[AussieMick]

We shouldn’t split the discussion into “science” and “belief”. I trust what I’ve heard and I know many others have heard it as well. I’d like a well designed scientific exploration into why that is, so that we can take the anecdotal evidence and use scientific principles to prove it.

The absence of evidence isn’t the same as disproof, just as anecdotes aren’t proof.

My turn to be sceptical; whenever the “it’s not science” argument is used it refers specifically to that which has yet to be proven and, troublingly, assumes all is known to such an extent that it’s not going to change. Amir does one suite of measurements, that might not even be relevant to the task, and that’s taken as “proof”. Where are the peer reviewed articles backing it up? In my experience the “science” follower often asks this of any company claiming something that might be a little different.
An example is Akiko Audio (I return to them because I’m familiar but I’m sure there are other examples). They publish the results of two sets of tests and get criticised for it not being peer reviewed. Amir publishes one test and it’s taken as truth. That seems a rather unscientific unfairness.
“Scientism” is a bit of an ugly word but it’s often appropriate. Sometimes I read posts and think there’s a little, “science, but on my terms”. Science is a great democratiser!

Observation of phenomena - thoughtfulness - design experiment - test thoughts - repeat.

We have the observation part already. The inquisitive scientific mind doesn’t dismiss on the basis of the known, it wonders if there’s something unknown and attempts to prove it.

 

[Mr BubbaHyde]

We shouldn’t split the discussion into “science” and “belief”. I trust what I’ve heard and I know many others have heard it as well. I’d like a well designed scientific exploration into why that is, so that we can take the anecdotal evidence and use scientific principles to prove it.

The absence of evidence isn’t the same as disproof, just as anecdotes aren’t proof.

My turn to be sceptical; whenever the “it’s not science” argument is used it refers specifically to that which has yet to be proven and, troublingly, assumes all is known to such an extent that it’s not going to change. Amir does one suite of measurements, that might not even be relevant to the task, and that’s taken as “proof”. Where are the peer reviewed articles backing it up? In my experience the “science” follower often asks this of any company claiming something that might be a little different.
An example is Akiko Audio (I return to them because I’m familiar but I’m sure there are other examples). They publish the results of two sets of tests and get criticised for it not being peer reviewed. Amir publishes one test and it’s taken as truth. That seems a rather unscientific unfairness.
“Scientism” is a bit of an ugly word but it’s often appropriate. Sometimes I read posts and think there’s a little, “science, but on my terms”. Science is a great democratiser!

Observation of phenomena - thoughtfulness - design experiment - test thoughts - repeat.

We have the observation part already. The inquisitive scientific mind doesn’t dismiss on the basis of the known, it wonders if there’s something unknown and attempts to prove it.

I tend to agree with your assessment, I was an avid reader of ASR but years ago I noticed the same type of thing. At that time I was always interested in the latest DAC that got the best "Measurements" and there is nothing wrong with that, nothing whatsoever!

Then one day I decided to buy a Denafrips Ares II, and I fell in love with the sound! If I remember correctly (and I may not) but there this DAC was a "poorly" measuring DAC and not "up to snuff", it was also around that time that I started to split and decided I was going to take the science into account but always trust my ears and what I enjoyed hearing.

I am still there, I am willing within "reason" (meaning my wallet, lol) to try something to see if it makes an audible difference or changes my enjoyment of a product.

Cheers!!

 

[vext01]

Amir publishes one test and it’s taken as truth. That seems a rather unscientific unfairness.

Although Amir's experiment looks pretty rigorous, Mick is right: more people should measure this.

Ifi, can you please send me one?

 

[rdale]

Although Amir's experiment looks pretty rigorous, Mick is right: more people should measure this.

Ifi, can you please send me one?

I remember iFi responding to some review at ASR measuring the performance of one of their USB noise filters where they said the level of skill needed to get meaningful measurements was very high, even when you have the best possible test gear. I got the impression it was almost as hard to measure products like this as it is to design them in the first place. As I’ve got a large number of iFi products already (along with a couple of LAN iSilencers on order too that I’m looking forward to trying), I have plenty of confidence that they know what they are doing.

 

[vext01]

they said the level of skill needed to get meaningful measurements was very high

How convenient!

