its kind.About the Intona, I like it. It's just a hassle free USB hub. It adds galvanic isolation and also filters the power. It's made for industrial purpose, so no audiophile hocus pocus / marketing. The fact that many audiophiles buy it was just a nice surprise for the manufacturer.
BUT. Tell me to remove it from the chain and if I hear any difference. probably... no. I bought it second hand for little money though. so it's staying.
Sorry, not true with Singxer SU-1. Read also the quoted text in the post #1165
Yes, I know, I own one, and that's also literally what I said though?
the SU-1 does need it(=power), but that was the first generation of its kind.
There's now an SU- two and SU- six out as well. The SU-1 is what, 4 years old now?
SU-6 also needs 5V from the computer. What SU-6 has more besides polishing marketing data and inflating the price is the possibility of external PSU, which BTW again, prove my point that galvanic isolation is not a silver bullet and that the better LPSU can contribute a lot in suppressing the noise.
Hope we are not hijacking this thread completely...I am trying to involve Allo but no success.
In a previous post you said At last...do not use CLK division on ess IC , it seems to veil the sound (while THD barely changes)
Does this mean that synchronous mode is worse, unless 352khz rates are used?
Good question and eye
I am still doing tests and will ask from input from other listeners ...but yes thats my personal opinion .
The sound changes when CLK division is used , but we all know its hard to explain how..
I have another week though to tweak Revolution so lets see..
Meanwhile both modes will be available synch (dpll at 0)(128fs with clock division up to 192Khz and 128fs no clk division for rest) and async(no division up to 192Khz but dpll at 1 and rest in 128fs with no clk division and dpll 0 of course)
I am still doing tests and will ask from input from other listeners ...but yes thats my personal opinion .
The sound changes when CLK division is used , but we all know its hard to explain how..
I have another week though to tweak Revolution so lets see..
Meanwhile both modes will be available synch (dpll at 0)(128fs with clock division up to 192Khz and 128fs no clk division for rest) and async(no division up to 192Khz but dpll at 1 and rest in 128fs with no clk division and dpll 0 of course)
Good question and eye
I am still doing tests and will ask from input from other listeners ...but yes thats my personal opinion .
The sound changes when CLK division is used , but we all know its hard to explain how..
I have another week though to tweak Revolution so lets see..
Meanwhile both modes will be available synch (dpll at 0)(128fs with clock division up to 192Khz and 128fs no clk division for rest) and async(no division up to 192Khz but dpll at 1 and rest in 128fs with no clk division and dpll 0 of course)
Maybe sound changes with clk division due to jitter?