With a natural obsession with testing this kind of stuff, I tried a number of configurations in the network. I even had cable (Spectrum) and fiber (AT&T) internet services active at the same time to test differences.
Keep in mind, these are just my findings. Please let me know what your experiences are in the comments. YMMV.
The power supply to your modem and router have a much larger impact over your Ethernet cable.
Try plugging your equipment into a UPS battery. This is the one I use. I went a step further and connected them to a Paul Hynes SR4. The differences are night & day.
Fiber internet sounds quieter, tighter, and more detailed – but thinner. Cable internet is softer and warmer – but doesn’t have as much detail or depth. This is even obvious when streaming music to smart devices such as Google Home and Minis.
Although more subtle, changing the Ethernet cable between the modem and router – affected the sound even when the router is connecting to the Innuos via a wireless bridge.
This part didn’t make any sense to me. Especially when the bridge is galvanically isolated from the main system. The only thing I could think of is how the interaction between the radio antennas was impacted by the Ethernet cable.
Tonal differences were subtle but there’s no doubt that the “detached” Ethernet cable still impacted the music server that was connected wirelessly.
This was interesting. I feel the tonal quality is better with a direct connection – even when using generic Ethernet cable. It just sounded more solid. But the sound was pancake flat and transients were obliterated.
Using the bridge had a more “HiFi” sound. It was much quieter, had a larger soundstage, and provided much better focus. However, it was cooler, coarser, and thinner. A SOtM iSO-CAT6 did help smooth it out a bit – but adds to the cost.
I was a bit torn here. I enjoy the richer tone of the direct connection, but the delineation and outlining with the bridge was much better.
Now, what would happen if I power the bridge with a better power supply? I feel it’ll make the wireless bridge the better (and more convenient) option. But that’s an experiment for another day.
Mr. Audio Bacon himself. An open-minded electrical engineer and software developer by trade. I have an obsession with the enjoyment of all things media - specifically in the realm of music and film. So much heart and soul (and money) go into the creation of this artistry. My aim is to find out which products get me closer to what the musicians and directors intended.
Thanks for this review Jay! I've not yet pursued any ethernet cable upgrades over standard cables. Question for you: I currently have a Chord Dave and M Scaler (connected with some nice Black Cat Tron SPDIF cables that sound great to me). But I connect my M Scaler to the source (Macbook Air / Roon via USB hub that has an iFi DC iPurifier 2 for power) via USB cables (upgraded and with iFi iPurifier3 to the M Scaler and other iFi iSilencers throughout the USB system). I'm wondering if you think this USB to the M Scaler approach would be bested by some sort of Ethernet connector (like to an Ultra Rendu or maybe a Stack Audio Link)?? Also, have you upgraded your M Scaler to some other power supply and if so which one? It is indeed amazing how much difference cables can make. Currently my Dave is using a PS Audio AC12.
This will void your warranty - but I'm trying out a few battery packs with the HMS. The pros outweigh the cons so far but I'll have to do more digging. Interestingly, the DC cable used with these battery packs seems to determine tonal color more than anything.
Remote controlling your music via a music server/streamer on your network (rather than your laptop) typically yields better performance. The USB output of the server/streamers should be better than that out of a laptop.
Thanks. That is intersting about the DC cable to the HMS! I look forward to your review on this at some point. I did compare a while back my Laptop to the Roon Nucleus (same Roon software, same source content) and also to the Aurender server. The Aurender sounded somewhat (but not a lot) better, but the Roon Nucleus and MacBook Air were almost identical (with an iPurifier 3 inline) and that's with Dave sourced directly via USB (before using the M Scaler into the Dave). Something like an Ultra Rendu or other streamer/server may be in the cards for me. But I wonder is USB to the HMS optimal (given that Dave gets its signal from the HMS via SPDIF anyway) or is the HMS best sourced by SPDIF or optical or does it really matter? I will say that so far other than the obvious benefit of the HMS and the SPDIF cables, the biggest effect I've seen was the better power cable to Dave - that made a very big difference.