 

[rdale]

I remember iFi responding to some review at ASR measuring the performance of one of their USB noise filters where they said the level of skill needed to get meaningful measurements was very high, even when you have the best possible test gear. I got the impression it was almost as hard to measure products like this as it is to design them in the first place. As I’ve got a large number of iFi products already (along with a couple of LAN iSilencers on order too that I’m looking forward to trying), I have plenty of confidence that they know what they are doing.

I found this paper on the iFi site about how difficult it is to measure noise in digital circuits:

https://ifi-audio.com/wp-content/uploads/2018/07/iPOWER-Much-Ado-About-Nothing.pdf

I’ve just heard that my order for a couple of LAN iSilencers has been delayed until mid to late April because of high demand, and reading about measuring noise isn’t quite the same as trying them out

 

[mfdoom808]

 

[Moonglum]

Set your audio free with blissful, noise-free sound​

When used for audio, LAN interfaces are inherently ‘noisy’. iFi’s LAN iSilencer removes electrical noise that distorts the audio signal to unleash the connected equipment’s sonic potential.

Southport, England – In recent years, more and more music lovers have taken to streaming music as their primary audio source. These streamers are often connected to audio gear via network routers. There are pros and cons to this. Ethernet interfaces can pass hi-res audio data at maximum resolution and, allow bit-perfect conveyance of the original source data – which is ideal, on paper. In practice, LAN is a complex, multi-purpose computer interface – not one that was created specifically for audio. It is usceptible to electrical noise, not least from the router’s power supply, which can adversely affect sound quality.

iFi makes several devices that enhance audio quality. Two of the most popular are the iSilencer+ and iDefender+. Joining that illustrious family is the new LAN iSilencer, that boasts premium galvanic isolation technology, where electrical circuits are separated to eliminate stray currents.

The LAN iSilencer features Gigabit Ethernet support, with no speed restrictions. It is constructed using a shielded socket and plug with integral ground connections, that blocks external interference, minimises packet loss and ensures stable signal transmission.

Whilst isolating and removing system noise, the LAN iSIlencer reduces the overall jitter in the system – a form of digital distortion that leads to timing errors and results in cold, harsh sound. It also re-balances the signal via iFi’s REbalance technology, ensuring that the digital signal received by the streamer is identical to the original source.

The LAN iSilencer can be connected to a router, network switch or ethernet wall socket. No power source is needed. The precise effect of the LAN iSilencer on sound quality varies from system to system; typically, one can expect to hear more clarity, greater dynamic extension and sharper leading edges, making the music sound more vivid and open.

The iFi LAN iSilencer is available from Amazon and other selected retailers from today, priced at £89, (€89, $89).

I tried one of these connected to my BT router. Felt it smoothed out the sound, made it less digital sounding with a blacker background. Bought a second and connected to my EE 8 switch and that amplified the observed effect. Nice ! Using an Innuos Zen mk3 streamer and my cables are high end Audioquest ones - I have linear power supply on my switch and on my router and take streaming seriousl. This is a good value tweak in my opinion. Over on the Innuos Facebook group there are folks even with Innuos Phoenix Net high end switch who added this little ifi product and reported an uptick in sound in a similar way to what I report above - less ‘digital hash’. Bravo ifi !

 

[Mr BubbaHyde]

I tried one of these connected to my BT router. Felt it smoothed out the sound, made it less digital sounding with a blacker background. Bought a second and connected to my EE 8 switch and that amplified the observed effect. Nice ! Using an Innuos Zen mk3 streamer and my cables are high end Audioquest ones - I have linear power supply on my switch and on my router and take streaming seriousl. This is a good value tweak in my opinion. Over on the Innuos Facebook group there are folks even with Innuos Phoenix Net high end switch who added this little ifi product and reported an uptick in sound in a similar way to what I report above - less ‘digital hash’. Bravo ifi !

Thanks for the feedback, and your specific use situation!

We appreciate it!

Cheers!!

 

[mfdoom808]

I tried one of these connected to my BT router. Felt it smoothed out the sound, made it less digital sounding with a blacker background. Bought a second and connected to my EE 8 switch and that amplified the observed effect. Nice ! Using an Innuos Zen mk3 streamer and my cables are high end Audioquest ones - I have linear power supply on my switch and on my router and take streaming seriousl. This is a good value tweak in my opinion. Over on the Innuos Facebook group there are folks even with Innuos Phoenix Net high end switch who added this little ifi product and reported an uptick in sound in a similar way to what I report above - less ‘digital hash’. Bravo ifi !

Could it be placebo maybe?

