A relay is a switch too, just operated using a coil instead of a physical switch.
There's only one contact resistance point in SMSL, the relay. Plus, it is positioned right where it should be; the signal does not have to travel via long PCB traces to reach switches. The iFi uses an additional contact resistance point, where the header is inserted into the sockets, which will provide connectivity to the front panel placed switches. Another point of contact resistance.
NOTE: iFi may be using LS47 family logic circuits on the PCB.. in which case the signal is not routed - it is a 5V (high) that is routed. But the 74 logic will bring their own issues...
A volume control IC is MUCH more complex and adds more distortion and noise, audio passes through a lot more components than a potmeter.
Channel imbalance is bettter and easy to use remote control and volume presets.
iFi uses a Bourns (looks like Bourns) potentiometer which should be a conductive plastic film potentiometer. The wiper sets the volume by scrubbing/scratching action on top of that plastic film. This is producing noise; noise that will only become worse with time. This is vastly inferior to a volume control IC. Have a look at their (volume control ICs) specifications sheet. The actual sound path is extremely short and simple and consists of a triplet of resistors at any given setting.
I'll give you some time here to read the spec sheet and absorb it. Then, if you still believe that the volume control IC is inferior to the particular potentiometer that iFi is using (miniature Bourns), I am happy to discuss it further with you. High-quality ladder-switched volume potentiometers might be a match for volume control ICs... but these potentiometers are around $500 for a quad pot (fully balanced). This is not what we see in iFi...
A linear power supply can be better than a linear power supply.
In general SMPS do emit more common mode noise and can have a higher leakage.
Also (cheaper) SMPS have not the same longevity as as linear.
SMPS are cheaper, smaller light-weight and can have a wide input voltage range.
They are noise generators that should never be used in high-quality audio.
The 'should' part is just an assumption though.
For filters and specific circumstances polyprop can be measurable better than say an electrolytic, tantalum or ceramic multilayer. Doesn't have to be a sound determining element though.
From my experience, and from the measured results, the polypropylene film caps have lower distortion and are much better suited to sound coupling than SMD devices used in iFi (I did not see any polypropylene SMD cap on iFi PCB).
NOTE: iFi may have implemented a DC-coupled signal path... may have. Once again, I'd need to see the PCB myself.
Other circuits also can be exceptionally good as well, as you said... implementation is everything, not so much the used technique.
I know those current feedback ICs... I worked with them a lot (even with the legendary AD811 in I/V converters - extremely hard to implement properly). When implemented correctly, they are super-sounding devices. I believe SMSL managed to do a proper implementation (based on what the thread started reported here -> a much better sound experience with SMSL...)
Is that so ? and even when that would be the case why would that be an advantage ?
The longer signal path scenario must be taken into consideration from the added inductance and capacitance, as the result. The signal traces are surrounded by the ground fill. The best results are always obtained if these factors are removed from the direct signal path. SMSL has a straightforward signal path from the input towards the output.
NOTE: I am not 100% sure if iFi has a longer signal path... It looks like... I'd need to see the PCB myself.
Just a power supply rail transforming signal fidelity 'completely' ?
You need to try a good quality liner power supply: CLC input (L will remove rubbish above 100kHz), followed by a series regulator with let's say 2 TO3 transistors for high currents of around 6A, or (multiples of) LT3045/LT3094 for lower currents - around 0.5 - 2A or so. These LT devices are extremely efficient for noise frequencies up to 100kHz (see the specifications sheet). Combined with a choke (L) at the input, very broad-spectrum noise attenuation is possible. They sound exceptionally good.
Any measurements or nulls that confirm this statement ?
Read this if you are really interested:
No, this peak does not come from averaging in the scope, it is the noise (smoothed somewhat due to averaging). In a single shot the bumps are already there. IMHO, the shift of the bumps comes from the changing impedance the SMPS sees at the output, when the filter is inserted between output and...
www.diyaudio.com
... and you'll understand why I call SMPS the noise generators, and why they should never be used in high-end audio. They do have their place in high density, high voltage/high current rail ability, required for class D amps. Why not? The class D amps are already sort of a noise generator in itself, a match made in heaven.