Wouldn't it be wiser to use sfp fiber and electricity as stable as possible like from off grid solar setup with PSU in the middle?
I'm actually going to install solar to test a few things out. Although STP "should" improve, this really depends on the switches. For the SOtM, the Ethernet ports are better.
All standard solar inverters will copy the ac frequency on the line. This is an anti islanding feature to stop you frying the lineman's working up the power pole when he turns off the power line. Just use a line interactive UPS if you want a corrected voltage.
So MRI brain scans, bank transactions, data base backups and crc checked files move over standard cat 5~6 just fine and with zero corruption or degradation. But your audio file is some how alterted? Give me a break clearing you don't understand TCP_IP or are a shill
Ok I'm sorry, but how can you seriously claim you work in any field related to ethernet and then claim that ethernet cables have any sound to them. Yes exactly, it's digital and not just digital, its error correcting, there is NO WAY that there is any difference. If they pay you for this review just tell the people. If anyone doesn't trust a normal cable, get a cat.6a cable, it's individually shielded and basically immune to interference on short runs (100m). The reason you don't want to do any measurements is not that you trust your ears more but that every measurements even before error correction would disprove everything you say. Is money really worth deceiving people that don't know better and waste 1000$ on cheap ethernet cables that cost 4$ in production? 1,50 per connector + 1$ in cable and 2-3 minutes actually assembling the cable.
Yep. Perhaps the funniest bit, though there is plenty of competition, is this. -
“ Fiber internet sounds quieter, tighter, and more detailed – but thinner. Cable internet is softer and warmer...”
There is no bigger con in the entire commercial world than that aimed at so called “audiophiles”, especially in the area of cables in their various forms.
Jay, you're contradicting 50+ years of work by thousands of engineers with an anecdotal argument. Without data to back up your claims it seems you're either preying on the ignorance of people with money by reinforcing the patterns that they're familiar with (and coincidentally spreading misinformation about how the world works) or you're not understanding how data is transferred and stored. It's easy to verify if there's any change in the data by hashing the data at each end of the transfer and comparing the results. If the hashes are the same, then the data is the same and the audio is the same. This validation removes all data transport and storage complexity from the equation and then no one needs to buy insanely overpriced cables. I'm not going to return to this site, but I'm tempted to return just to read the responses of more highly qualified and less patient people.
This is complete nonsense. I question your claim of being an EE/CE.
Even if noise is introduced into an Ethernet cable, the transceiver only cares about the bits. As long as it can discern the signal over the noise, it will send on a perfect recreation of the data. Ethernet and IP have built in error checking and frames/packets will be discarded if there are any errors introduced along the path. TCP or the application will simply retransmit the data.
No, see, the cable transforms the bits such that the hashes collide but it sounds the same. Better. So it sounds better. You should give them $1,000 for this cable.
Bob, take a look at the yamaha wxa50 for HD music streaming.
Absolute rubbish.
The 1's and 0's either get to the device or they don't. Additional 1's or 0's that aren't required will be rejected.
There will be error checking algorithms performed and the data is the data - nothing more, nothing less.
Don't believe me? Then transfer a file from a super quick new PC to an slow buggy 10yr old one and compare the two files. Are the files identical, or does using the ethernet cable add junk data or lose some of the information? The answer is neither. The data bits are checked in and checked out perfectly. That's why when you open a photo it's not magically less vibrant due to data loss or has weird characters and grain embedded.
In the digital domain you get 100% of what you put in, nothing more, nothing less.
This article is a travesty. Focus on analogue and noise reduction in the electrical paths if you want to have identifiable differences.