 

[kingoftown1]

I thought I got hit by a car and flung into a ditch one time, but it turned out to just be placebo.

 

[mfdoom808]

As has been said many times already, nobody, not one single person, is saying it has anything to do with data. It doesn’t. The data is unaffected. This device does not affect data accuracy.
Digital signals are transferred via analogue means. Those electrical representations of one’s and zeros are anslogue and as such are not immune to having noise ride on the line. Electrical noise and RF and EM. That noise does NOT corrupt data - as you correctly point out the protocols are extremely effective at ensuring data integrity.
However, they don’t filter out all of the noise accrued along the way and that noise makes it’s way into the anslogue circuitry further down the line, be it the DAC, preamp, power amp or associated cabling.
This noise is very audible as glare, greyness, haze, lack of transient definition, compromised low level definition/resolution, or a combination of these.
I don’t personally know if this product works because I haven’t personally tried it, and until you try it you don’t either. Accusing a company of fraud is a serious thing, especially when you’ve not tried it. You also criticise it for not being able to do something it’s not even designed to do.
However, there is more than enough evidence around and reading available to suggest that this product belongs within the group of Products designed to tackle noise in hifi systems. On the basis of that it’s worth a try if you don’t already have your system sorted out in this area (power conditioners/regenerators, power cables, equipment isolation, RF and EM isolation).

Please stop dismissing products like this on the basis of data corruption.

It’s not about data corruption.
It’s not about data corruption.
It’s not about data corruption.

Ethernet is galvanically isolated by design. It's in the 802.3 standard: https://grouper.ieee.org/groups/802/3/af/public/mar00/stapleton_1_0300.pdf

 

[kingoftown1]

Ethernet is galvanically isolated by design. It's in the 802.3 standard: https://grouper.ieee.org/groups/802/3/af/public/mar00/stapleton_1_0300.pdf

i cant comment specifically about what this device does, but Galvanic Isolation with ethernet is physically implemented as a high frequency transformer on each pin. The reason it’s a high frequency transformer is to pass through the data signal (which is a high frequency) but block DC and low frequency noise. They don’t block frequencies higher than the Ethernet signal as they are ‘high pass’ transformers, not ‘band pass’ transformers tuned for just the Ethernet frequency. Therefore, all RFI mixed in with the Ethernet signal will pass through.

 

[mfdoom808]

The transformer isolates the power supplies on the sender and receiver ends, this way you won't have any conducted emissions affecting the receiving end, like common-mode currents for example. Whatever high-frequency EMI you have on the data lines only affects the data lines themselves and won't affect the power supply feeding the ethernet controller chip, especially since the data lines are implemented as differential signals. Besides, the signals themselves are pretty much square waves in the MHz range, which per se have a very wide harmonic content, making what you claim as "EMI" to be completely insignificant from a perspective of coupling to the rest of the circuit. Since the subject here is not really data corruption, I don't see how this product would affect the sound otherwise.

I won't get into the psychological considerations here because that's outside my academic background, but it's widely studied that we can hear differences by simply believing they will be there.

A much more interesting approach than what has been reported by customers in this thread would be someone unknowingly switching between two similar sources, with and without the product, and then identifying with proper statistical significance the source that has this product installed. More simplistic reports like the ones in this thread where someone installs it and hears a difference have no value on discerning whether this product does something or not. It's the same as putting a stone over your equipment and swearing you hear a difference: You may hear it, but is it really there?

 

[Moonglum]

Could it be placebo maybe?

Well perhaps. I do find the discussion of placebo and confirmation bias on audio forums amusing though as I am a Psychology Professor at a leading European Psychology department and have published peer reviewed papers on cognitive biases. Every time I jump into these threads about tweaks, cables etc I get ‘schooled’ by others about these things or told we should be doing the equivalent of double blinded randomised experiments. Makes me smile - how many actually do that when auditioning hi fi gear ?! I mean I know how to, I conduct experiments day in day out and analyse them using inferential statistics like ANOVA and regression models. Still to me there remains value in subjective opinion.

A long answer to the question, forgive me. So yes, placebo, possibly! Confirmation bias ? No not really as tbh I approached this tweak with a rather sceptical attitude and was rather taken aback by its apparent impact. Is its effect measurable? No idea and the devil’s in the detail of knowing what to measure and then employing suitable methods. Will Amir at ASR measure it and claim it debunked ? Possibly! Will I care ? Nope ! I happily use some DACs that he claims measure badly. They sound great !