Stop trying to pass of this snake oil - there is nothing to debate here whatsoever. The only way your ethernet cable will make a difference is if it isn't shielded and causes some kind of ground loop him. However cheap cat5+cat6 cables are used daily in multi billion dollar data centres daily. Google and Microsoft are not rushing out to improve the data quality of their information transports for a good reason. The cable is simply the carrier of digital data which is checked in and out 1000 per second in devices with 1ms response times.
I'm so angry at this absolute con of a cable and concept.
Credentials : 25+ years being an audiophile reviewer and the same being a data centre manager
I agree with most of the statements, however, I think in "digital" transmission, on the receiving end, you either get 100% or 0% (complete and identical info as was sent or nothing at all) depending if the "packet" of information was received successfully w/o any errors or there were enough issues or errors that, even after attempting to "fix" the errors, the transmission is declared unsuccessful and "nothing" is received! So, the overall transmission is either "1" for total success or "0" for a complete failure!
You are completely correct, I work as an Automation Engineer and Shielded cables are only useful when you have A LOT of electrical noise around the cable and trust me you don't have that at home, we usually only use them inside of Motor Control Centers to communicate control equipment, and I am talking of 480 V panels with currents in the hundreds if not thousands of amperes.
An idiot is born everyday. Only few of those spend thousands of dollars to buy snake oil products.
Dumbest thing I've seen. A $1 cable works just as good. There is no difference. Hiccups don't matter! The devices have a cache that holds the entire file after a second!
This is an amazing troll post. EMI differences won't matter at all if the music is digital, transmitted to a device, and then cached there. Talking about how Ethernet cables "sound" is in the same realm as the "WiFi allergy" people: it's all imagined and in your head.
Let me guess next you'll be on about how the fiber cable quality effects an audiophile's enjoyment... Let's do a double blind experiment, 10 Grand to you if beat chance consistently, but 10k to me when you fail.
Hilarious! I loved the subtle-not-subtle sarcasm throughout the whole article.
Maybe one day audio measurement tools will finally prove once and for all that the bragging about overpriced components is for nought. Until then l guess I'll keep financing my doctor's spending obsession.
If you guys want real high quality Ethernet cables, try R&M (Reichle and De-Massari).
1000 bucks for an Ethernet is for utter fools, R&M is swiss and ultra high quality and uses IDC in plugs, not IPC like this crap, but not 1000 dollars .
This is the biggest joke of an article I have ever seen come out of an audiophile site. You explain nothing about the underlying IP stack to the OS you are using, which has way more impact to streaming audio than the Ethernet cable you are using (which has no impact, assuming you have 0 packet loss, which 99.99999999999% of cables are capable of.
And again, yes, it’s digital, it doesn’t fucking matter. Keep making up whatever you want to justify your non-sense, but it doesn’t change the laws of physics and how computers work on a fundamental basis.
If you notice any difference, it’s because of the equipment you are hooking the cables up to is crap. How many times do people with PhDs have to explain this to audiophiles?
You disgrace the term “engineer” along with those that actually do the real work.
Streaming vs downloading would not effect playback as far as different Ethernet cables go. I'm sorry but this is a cash grab my friend.
View Comments
Thanks for this review Jay! I've not yet pursued any ethernet cable upgrades over standard cables. Question for you: I currently have a Chord Dave and M Scaler (connected with some nice Black Cat Tron SPDIF cables that sound great to me). But I connect my M Scaler to the source (Macbook Air / Roon via USB hub that has an iFi DC iPurifier 2 for power) via USB cables (upgraded and with iFi iPurifier3 to the M Scaler and other iFi iSilencers throughout the USB system). I'm wondering if you think this USB to the M Scaler approach would be bested by some sort of Ethernet connector (like to an Ultra Rendu or maybe a Stack Audio Link)?? Also, have you upgraded your M Scaler to some other power supply and if so which one? It is indeed amazing how much difference cables can make. Currently my Dave is using a PS Audio AC12.
This will void your warranty - but I'm trying out a few battery packs with the HMS. The pros outweigh the cons so far but I'll have to do more digging. Interestingly, the DC cable used with these battery packs seems to determine tonal color more than anything.
Remote controlling your music via a music server/streamer on your network (rather than your laptop) typically yields better performance. The USB output of the server/streamers should be better than that out of a laptop.
Thanks. That is intersting about the DC cable to the HMS! I look forward to your review on this at some point. I did compare a while back my Laptop to the Roon Nucleus (same Roon software, same source content) and also to the Aurender server. The Aurender sounded somewhat (but not a lot) better, but the Roon Nucleus and MacBook Air were almost identical (with an iPurifier 3 inline) and that's with Dave sourced directly via USB (before using the M Scaler into the Dave). Something like an Ultra Rendu or other streamer/server may be in the cards for me. But I wonder is USB to the HMS optimal (given that Dave gets its signal from the HMS via SPDIF anyway) or is the HMS best sourced by SPDIF or optical or does it really matter? I will say that so far other than the obvious benefit of the HMS and the SPDIF cables, the biggest effect I've seen was the better power cable to Dave - that made a very big difference.
Wouldn't it be wiser to use sfp fiber and electricity as stable as possible like from off grid solar setup with PSU in the middle?
I'm actually going to install solar to test a few things out. Although STP "should" improve, this really depends on the switches. For the SOtM, the Ethernet ports are better.
All standard solar inverters will copy the ac frequency on the line. This is an anti islanding feature to stop you frying the lineman's working up the power pole when he turns off the power line. Just use a line interactive UPS if you want a corrected voltage.
So MRI brain scans, bank transactions, data base backups and crc checked files move over standard cat 5~6 just fine and with zero corruption or degradation. But your audio file is some how alterted? Give me a break clearing you don't understand TCP_IP or are a shill
Ok I'm sorry, but how can you seriously claim you work in any field related to ethernet and then claim that ethernet cables have any sound to them. Yes exactly, it's digital and not just digital, its error correcting, there is NO WAY that there is any difference. If they pay you for this review just tell the people. If anyone doesn't trust a normal cable, get a cat.6a cable, it's individually shielded and basically immune to interference on short runs (100m). The reason you don't want to do any measurements is not that you trust your ears more but that every measurements even before error correction would disprove everything you say. Is money really worth deceiving people that don't know better and waste 1000$ on cheap ethernet cables that cost 4$ in production? 1,50 per connector + 1$ in cable and 2-3 minutes actually assembling the cable.
Yep. Perhaps the funniest bit, though there is plenty of competition, is this. -
“ Fiber internet sounds quieter, tighter, and more detailed – but thinner. Cable internet is softer and warmer...”
There is no bigger con in the entire commercial world than that aimed at so called “audiophiles”, especially in the area of cables in their various forms.
Jay, you're contradicting 50+ years of work by thousands of engineers with an anecdotal argument. Without data to back up your claims it seems you're either preying on the ignorance of people with money by reinforcing the patterns that they're familiar with (and coincidentally spreading misinformation about how the world works) or you're not understanding how data is transferred and stored. It's easy to verify if there's any change in the data by hashing the data at each end of the transfer and comparing the results. If the hashes are the same, then the data is the same and the audio is the same. This validation removes all data transport and storage complexity from the equation and then no one needs to buy insanely overpriced cables. I'm not going to return to this site, but I'm tempted to return just to read the responses of more highly qualified and less patient people.
This is complete nonsense. I question your claim of being an EE/CE.
Even if noise is introduced into an Ethernet cable, the transceiver only cares about the bits. As long as it can discern the signal over the noise, it will send on a perfect recreation of the data. Ethernet and IP have built in error checking and frames/packets will be discarded if there are any errors introduced along the path. TCP or the application will simply retransmit the data.
No, see, the cable transforms the bits such that the hashes collide but it sounds the same. Better. So it sounds better. You should give them $1,000 for this cable.
Bob, take a look at the yamaha wxa50 for HD music streaming.
Absolute rubbish.
The 1's and 0's either get to the device or they don't. Additional 1's or 0's that aren't required will be rejected.
There will be error checking algorithms performed and the data is the data - nothing more, nothing less.
Don't believe me? Then transfer a file from a super quick new PC to an slow buggy 10yr old one and compare the two files. Are the files identical, or does using the ethernet cable add junk data or lose some of the information? The answer is neither. The data bits are checked in and checked out perfectly. That's why when you open a photo it's not magically less vibrant due to data loss or has weird characters and grain embedded.
In the digital domain you get 100% of what you put in, nothing more, nothing less.
This article is a travesty. Focus on analogue and noise reduction in the electrical paths if you want to have identifiable differences.
Stop trying to pass of this snake oil - there is nothing to debate here whatsoever. The only way your ethernet cable will make a difference is if it isn't shielded and causes some kind of ground loop him. However cheap cat5+cat6 cables are used daily in multi billion dollar data centres daily. Google and Microsoft are not rushing out to improve the data quality of their information transports for a good reason. The cable is simply the carrier of digital data which is checked in and out 1000 per second in devices with 1ms response times.
I'm so angry at this absolute con of a cable and concept.
Credentials : 25+ years being an audiophile reviewer and the same being a data centre manager
I agree with most of the statements, however, I think in "digital" transmission, on the receiving end, you either get 100% or 0% (complete and identical info as was sent or nothing at all) depending if the "packet" of information was received successfully w/o any errors or there were enough issues or errors that, even after attempting to "fix" the errors, the transmission is declared unsuccessful and "nothing" is received! So, the overall transmission is either "1" for total success or "0" for a complete failure!
You are completely correct, I work as an Automation Engineer and Shielded cables are only useful when you have A LOT of electrical noise around the cable and trust me you don't have that at home, we usually only use them inside of Motor Control Centers to communicate control equipment, and I am talking of 480 V panels with currents in the hundreds if not thousands of amperes.
An idiot is born everyday. Only few of those spend thousands of dollars to buy snake oil products.
Dumbest thing I've seen. A $1 cable works just as good. There is no difference. Hiccups don't matter! The devices have a cache that holds the entire file after a second!
This is an amazing troll post. EMI differences won't matter at all if the music is digital, transmitted to a device, and then cached there. Talking about how Ethernet cables "sound" is in the same realm as the "WiFi allergy" people: it's all imagined and in your head.
Let me guess next you'll be on about how the fiber cable quality effects an audiophile's enjoyment... Let's do a double blind experiment, 10 Grand to you if beat chance consistently, but 10k to me when you fail.
Hilarious! I loved the subtle-not-subtle sarcasm throughout the whole article.
Maybe one day audio measurement tools will finally prove once and for all that the bragging about overpriced components is for nought. Until then l guess I'll keep financing my doctor's spending obsession.
If you guys want real high quality Ethernet cables, try R&M (Reichle and De-Massari).
1000 bucks for an Ethernet is for utter fools, R&M is swiss and ultra high quality and uses IDC in plugs, not IPC like this crap, but not 1000 dollars .
This is the biggest joke of an article I have ever seen come out of an audiophile site. You explain nothing about the underlying IP stack to the OS you are using, which has way more impact to streaming audio than the Ethernet cable you are using (which has no impact, assuming you have 0 packet loss, which 99.99999999999% of cables are capable of.
And again, yes, it’s digital, it doesn’t fucking matter. Keep making up whatever you want to justify your non-sense, but it doesn’t change the laws of physics and how computers work on a fundamental basis.
If you notice any difference, it’s because of the equipment you are hooking the cables up to is crap. How many times do people with PhDs have to explain this to audiophiles?
You disgrace the term “engineer” along with those that actually do the real work.
Streaming vs downloading would not effect playback as far as different Ethernet cables go. I'm sorry but this is a cash grab my friend.